Athenian Democratic Origins: and other essays

ATHENIAN DEMOCRATIC ORIGINS

This page intentionally left blank

Athenian Democratic Origins and other essays G. E. M. DE STE. CROIX Edited by David Harvey and Robert Parker With the assistance of Peter Thonemann

1

3

Great Clarendon Street, Oxford ox2 6dp 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 in Oxford New York Auckland Cape Town Dar es Salaam Hong Kong Karachi Kuala Lumpur Madrid Melbourne Mexico City Nairobi New Delhi Shanghai Taipei Toronto With offices in Argentina Austria Brazil Chile Czech Republic France Greece Guatemala Hungary Italy Japan Poland Portugal Singapore South Korea Switzerland Thailand Turkey Ukraine Vietnam Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries Published in the United States by Oxford University Press Inc., New York ß The Estate of G. E. M. de Ste. Croix 2004 The moral rights of the author have been asserted Database right Oxford University Press (maker) First published 2004 First published in paperback 2005 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, 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 book in any other binding or cover and you must impose this same condition on any acquirer British Library Cataloguing in Publication Data Data available Library of Congress Cataloging in Publication Data Data available Typeset by SPI Publisher Services, Pondicherry, India Printed in Great Britain on acid-free paper by Biddles Ltd., King’s Lynn, Norfolk ISBN 0–19–925517–2 ISBN 0–19–928516–0 (Pbk.)

978–0–19–925517–7 978–0–19–928516–7 (Pbk.)

1 3 5 7 9 10 8 6 4 2

Contents

Editors’ Introduction

1

1. The Solonian Census Classes and the Qualifications for Cavalry and Hoplite Service

5

(i) The Late Fifth- and Fourth-Century Evidence (ii) Military Categories in the Fifth and Fourth Centuries (iii) Tele¯ and Military Categories in the Fifth and Fourth Centuries (iv) The Original Qualifications of the Tele¯ (v) The Original Qualifications of the Pentakosiomedimnoi (vi) The Original Qualifications of the Hippeis and Zeugitai (vii) Were the Qualifications of the Tele¯ Changed? (viii) Pollux VIII 130 (ix) Miscellaneous Endnotes

Appendix: The Dedication of Anthemion (Arist., Ath. Pol. 7.4) Afterword

2. Five Notes on Solon’s Constitution (i) (ii) (iii) (iv) (v)

The Introduction of Majority Voting at Athens The Date of Solon’s Nomothesia The Eupatrid Monopoly of the ‘State Machine’ The Solonian Council of Four Hundred ’ prokr
itvn [‘sortition from pre-selected
rvsiB ek klZ candidates’] (Ath. Pol. 8.1) Appendix: Election to High Office by Lot Elsewhere than in Greece Afterword

3. Solon, the Horoi and the Hektemoroi Afterword

8 14 19 28 32 46 51 56 60 63

70 72

73 73 75 80 83 89 104 107

109 127

vi

Contents

4. Cleisthenes I: The Constitution (i) Cleisthenes in Herodotus and Aristotle (ii) The Constitution of Cleisthenes Endnotes Afterword

129 129 136 172 177

5. Cleisthenes II: Ostracism, Archons and Strategoi

180

(i) The Law of Ostracism: Its Date and Purpose (ii) Archons and Strategoi

180 215

Afterword

6. The Athenian Citizenship Laws Afterword

7. The Athenaion Politeia and Early Athenian History (i) (ii) (iii) (iv)

The Athenaion Politeia and the Politics The Sources of the Athenaion Politeia Aristotle and the Atthidographers Aristotle and the Documentary Sources Endnotes Afterword

229

233 251

254 256 277 286 307 323 325

8. The metra in Aristotle, Eth. Nic. V vii 5, 1134b35–1135a3

328

Appendix: Measures of Capacity

341

Afterword

9. How Far was Trade a Cause of Early Greek Colonisation? Afterword

10. But what about Aegina? Afterword

347

349 367

371 411

Contents 11. Herodotus and King Cleomenes I of Sparta Afterword

vii 421 438

Abbreviations and Bibliography

441

Index

447

This page intentionally left blank

Editors’ Introduction

The core of this book is a collection of essays on which Geoffrey de Ste. Croix was working in the 1960s. They were interconnected in theme, and as numerous cross-references show were envisaged as a single book. A colleague rashly referred to them in 1966 as ‘to be published shortly’;1 and the essays that now appear as chapters 1–2, 4–5, and 7 existed by then as typescripts with full notes. Chapter 8 too was very likely designed for a place in the volume, since it connects thematically with chapter 1. Publication was delayed by the circumstance that one chapter, entitled ‘The Causes of the Peloponnesian War’, grew into the large book The Origins of the Peloponnesian War (1972). Ste. Croix would probably have returned to the essays on completion of Origins of the Peloponnesian War, had not an invitation to give the J. H. Gray lectures for 1972–3 in Cambridge intervened; from that lecture course emerged the gigantic Class Struggle in the Ancient Greek World (1981), and the old interests in Christianity and the Later Roman Empire revived by the writing of that deceptively-titled book claimed all Ste. Croix’s attention for the rest of his life.2 The essays, however, still appear as ‘forthcoming’ in the authorised bibliography included in the Festschrift for Ste. Croix.3 In that list, the essays mentioned above have been joined by two that exist only as working drafts but may always, given their themes, have been envisaged as forming part of the book (chapters 3 and 6), and by four on different topics. We here publish all these except one, which is quite different from the others in period and theme (a lecture on ‘Some Greek views on the origin of man and civilisation’). We have grouped at the start, and in the intended order, those that certainly or possibly had a place in the volume planned 1

Forrest, EGD 245. For a biographical sketch, with a list of published obituaries, see Parker in PBA 111, 2000 Lectures and Memoirs, 447–78, or more briefly in the forthcoming Oxford Dictionary of National Biography. Those citing his works are asked to note that the full point after Ste. in Ste. Croix was defended with great vigour against editors by the name’s owner. 3 Crux, xii. 2

2

Athenian Democratic Origins

in the 1960s.4 Chapters 4 and 5 have been created by division of what Ste. Croix designed as one huge chapter entitled ‘Cleisthenes, ostracism, archons and strategoi’. The title of the volume, too, is ours: for Ste. Croix these were just ‘Essays on Greek History’. The subject of these chapters was one of the highest significance to Ste. Croix, most democratic of Marxists. He used to speak with amused frustration of an Indian pupil who would regularly conclude his undergraduate essays with the thought ‘And so we see that it all came to the same in the end.’ Such was not the attitude of Ste. Croix. The reader will regularly find in these chapters the language not just of change but even of progress: the reforms of Cleisthenes, for instance, brought ‘three quite different changes for the better, operating from the very first’ (p. 142). And things got better by design, not through happy chance. A second leitmotif of the essays is a defence of the political intelligence of the Athenians against their many ancient and modern critics. ‘Again, the good sense of the sixth-century Athenians shows itself conspicuously’, we read, or ‘Again we have vindicated the good sense of the Athenians.’ The influence in all this of the great liberal historian of Greece, George Grote, is obvious and explicit. But if the ultimate aim of the essays is to school the political intelligence of their readers, their method is scholarship of a detailed and rigorous kind. They are largely exercises in ‘source criticism’ (Quellenkritik) in its twin branches: study of the ancient evidence with a view to establishing who said what, and on what authority; and criticism of the data thus secured in the light of the observable political behaviour of human actors. Much of the genius of Ste. Croix lay in the energy, thoroughness and realism with which he applied this second criterion even to technical topics which offer little purchase to lazier imaginations. Again and again current assumptions are exposed, when once they are made explicit, as slack and implausible. Mommsen once wrote in rather apocalyptic tones of the great Niebuhr, always accounted the founder of ‘source criticism’ in classical studies: ‘He was the first to have the courage to test historical tradition by the logic of facts, 4 It is in fact possible that Chapters 9 and 10 were also envisaged even in the sixties as part of the Greek Essays, because they link thematically with the originally planned chapter on the causes of the Peloponnesian war. But the order envisaged in 1985, with what is now chapter 10 appearing first, was certainly not that intended in 1965.

Editors’ Introduction

3

to exclude what is in itself impossible from the gloomy waste of misunderstood and incomprehensible tradition, to postulate what is required by the necessary laws of development even where in the tradition it is confused or has vanished.’5 It is in that critical tradition that Ste. Croix stands. One might also say of him what Wilamowitz said of Grote: ‘Here at last was political history handled by a man who understood politics.’6 The other essays are of diverse origin, but they exhibit the same virtues. The piece on Greek trade and colonisation (ch. 9), delivered to a student society, demolishes facile assumptions about the nature of Greek trade and its role in early Greek settlements abroad. It remains valuable despite—or in view of—subsequent debate and research, which has ranged from the reassessment of concepts to the publication of fresh excavations. The essay on Aegina (ch. 10) mounts a multi-pronged attack on the view that the ruling class of Aegina was—uniquely—a mercantile aristocracy. That remains the standard view, and has rarely been challenged or examined as thoroughly as it is here. Finally, the lecture on Cleomenes (ch. 11) makes a powerful case for attributing a coherent anti-Persian policy to the maligned Spartan king. It was written for a schools audience, and comes closer to catching the characteristic tone of voice of its author than anything else in this volume. Ste. Croix’s Nachlass was passed to us by his widow Margaret after his death in February 2000.7 It was immediately clear both that the Greek essays ought to be published, and that they should not be updated except in trivial particulars. Only a very temerarious person would attempt to predict what new argument or new item of evidence might, or might not, have persuaded Geoffrey de Ste. Croix to change his mind. But we have indicated in ‘Afterwords’ the main contributions that have appeared since he laid the essays aside. These are necessarily incomplete; omissions indicate the limitations of the editors’ knowledge, not silent condemnation. To modernise the footnotes would have been less hybristic, but 5

Citation from K. Christ, Von Gibbon zu Rostovtzeff, 3rd edn. (1989), 49. U. von Wilamowitz-Moellendorff, History of Classical Scholarship, tr. A. Harris (1921/1982), 153. 7 The early Christian material was subsequently passed by us to Professor L. M. Whitby of the University of Warwick. It is hoped to bring out a volume of Ste. Croix’s writings on early Christianity, including some unpublished pieces. 6

4

Athenian Democratic Origins

here too the advantage to the reader of knowing whose mind made the decisions seemed to outweigh the gains that revision might have brought. All that we have done silently (apart from a few trivialities) is to replace references to outdated editions with those that are now in use. We have omitted one or two now outdated ‘Long Notes’ of merely bibliographic character, or which attacked new theories of the ’50s and ’60s which no one has taken up. We have also added translations. Other interventions are indicated by square brackets, and are for the most part occasioned by a note such as ‘check’ or ‘more’ in Ste. Croix’s typescript. Our editorial task has been very greatly eased by prompt and expert responses to various enquiries from Dr J. Black, Professor J. K. Davies, Dr R. P. Duncan-Jones, Dr P. Haarer, Professor G. A. Holmes, Dr C. J. Howgego, and Professor R. G. Osborne. On behalf of Ste. Croix we also thank Dr Duncan-Jones for the advice quoted in the Afterword to chapter 8. We owe particular gratitude to Professors P. J. Rhodes and C. M. Reed for supplying Afterwords to chapters 7 and 9, and to Professor P. A. Cartledge for good will and judicious advice throughout. For secretarial assistance we thank Mrs Mary Lale. Hilary O’Shea’s enthusiasm for the project at OUP has been most encouraging, and we are grateful also to Lavinia Porter, Enid Barker, and Dorothy McCarthy of the Press; also to our excellent copy-editor (John Cordy) and indexer (Rosemary Dear). A research grant to Parker from the Faculty of Classics in Oxford permitted professional transcription of most of Ste. Croix’s typescript onto disk, and the employment of a graduate assistant to enter and translate the Greek and perform a thousand other tasks in tidying the result. Peter Thonemann (then of Lady Margaret Hall; now of All Souls College) fulfilled that function with exemplary care and acuity, and some addenda or corrigenda provided by him are distinguished by his initials PJT.

1 The Solonian Census Classes and the Qualifications for Cavalry and Hoplite Service

In 1893 Eduard Meyer observed that the problems concerning the Solonian census classes appear simple but are in reality very complicated and difficult to solve.1 Meyer himself seems to have made no special study of the subject, and since his time there has been no attempt to deal with it comprehensively, although many scholars have given general opinions about it and several have written on individual points. I am not satisfied with any solution yet offered, and I wish to propose a radical reinterpretation of the evidence, which will also deal with the qualifications for cavalry and hoplite service, a subject which has not been treated in any detail, or properly linked with the census classes. As will become clear at various points in these essays, I feel that too many modern scholars conceive the economic situation even in fifth- and fourth-century Greece, and still more in the archaic age, in a thoroughly anachronistic manner, largely in terms of categories derived from the modern or the medieval world which have no application to ancient Greece. I would regard, for example, any theory that the Athenian census classes could ever have been defined in terms of money income as a sign of a serious misunderstanding of ancient Greek economics. If in this first essay I devote much more space to the problems of the census classes than many people might have thought necessary, it is partly because I believe that a proper understanding of the census classes (t
elZ) may make it possible to get rid of some basic misconceptions about the economy of classical Athens.

1

GdA II1 656, repeated in III2 (1937) 608 n. 1.

6

Athenian Democratic Origins

It is not necessary for me to attempt to justify in detail in these essays my general conception of the character of the constitutional reforms of Solon, since my own view is very close to that of a whole group of scholars, among whom I would single out Jacoby and Wade-Gery; and for the defence of my general attitude to Solon I can rely upon their work. The fundamental point has been well put by both these scholars. As Wade-Gery has said,2 ‘Solon’s most drastic political change was to ignore the Eupatrid Order.’ Jacoby, in an article entitled ‘GENESIA’,3 which has attracted less attention from historians concentrating on archaic Attica than his Atthis and his great commentary on the Atthidographers, has a paragraph which seems to me profoundly true. ‘Seen historically’, he says, ‘the tax-classes of Solon’s constitution . . . are obviously a step forward in the development of the Athenian State: they signify the emancipation of the plebs (this word still remaining the best & translation of plZuoB which is not a proletariat either in Athens or in fourth-century Rome), . . . and they are the preliminary condition for the formation of a homogeneous citizenship . . . I do not believe in a pre-Solonian timocracy,4 although I confess that I do not know at all whether the Clan State of the seventh century B.C. made personal, as distinct from real, property serve the purposes of the State; nor whether it conceded a share in the government to some ‘plebeian’ families. I consider these two assumptions to be ’ improbable. The formula arist
indZn kai ployt
indZn [‘according to birth and wealth’] (Arist., Ath. Pol. 3.1; 6) does not prove anything; it is not old, and wealth is only another aspect of aristocoploytoi [‘men of ancestral wealth’] racy;5 it is the genuine palai
’ who hold the offices and form the State in the arxa
ia polite
ia [‘ancient constitution’]. But even if in isolated cases plebeian families were received into the ruling classes before 594/3 B.C.,6 that would not alter the fact which in my opinion is certain, that it was Solon who achieved the fundamental change by making property instead of blood the standard of political rights. Principles always have a previous history, actual or ideal, the decisive step is their being publicly acknowledged: the change from the ‘Patriciate’ to the ‘Nobility’ begins with Solon. During the sixth 2 4 5 6

3 EGH 86 ff., at p. 101. In CQ 38 (1944) 65 ff., at p. 74. Here Jacoby refers to Busolt-Swoboda, GS II 824, and Adcock in CAH IV 47. And see the similar opinion expressed by Wade-Gery, EGH 100–1. See Endnote i (p. 63 below).

The Solonian Census Classes

7

century the ruling class becomes more and more what in France and Alsace one called the Notables who absorb the aristocracy without destroying it.’ In recent years a number of historians have carried mistrust for the sources for early Athenian history to what seem to me excessive lengths. In a later essay7 I shall examine some of these sources in detail. At this point I need only say two things. First, I think there are good reasons for attaching more weight to the statements about Athenian constitutional history made by Aristotle than by any of the Atthidographers who preceded him; but he must not be slavishly followed: every statement in the Politics or the Ath. Pol. must be examined separately on its merits. Secondly, there is no ground whatever for denying, as several scholars have done during the last few years, that Solon’s n
omoi included laws materially changing the constitution, the polite
ia, as well as laws which did not affect the constitutional framework but functioned within it.8 We have every right to speak of a Solonian politeia. The four tele¯, the pentakosiom
edimnoi, i‘ ppe&iB, zeyg&itai and & uZteB (terms I shall as a rule simply transliterate, as Pentakosiomedimnoi, Hippeis, Zeugitai and Thetes), were created as such by Solon, whose constitution was promulgated in 594/3, the year of his archonship.9 Some would take the tele¯ back into the seventh century, to Draco or beyond; but this seems to me certainly wrong. To conceive the origin of the tele¯ as such as pre-Solonian is to rely entirely either upon conjecture or upon the bogus10 constitution of Draco inserted in chapter 4 of Aristotle’s Athenaion Politeia and the purely consequential addition of the words ‘just as they were divided before’ in Ath. Pol. 7.3. As will become clear later, I believe that with the exception of the Pentakosiomedimnoi, whose name is of a different kind from the others, the tele¯ already existed as recognisable military and social groups. Down to the last days of the Athenian democracy as we know it (i.e. until 322 B.C.), the four census classes remained in existence, at least in theory, as the basis of qualification for certain political offices: this was certainly their chief original function, but by the fourth century, and perhaps earlier, the rules were little regarded. 7

8 9 Ch. 7. On this point, see esp. pp. 310–17 below. See ch. 2 (ii). That the ‘constitution of Draco’ is a forgery has been so amply proved, and is now so generally admitted, that I can regard it as a fact. It will be sufficient to refer to the admirable analysis by Fuks, AC 84–101, where a full bibliography is given. 10

8

Athenian Democratic Origins

The classes also entered into certain administrative regulations, in ways that we shall see. In the early days, they may well have had something to do with the taxation system. Many people have believed that they also provided—always, or only at the beginning or later—the foundation of the military system, so that the Hippeis (and Pentakosiomedimnoi) provided the state cavalry, the Zeugitai were the hoplites, and the Thetes served either as light-armed (cilo
i) or as sailors in the fleet. In the late fifth and fourth centuries, for which our information is so much more plentiful than for the sixth and early fifth, there are a few references to the census classes in relation to contemporary affairs. It will be well to look at this evidence first, as it is naturally more reliable than statements about the origin of the classes, which as we shall see could not have been based upon any very secure foundation. We can divide these passages into four distinct groups: (1) those concerned with qualifications for political office, (2) those relating to colonies or cleruchies, (3) those mentioning administrative regulations which distinguish between the respective tele¯, and (4) those which refer—or may refer—to the tele¯ in connection with military or naval service. I shall review these in turn.

(I) THE LATE FIFTH- AND FOURTH-CENTURY EVIDENCE 1. (a) According to Aristotle,11 when the question po&ion t
eloB tele&i [‘Of which census class are you a member?’]12 was put in ’ Zn [‘a candidate his own day to t on m
ellonta klZro &ysua
i tin ’ arx
for allotment to office’], no one would say that he was a Thes. This may suggest that many offices were still open in theory to none but the three top classes; but only of the nine archons can we say that this is partly confirmed by other evidence,13 and only in one case does Aristotle specifically mention membership of a telos as a qualification: the treasurers of Athena, says Aristotle,14 still had ’
in theory to be Pentakosiomedimnoi, although in practice arxei d’ ’ p
o‘ lax vn k an any p
enZBfi Z
[‘the man allotted holds office even if 11

12 Ath. Pol. 7.4 fin.; cf. Poll. VIII 86. See Endnote i i (p. 64 below). i.e. the fact that Arist. (Ath. Pol. 26.2) records the opening of the archonship to the Zeugitai in 457/6 and does not mention the admission of the Thetes. 14 Ath. Pol 8.1; 47.1. 13

The Late Fifth- and Fourth-Century Evidence

9

he is extremely poor’] (Ath. Pol. 47.1). Isaeus VII 39, discussed in section (b) below, suggests that in the first half of the fourth century certain offices were limited to Pentakosiomedimnoi and Hippeis. Aristotle, Ath. Pol. 7.3, also speaks of the census classes as being concerned under Solon’s constitution in the qualification of the Poletai, the Eleven and the Kolakretai, as well as the archons and treasurers, but there is no evidence how long this situation continued. My own belief is that long before Aristotle’s day the necessity for these magistrates to belong to one of the higher classes had been legally removed.15 Aristotle’s statements about the situation in his own day are often wrongly interpreted. Thus Busolt16 believed him to be saying, in Ath. Pol. 47.1, that even very poor men were able to ‘belong to the class of Pentakosiomedimnoi’ (‘geho¨ren zu den Pentakosiomedimnoi’), and the same view has been taken by many other scholars, whereas in fact Aristotle is surely saying that the poor men in question are accepted even if they are clearly not Pentakosiomedimnoi. In Ath. Pol. 7.4, quoted above, he implies that the candidate for an archonship might give a false answer to the question po&ion t
eloB tele&i [‘Of which census class are you a member?’]; and by analogy a candidate for the treasurership of Athena would say he was a Pentakosiomedimnos even if in reality he was far from being one. The only alternative is to suppose that breaches of the law were openly winked at, and that the aspiring treasurer admitted he was only a Zeugites, for example, but was accepted nevertheless, perhaps in the absence of properly qualified candidates. At any rate, Busolt’s position is patently untenable. In the mid-fourth century, as we shall see in a moment, a rich Athenian, Pronapes, could be accused in court of giving himself a low17 property-assessment and yet aspiring to hold office as if he belonged to the class of Hippeis;18 and this clearly implies that at 15 I base this view on Aristotle’s failure to mention the question of telosmembership in Ath. Pol. 47.2 and 52.1 (sortition of the Poletai and the Eleven respectively), and his specific references in 8.1 and 47.1 to the theoretical necessity for the treasurers of Athena to be Pentakosiomedimnoi. (The Kolakretai had disappeared in the late fifth century.) 16 Busolt-Swoboda, GS II 838 n. 1—twice, near the beginning and the end of the first paragraph of the note. 17 ’ The word mikr
on [‘low’] in apegr
acato . . . t
imZma mikr
on is to be understood in a relative sense, Pronapes being obviously a man of wealth. 18 Isae. VII 39.

10

Athenian Democratic Origins

that time a Hippeus would still be a man of considerable property. How early the situation described in Ath. Pol. 7.4 and 47.1 had arisen we do not know. The only other piece of evidence I can think of which may be relevant is the Old Oligarch’s statement that ’ vn in his day all Athenians thought they should tvn & arx & mete&inai ’ tfi Z& xeiroton
ifia [‘have a share in office by lot e’
n te tfiv& kl
Zrfiv kai en and by election’].19 This certainly suggests that the situation depicted by Aristotle may have come into existence already, or had at least begun to develop. (b) In Isaeus VII 39 (a speech delivered c. 354) we find the allegation that ‘Apollodorus did not, like Pronapes, assess his t
imZma at a low value and yet aspire to hold office as if he were a member of the class of Hippeis’. (This passage, as I have explained elsewhere,20 is often mistranslated, as for example in the Bude´ and Loeb editions.) We shall see later on that we must distinguish between what I call ‘fiscal-time¯ma’, the return of property upon which eisphora was calculated, and ‘political-time¯ma’, any valuation of property determining a man’s telos,21 and that the time¯ma which Pronapes is accused of undervaluing is fiscal-time¯ma. And in any event, there is no necessary implication that Pronapes has actually returned a political-time¯ma high enough to put him in the ‘ i‘ pp
class of Hippeis: the speaker merely says that vB ada de telvn & ’
rxein Zj
’ ioy t ’ aB [‘he aspired to hold office as if he were a a aB arx
member of the class of Hippeis’], and this need not mean more than that Pronapes, when applying for office, represented himself to be a Hippeus—compare section (a) above, where we saw that misrepresentation of their tele¯ by candidates for office apparently did take place. (c) Plato, Laws III 698B,22 uses in relation to the constitution ’ timZm
’ of Athens during the Persian wars the phrase ek atvn arxa
i tineB tett
arvn [‘certain offices were held on the basis of the four time¯mata’]. It is interesting that we should find the time¯mata (i.e. the system of tele¯) mentioned as qualifications for office only. 19

Ps.-Xen., Ath. Pol. I 2. I would date this work either during the first half of the Archidamian war or in the years following the Peace of Nicias [see Ste. Croix, OPW, 308–10]. 20 DTAEFC 43–4. 21 In fact this distinction is purely notional, since I do not admit the existence of ‘political-time¯ma’, but it is a useful one just the same. 22 Cf. V 744B–5A.

The Late Fifth- and Fourth-Century Evidence

11

’ t^ (d) It is just possible that we should restore ek on [hipp
eon] [‘from the [Hippeis]’] in IG I3 82.18, an inscription of the year 421/0. The Hipparchs and Phylarchs who commanded the Athenian cavalry would of course have to be cavalrymen themselves, but we never hear of any connection between them or the Strategoi (or any other military officers) and the census classes, or of any property qualification for holding any of these military posts, except in the unhistorical ‘constitution of Draco’.23 The fact that we never hear of any rule that the Strategoi—first created in 501/0, I believe, as a General Staff and not as tribal commanders24—had to be drawn from one of the higher classes may perhaps afford some slight indication that the census classes were already becoming less important by the end of the sixth century. The Strategoi would always of course belong at least to the zeugite class.25 2. (a) The rider of Phantocles to the Brea decree, of the midfifth century,26 allows colonists to go to Brea ‘from Thetes and Zeugitai’. The purpose of this rider cannot have been to exclude the Pentakosiomedimnoi and Hippeis, as has often been supposed. No member of one of the two highest classes would want to emigrate, especially to such an uncivilised area as Thrace, and lose his Athenian citizenship—unless of course he went as oikistes, in which case he could hope for heroic honours after his death. We can only suppose that the lost earlier part of the main decree provided that Thetes alone should participate in the colony:27 the order of the words in the rider, mentioning the Zeugitai after the Thetes, is in favour of this. Or perhaps the colonists in some other recent case or cases had been drawn from the Thetes only, and Phantocles wished to make it clear that this was not a precedent to be followed. (b) In an Athenian decree which seems to me to be of the year 387/6, preserved only in sadly mutilated fragments yielding no continuous text, the word pentakosiomed
imnvn occurs in an unrestorable context.28 By analogy with the Brea decree, Luria 23

24 Arist., Ath. Pol. 4.2: see n. 10 above. See pp. 225–6 below. And see Deinarch. I (c. Dem.) 71. 26 M/L 49 ¼ Fornara 100. 39–42 (¼IG I3 46. 43–6). 27 This view is endorsed by Jones, AD 168. I now find it was expressed by Jac., ii 379 n. 25. 28 IG II2 30 [¼Ag. XIX L3] a. 12. The date is from b. 5: [- v]noB & mZn oB to met a ’
Qe
o[doton arxonta-]. Kirchner and others have taken 386/5 to be the date of the 25

12

Athenian Democratic Origins

has plausibly suggested that we should read [-pl Zn i‘ pp
evn ka]i pentakosiomed
imnvn [‘[apart from the Hippeis an]d Pentakosiomedimnoi’].29 The decree evidently concerns Lemnos, conceded to Athens by the King’s Peace, and there are several references to cleruchs.30 This is the last piece of evidence we have for the actual use of the census classes in practice otherwise than as qualifications for office or (see the examples in para. 3 below) under some old administrative law. 3. (a) Demosthenes XXIV 144 (a speech delivered c. 353/2) quotes a law forbidding (with certain exceptions) the arrest of an ‘ an ’ eggyZt ’ ’ o t
eloB telo &yntaB Athenian oB o ayt aB tre&iB kauistfi Z& t [‘who presents three sureties of the same telos (as himself)’]. (That this law does refer to sureties of the same census class receives support from Ath. Pol. 4.2, notwithstanding that this ‘constitution of Draco’ is fictitious.) It is interesting to note that the new law of Timocrates, against which Demosthenes is speaking, clearly had no corresponding provision. We might not feel positive about this if we had no more to go by than the passages Demosthenes quotes from the new law,31 but the expectation expressed in ch. 85 that ’ men of straw (wa
yloyB anur
vpoyB) would invariably be offered as sureties under the new law places the matter beyond doubt. (b) Ps.-Dem. XLIII 54 (a speech delivered apparently in the late 340s) quotes a law prescribing the dowry which had to be provided for an epikleros of thetic status by a next of kin who did not wish to marry her: if he were a Pentakosiomedimnos he had to find 500 drachmae, if a Hippeus 300 drachmae, if a Zeugites 150 drachmae. I think it is very probably a reference to this law which was the occasion for the use of the word pentakosiom
edimnon by Lysias32 in a speech apparently delivered in an inheritance case involving an epikleros.33 decree, but surely the formula quoted implies that Theodotus’ year (387/6) is not yet over and Mystichides (386/5) probably not yet elected. [R. S. Stroud dates the decree to 387/6 in his re-edition, Hesp. XL (1971) 169.] 29 See SEG III.73. [Luria’s supplement has been widely if not universally accepted: see J. Cargill, Athenian Settlements of the Fourth Century B.C. (1995), 192 n. 27.] 30 Frr. a. 13, 20, 22; b. 6; c. 5 [Continuous numeration in later editions: b. 33, c. 42; add e. 47]. 31 §§ 41, 72, 79, 82–87, 89, 93, 100 and 122, even if we reject 39–40 as spurious. 32 Fr. 207 Sauppe, ap. Harp., s.v. pentakosiom
edimnon. 33 & ’Onomakl
eoyB uygatr
The title is Peri tZB oB [‘On the daughter of Onomakles’].

The Late Fifth- and Fourth-Century Evidence

13

Poseidippus, a poet of the New Comedy who wrote in the third century (after the fall of the Athenian democracy), is quoted by Harpocration as saying that it was obligatory upon to&iB e’
ggista ’ lamb
’ e0 mn &aB did
onai g
enoyB t aB u
ZssaB Z anein pr oB g
amon Z [‘the next of kin of girls of the thetic class either to marry them or to give them five minae’];34 and Diodorus35 credits Charondas with a ’ & law peri tvn & epikl Zrvn, o‘ kai par a S
olvni ke
imenoB [‘concerning heiresses, which is also found in Solon’s laws’]: that the next of ‘ penixr ’ iklZroB [‘the poor heiress’] or kin must either marry Z a ep
furnish 500 drachmae towards her dowry. 4. The following are all the sources I know of which mention any of the four census classes in a military context: & (a) Harpocration (cf. Hesych., Phot., Suid.), s.v. uZteB kai uZtik
on cites two late fifth-century writers: (i) Aristophanes, fr. 248,36 from the Daitaleis of 427. Harpo’ estrate
’ & cration simply says, o
‘ ti de o yk yonto (sc. oi‘ uZteB) ’ ’ ’ e
irZke kai Aristow
anZB en Daitale &ysin [‘Aristophanes in the ‘‘Banqueters’’ says that they (i.e. the Thetes) did not serve in the army’]. Unfortunately Aristophanes’ actual words are not preserved. ’ tfiv& Kat & en a Fil
inoy (ii) Antiphon, fr. 61 (Thalheim):37 ’Antiwvn ‘ itaB poiZsai’ & & wZs
i
‘to
yB te uZtaB a
‘ pantaB opl
[‘Antiphon in the ‘‘Against Philinus’’ says ‘‘ . . . and to make all the Thetes hoplites’’ ’]. By itself, the quotation from Antiphon need not mean more than that not all the Thetes served as hoplites in the late fifth century; but unless Harpocration made a wrong inference from Aristophanes the only conclusion we can honestly draw from the two statements taken together is that the Thetes as a class were not liable to hoplite service at all—that is to say, that they formed a category entirely below the hoplite class. (b) Thucydides III 16.1: all Athenians went on board the ships in the emergency in 428, except the Hippeis and Pentakosiomedimnoi.

34 35 37

& Poseidipp. Com. Fr. 38 K/A, ap. Harp. s.v. QZteB kai uZtik
on. 36 XII 18.3. 248 K/A. ¼ fr. 63 Sauppe ¼ fr. B 6 Maidment.

14

Athenian Democratic Origins

(c) Thuc. VI 43 (lines 15–18 OCT): out of the total number of ‘ itai [‘hoplites’] in the great Athenian expedition to Sicily 5,100 opl& ’ in 415, there were ‘of the Athenians themselves, 1,500 ek ’ katal
ogoy [‘‘from the katalogos’’], and 600 Thetes epib
atai tvn & nevn & [‘‘as marines’’]’,38 the rest being allies and mercenaries. Compare Thuc. VIII 24.2, where Leon and Diomedon in 412 ’ ‘ ’ katal
’ have epib
ataB tvn & oplit vn & ek ogoy anagkasto
yB [‘hoplites from the katalogos serving as marines under compulsion’]; ‘ itai ek ’ katal
also VII 16.1; 20.2, where the opl& ogoy [‘hoplites from the katalogos’] appear again. I shall return to the interpretation of these passages presently. After this there seems to be only one passage in which any of the tele¯ might be appearing in a military context. This is Xen. Hell. ’ tfi Z& Zlik
‘ ifia o
’ ntaB I vi 24: when the Athenians put on board toyB en ’ eroyB [‘all those of military age, a
‘ pantaB kai do
yloyB kai eleyu
both slaves and free men’], for the Arginusae campaign in 406, ei’ s
ebZsan de kai tvn & i‘ pp
evn pollo
i [‘and many Hippeis also came on board’]. And here I see no reason at all to suppose that the i‘ ppe&iB concerned are anything but the state cavalry. This text is therefore irrelevant to our consideration of the tele¯.

(II) MILITARY CATEGORIES IN THE FIFTH AND FOURTH CENTURIES I shall speak of three military categories: 1. ‘Cavalry’, meaning as a rule those actually serving in the state cavalry. They are of course i‘ ppe&iB, but to avoid confusion with the telos of Hippeis I shall invariably call them ‘cavalry’. 2. ‘Hoplites’, meaning those on the state register (kat
alogoB) of heavy-armed infantry. I shall sometimes refer to ‘the hoplite class’, an expression I shall generally use to include also those who were once hoplites but had passed the age limit of 6039 and ‘ er t become yp on kat
alogon [‘above/beyond the katalogos’]40 38

These will be the 60 taxe&iai. For 60 ships we would have expected 600 ‘ marines, and although most of the MSS have eptak
osioi the true reading is ‘ probably ejak
osioi, as in one good MS, Hsl—which alone, as K. J. Dover has kindly pointed out to me, has also preserved in VI 96.3 the correct (cf. 97.3; VII ‘ ‘ 43.4) reading ejakos
ioyB, the other MSS all being corrupted to eptakos
ioyB. 39 See Arist., Ath. Pol. 53.4; Poll. II 11; Plut., Phoc. 24. 40 See Dem. XIII 4; Poll. II 11.

Military Categories in the Fifth and Fourth Centuries

15

’ ynatoi [‘incapacitated’] who had sufficient propand those ad
erty to qualify for hoplite service.41 3. ‘Those below the hoplite class’, an expression which speaks for itself. I shall now review each of these categories in turn. 1. It is generally agreed that there was no organised cavalry force at Athens until the mid-fifth century. Our earliest conclusive evidence of the existence of an Athenian military force known as & (with oi‘ i‘ ppe&iB is a dedication on the Acropolis by hoi hippZB 42 their three Hipparchs named), dated by its letter-forms to the 450s or perhaps the early 440s: IG I3 511 ¼ DAA 135, 135a, 135b.43 It is worth mentioning that Sparta apparently had no organised cavalry until 424.44 Nevertheless, there is not the least reason to doubt that there were always—before as well as after Solon45—individual Athenians who possessed warhorses and used them on campaign, as ‘mounted infantry’: at least they could arrive at the battlefield fresher than the ordinary hoplite,46 and once the enemy was in flight they could mount their horses again and lead the pursuit— cavalry were invaluable for this purpose, as the Greeks well knew.47 The horse was the great ‘status symbol’ of the Greek 41 I do not suppose they will have been a large class: here I agree with Jones, AD 162–3. 42 Cf. Andoc. III 5 ¼ Aesch. II 173: 300 i‘ ppe&iB at first. Contrast Schol. Ar., Eq. 627b: ‘the Athenian cavalry were at first 600 in number.’ But the scholion is a very unintelligent one. [Ste. Croix is translating pr vton & ‘at first’, but the latest edd. of the scholia rightly read pr
oteron ‘formerly’, with better MSS support. PJT] 43 Cf. IG I3 512 ¼ DAA 141, which must refer to a hipparch or hipparchs, and comes from the early part of the second half of the fifth century. It will be sufficient to refer to the notes in DAA, ad locc., also Gomme, HCT I 316, 328; II 77, 101–2; Busolt-Swoboda, GS I 344; II 824 & n. 1, 978, 1128. Paus. I 29.6 and IG I3 1181 may refer to 431; but see Wade-Gery in JHS 53 (1933), at pp. 78–9 [cf. D. L. Page, Further Greek Epigrams (1981), 274 n. 1]. 44 Thuc. IV 55.2; cf. Xen., Hipparch. IX 4. The 300 so-called i‘ ppe&iB at Sparta were merely a corps d’e´lite and fought as hoplites: see Thuc. V 72.4; Strabo X 4.18, pp. 481–2; Hesych., s.v. i‘ ppagr
etaB. They are known from at least as early as 480 (Hdts VIII 124), and are doubtless much older. 45 See Endnote i ii (p. 64 below). 46 Cf. the advice of Aen. Tact. XII 13–15 to a hoplite army: to arrive fresh for an engagement, go by sea or on wagons. 47 See e.g. Xen., Anab. II iv 6; III i 2. Note also the anecdote told by Hdts (I 63.2) of how Peisistratus caused his sons to mount after the battle of Pallene and sent them after the fleeing enemy, to do some useful propaganda. And see Endnote iv (p. 65 below).

16

Athenian Democratic Origins

world, above all perhaps in the archaic period, because in ancient Greece, as everyone knew, only the really rich man could afford to i‘ ppotrowe&in [‘keep/breed horses’]. We hear of several Greek states in this early period in which aristocracies of i‘ ppe&iB formed a ruling oligarchy, the best known being the Hippobotai of Chalcis.48 Aristotle was very much impressed by the examples he knew, of which he mentions Chalcis, Eretria, Magnesia on the Maeander ’
kai tvn & allvn polloi peri t Zn ’As
ian [‘and many others in Asia’], and he proclaims it as a general principle that the early ’ tvn Greek monarchies were succeeded by a polite
ia . . . ek & i‘ pp
evn [‘constitution . . . based on the cavalry’].49 The military character of this horsemanship cannot be doubted. For the official state roll of Athenian cavalry in at any rate the later fourth century the technical term seems to have been p
inaj.50 Aristotle, Ath. Pol. 49.2 gives a clear and precise description of how the cavalry were enrolled in his day, and such information as we have for earlier times51 suggests that substantially the same procedure was followed in at any rate the late fifth and early fourth centuries.52 A higher standard of physical fitness was needed for the cavalry than for the infantry, and cavalry service will probably have been given up by most men well before the age of 60, at which the hoplite went on the retired list. The retired cavalryman, unless altogether incapacitated, will surely have been added to the hoplite katalogos, remaining on it until he reached the age of 60. 2. The hoplites were those enrolled on the official register, the katalogos, and are consequently referred to in technical language as 48

Hdts V 77.2; VI 100.1; Arist., Pol. IV 1289b39, and ap. Str. X 1.8, p. 447; Plut., Per. 23.4. And see Busolt, GS I 210–11. 49 Pol. IV 1297b16–22; cf. 1289b35–40; VI 1321a8–11. 50 Arist., Ath. Pol. 49.2, lines 7 & 14 OCT. For the roll of cavalry under the Thirty the term san
ideB (Lys. XXVI 10; cf. san
idion in Lys. XVI 6) seems to have been used. On one or two occasions kat
alogoB is used of a particular list of cavalrymen, e.g. those selected for an expedition (as in Lys. XVI 13), or of new recruits (Arist., Ath. Pol. 49.2. Those on this kat
alogoB who do not secure exemption are entered ei’ B t on p
inaka). 51 ‘ ym ‘ vn e.g. Xen., Oec. IX 15; Lys. XIV 8, 10 (where yw’ & is to be understood as ’ astvn alone is relevant, explained by Sandys, ACA2 192 n.); XVI 13 (where adokim
the enrolment by Orthoboulos being obviously for the particular expedition, not for cavalry service in general). 52 On the question what qualified a man for cavalry service, see pp. 25–6, 48–9 below.

Military Categories in the Fifth and Fourth Centuries

17

‘ itai) ek ’ katal
(opl& ogoy [‘(hoplites) from the katalogos’].53 The word katalogos can also be used (a) of a list of hoplites enrolled for a particular campaign;54 (b) of a corresponding list of cavalry;55 (c) of the lists first compiled (as I think) in 362, with the aim of making those below the hoplite class liable to compulsory naval alogoB propservice;56 and (d) in various other ways;57 but o‘ kat
’ erly means the state register of hoplites, and the expression ek kat
alogoy normally refers to that register. Although we are well informed about the method of enrolling cavalry,58 we have no positive information about the way in which a man was registered in the hoplite katalogos,59 or about the qualifications which entitled him or obliged him to be so registered. This last question will be discussed later. I should like to make a general point which is commonly missed by those who speak of the necessity of the hoplite’s being able to provide himself with the standard armour. A hoplite in fact needed much more than the initial outlay required to buy his armour (which would no doubt last a long time in ordinary circumstances, and might be handed on from father to son): he needed sufficient property to be able to support himself and his household if, as might happen, he had to be away from home for long periods on service—probably unpaid until about the middle fifth century. And the ‘household’ would normally have to include a slave attendant for himself while on campaign and surely another slave or two at least to do the farm work.60 I should be very surprised indeed if any Athenian hoplite owned no slaves at all.61 53 Thuc. VI 43; VII 16.1; 20.2; VIII 24.2; Xen., Mem. III iv 1; Arist., Pol. V 1303a9–10; Ath. Pol. 26.1; Poll. II 11. Cf. Lys. XV 5. We also find the expression ‘ itai ek ’ t vn opl& & t
ajevn [‘hoplites from the ranks’] in Thuc. III 87.3. It has been suggested by Beloch and Busolt (see GS I 572 n. 1) that this expression differs from ‘ itai ek ’ katal
opl& ogoy only in excluding the ephebes. 54 As in Thuc. VI 26.2; 31.3; Ar., Eq. 1369–71; cf. Ach. 1065; Pax 1179–84 etc. I myself believe that Lys. XIV 6–7 and probably XV 7, perhaps 11 (but not 5), also fall into this category. And see Endnote v (p. 65 below). 55 Lys. XVI 13. 56 Ps.-Dem. L 6 ff.: see para. 3 below. 57 See e.g. Arist., Ath. Pol. 49.2 and n. 50 above. 58 See p. 16 above. 59 According to Lys. XV 5, it was the taxiarch who would strike his name off, and presumably therefore it was the taxiarch who enrolled him. 60 See my review of W. L. Westermann, The Slave Systems of Greek and Roman Antiquity, in CR 71 ¼ n.s. 7 (1957) 54 ff., at p. 58. 61 There is only one single text, namely Dem. XXIV 197, which Jones (AD 84) is able to cite in favour of his view that eisphora-payers, and therefore a fortiori

18

Athenian Democratic Origins

3. Those below the hoplite class are often referred to nowadays as Thetes, sometimes as cilo
i [‘light troops’] or as sailors; but we have not yet decided how far those below the hoplite class corresponded with the Thetes, and in my belief it is misleading to call them cilo
i or sailors, because they were not liable to regular conscription in either of these forms until the year 362 (or a little earlier).62 Thucydides expressly says that the Athenians never had any regular formations of cilo
i.63 Nor, in the fifth and early fourth centuries, was conscription used for the fleet, as far as I can see, except in emergencies, when there might be something like a leve´e en masse.64 I think regular conscription for the fleet was first introduced by the decree of Aristophon in the autumn of 362 (Ps.-Dem. L 6–7, 16).65 All passages referring to naval conscription, as far as I know, are without exception subsequent to this: i.e. Isocr. VIII 48 (c. 355); Dem. III 4 (referring to late 352); IV 36 (delivered in 351 or just after);66 Aesch. II 133 (referring to 346). Compare Dem. XXI 154–5, with Thuc. VI 31.3 and Lys. XXI 6. I realise that the view I have expressed is not the accepted one, but it seems to me very probable.67 We never hear of katalogoi (or a katalogos) for those below the hoplite class before 362, and surely what is described in Ps.-Dem. L 6–7 is not simply the extraction of lists of crews68 from registers already existing, but the actual compilation of such registers in the first place, in all the individual hoplites, might own no slaves, and this will not bear the interpretation he gives it— see Endnote vi (p. 66 below). 62

Not until c. 337, according to Kahrstedt, SSA 249 n. 1; UMA 21. ’ paraskeyZB ‘ & . . .vplism
Thuc. IV 94.1: ciloi . . . ek enoi oy
’ te t
ote [424 B.C.] ’ enonto tfi Z& p
& parZsan oy
’ te eg
olei [‘they had no regular light-armed troops with them on this occasion, nor did Athens ever have an organised force of this kind’]. 64 As perhaps in Thuc. III 16.1 (428); Xen., Hell. I vi 24 (406); ?V iv 61 (376). The mere use of the expression pandZme
i [‘in full force’] (Thuc. II 31.1; III 91.4; IV 90.1; Xen., Hell. II iv 43; IV iv 18) or panstratia&fi [‘with the whole army’] (Thuc. II 31.1; IV 66.1; 94.1) need not imply that there was a universal levy, extending below the hoplites (even if numbers of light troops—and even manual workers—are sometimes involved): cf. Xen., Hell. I iii 10; vi 18; II ii 7; IV vi 3; V iv 42. 65 ’ Note esp. the words katal
ogoyB poie&isuai t vn & dZmot vn & kai apow
erein na
ytaB [‘make katalogoi of the demesmen and return (a list of) sailors’] in § 6. 66 See G. L. Cawkwell in CQ 56 ¼ n.s. 12 (1962) 122–7, who makes a good case for 351. 67 ’ The existence of an offence of anaymax
ioy (for which see Lipsius, ARR 454 & n. 7, with refs.) need not imply more than what is understood by LSJ9, s.v.: ‘indictment of a taxiarch for keeping his ship out of action’. 68 Corresponding to the kat
alogoi mentioned in n. 54 above. 63

Tele¯ and Military Categories

19

demes, under the supervision of the demarch and councillors of each deme. Even if I am wrong about this, the naval conscription of 362 will have been the earliest actually recorded, and we need not suppose that the procedure had been instituted long before that date.69 It is precisely because there was not even a register of Thetes corresponding to the hoplite katalogos that Thucydides refrains from giving actual figures for the numbers serving70 or the losses in battle71 or by plague72 of those below the hoplite class, and contents himself with uncharacteristically vague expressions like ’
’ ol
’ igoB [‘a substantial crowd of light o‘ alloB o
‘ miloB cilvn & oyk ’
’ ’ troops in addition’], or to &y de alloy o
’ xloy aneje
yretoB arium
oB [‘a great and undetermined number from the rest of the popu’ lation’], or cilvn & de kai skeyow
orvn polyB arium
oB [‘a great number of light troops and baggage-carriers’].73

( I I I ) T E L E¯ A N D M I L I T A R Y C A T E G O R I E S IN THE FIFTH AND FOURTH CENTURIES What relationship, if any, existed between military categories and tele¯ in the late fifth and fourth centuries? The Pentakosiomedimnoi, for any other purpose than eligibility for office, can surely be regarded merely as the top section of the Hippeis, and for present purposes when I speak of Hippeis I must usually be understood to include the Pentakosiomedimnoi. What we have to decide is how far Hippeis corresponded with cavalry, Zeugitai with hoplites, and Thetes with those below the hoplite class. That there must have been at least a rough equivalence at all three levels74 is clear from 69

I suppose the year 378/7 is possible. Cf. my DTAEFC, at pp. 59–62, for comments upon the peculiar procedure adopted in 362 for levying a proeisphora, through the demes, in spite of the fact that quite different machinery, involving the symmories, had been devised in 378/7, though probably never used in the interval. [G. L. Cawkwell, CQ 34 (1984) 338, and V. Gabrielsen, Financing the Athenian Fleet (1994) 107, envisage conscription as exceptional even after 362.] 70 Thuc. II 31.2 (the invasion of the Megarid, 431); IV 94.1 (Delium, 424). 71 IV 101.2 (Delium). 72 III 87.3. 73 II 31.2; III 87.3; IV 101.2. 74 Even Hignett (HAC 101), in rejecting, with a reference to De Sanctis and Busolt-Swoboda (who, incidentally, give no real arguments), ‘the view that the classes had a military origin’, thinks it ‘possible that after they had been created the property-classes came to be used to decide the incidence of military obligations’. Contrast pp. 48–9 below.

20

Athenian Democratic Origins

the name of the telos of Hippeis and the passages quoted on pp. 13–14 above from Aristophanes, Antiphon and Thucydides. The question is whether the equivalence was complete. If it was not, we shall have to see if we can establish separate sets of qualifications for the tele¯ on the one hand and the cavalry and hoplites on the other. Most scholars have either taken the identification for granted or hedged. I have come across only one presentation of the case against making the identification which is worth taking seriously: that of Kahrstedt,75 who believes that military service was in earlier times tied to the tele¯ but that it became detached when the tele¯ became meaningless (‘gegenstandlos’), as Athens grew out of a natural economy.76 The tele¯ were superseded in practice for military purposes, Kahrstedt thinks, by a classification of citizens as ’ to &y katal
& ‘oi‘ ek ogoy (Landdienstpflichtige) und uZteB im neuen Sinn (Ruderdienstpflichtige)’ [‘those from the katalogos (liable to service on land) and thetes in the new sense (liable to service at sea)’].77 In fact there is no evidence that the term Thetes ever came to be used in any ‘new sense’, nor, as I have already shown, is there any evidence that the Thetes were under any duty, except in emergency, to serve in the fleet, as the hoplites were to serve in the land army. Kahrstedt’s theory is founded upon a conception which I believe to be false and which anyway remains to be investigated: the belief that the census classes were always defined in the way Aristotle says they were when they were first created by Solon, in terms of income from landed property. If indeed they were originally and always so defined, then of course well-to-do Athenians who owned no land, or only a little, would necessarily be Thetes, yet such men would surely have been liable to hoplite service,78 and this would destroy any equation between Zeugitai and hoplites. Let us return to the sources I have quoted on pp. 13–14 above. Thuc. VI 43, using technical language, draws a clear distinction ’ katal
between hoplites proper, ek ogoy [‘from the katalogos’], and 75

SSA 251–5; UMA 20–1, cf. 55, 61. 77 See esp. SSA 251. UMA 21. It would be absurd to pretend that there were no such men, or that they were an insignificant class, for by 431 at least 3,000 hoplites (see Thuc. II 31.2) and possibly many more were drawn from the metics, and therefore, with few or no exceptions, would have been entirely landless. 76 78

Tele¯ and Military Categories

21

’ & uZteB epib
atai [‘Thetes serving as marines’]. This is the only & occasion on which the word uZteB appears in Thucydides. In historical times, as far as our information goes, the word was used at Athens in only two different senses: it could mean the lowest of the four census classes, or it could be applied to misuvto
i, men who worked for wages. No one will suppose that Thucydides is using it in the latter sense here. I cannot imagine why he should want to call these men by the technical term Thetes if they were not precisely that, as it would be thoroughly misleading. The whole point of his statement is to explain the status of the two different groups which together made up those armed as hoplites on the expedition in question. He defines one group, the ’ katal
& hoplites, by a technical term, ek ogoy, and the word uZteB must similarly be technical: if it were not, it would tell us nothing about the status of the men concerned. And these ‘Thetes’, by ’ katal
contrast with the other group, are not ek ogoy, not enrolled on the hoplite katalogos, not members of the hoplite class. This strongly reinforces the tentative conclusion we drew from the passages quoted by Harpocration, that the Thetes were not hoplites but formed a class below them. But if the hoplites were the class above the Thetes, then it is an irresistible inference that the line (whatever it may have been) which divided Thetes from Zeugitai was the same as that dividing Thetes from hoplites, so that Zeugitai and hoplites (or the hoplite class)79 were identical—for at the next level Hippeis and cavalry can hardly be separated. It also appears from Thuc. VI 43 and VIII 24.280 that the ’ regular hoplites did not normally serve as marines (epib
atai) in the Athenian fleet, at least in the late fifth century and probably at all times, although they might on occasion be obliged to do so, but that marines were customarily drawn either from Athenian Thetes or (like many of the rowers)81 from other volunteers—no doubt mainly from the allies. Such men may well have been given their armour by the state. It is not possible to speak of a ‘rough equivalence’ between Zeugitai and hoplites, and between Thetes and those below the hoplite class: any such attempt to obscure the issue is excluded by 79

80 See p. 14 above, and cf. pp. 23–4 below. See p. 14 above. See the recent article by M. Amit, ‘The Sailors of the Athenian Fleet’, in Athenaeum 40 (1962) 157–78. 81

22

Athenian Democratic Origins

Thucydides’ quite explicit language. Again, it has been suggested to me that ‘calling these men Thetes is merely a reference to their financial status, men whom the state had to arm as hoplites’. The obvious reply is that being either a Zeugites or a Thes, or being either a hoplite or below the hoplite class, was indeed a matter of ‘financial status’, and anyone who could not afford to serve as a hoplite, and had to be armed by the state, was not a hoplite and is properly called a Thes. The significance of Thuc. III 16.1 is that it presents us with two census classes, Hippeis and Pentakosiomedimnoi, introduced in a military context: they receive exemption from service in an emergency in which everyone else is mobilised. Without the addition of kai pentakosiomed
imnvn [‘and the Pentakosiomedimnoi’], the words pl Zn i‘ pp
evn in this passage would most naturally be taken as referring to the actual serving cavalry, some such expression as pl Zn tvn & i‘ pp
ada telo
yntvn [‘except those assessed as Hippeis’] being necessary to signify clearly the census class of Hippeis, whereas if i‘ ppe&iB are joined with Pentakosiomedimnoi they can hardly be anything but the census class. (Realising this, Kahrstedt82 was driven to the desperate expedient of deleting kai pentakosiomed
imnvn as an interpolation.) At first sight one might think that it would have been sufficient to give exemption to the i‘ ppe&iB (the cavalry) alone. Would not the exclusion of the cavalry be understood to cover both Hippeis and Pentakosiomedimnoi? Was it really necessary, therefore, to mention the Pentakosiomedimnoi at all? In fact the decree exactly answered to the needs of the situation. To have said merely pl Zn i‘ pp
evn would have been ambiguous: prima facie the expression might well be thought to exclude only the actual serving cavalry, and this would in theory have made liable to service on the ships a fair number of wealthy Athenians who were members of the census class of Hippeis (or Pentakosiomedimnoi) but were no longer physically fit for cavalry service and had therefore been registered in the hoplite katalogos.83 Such men were not likely to be of much use as rowers. To say ‘except the Hippeis and Pentakosiomedimnoi’ would make it quite clear that they too were excluded. Thucydides’ wording, technical and precise, no doubt reproduces that of the actual decree. 82

SSA 253 n. 5. The extreme rarity of references to the Pentakosiomedimnoi makes the interpolation of their name very unlikely indeed. 83 See p. 16 above.

Tele¯ and Military Categories

23

A factor which seems to have influenced Kahrstedt in his attempt to separate the tele¯ from the military categories is that the tele¯ are mentioned in relation to military service by no ancient source but Thucydides. But why should we expect anyone else to mention them in this connection? Let us see precisely what Thetes, Zeugitai and Hippeis were. I suggest that the key to the whole problem is the realisation that we must begin with, and treat as primary, not the census classes but the military categories. Scholars who have identified tele¯ and military classes have begun with the tele¯ and have assumed that a man was drafted into the cavalry or the hoplites or exempted from military service according to whether he was qualified by his property assessment as a Hippeus, a Zeugites or a Thes. That would naturally make membership of a particular telos a very important factor in every man’s position. It then becomes very puzzling that we have no information whatever about the nature of the qualifications for the respective tele¯—or even a hint that such qualifications existed in the fifth and fourth centuries. Reserving for later treatment the position under the original Solonian constitution, I am going to argue that in the fifth and fourth centuries membership of all tele¯ except the highest was entirely dependent upon a man’s military classification, and did not determine that classification. We can postpone for the present the question how a man came to be placed in one military category rather than another—whether, that is to say, the hoplite was (as I believe) a man who could afford to serve in the infantry and a cavalryman one who could afford to serve on horseback, or whether there were precise quantitative qualifications, in terms of landed produce or of money (capital or income), and if the latter, what the figures were. 1. The Thetes, on my hypothesis, will have been those who were too poor (in the sense hereafter defined) to become hoplites. ’ katal
2. Of the Zeugitai the nucleus will have been oi‘ ek ogoy. From these serving hoplites we must deduct (a) those men, whether Hippeis or Pentakosiomedimnoi, who had when younger served in the cavalry but had retired from it as being insufficiently able-bodied and had been enrolled on the hoplite katalogos, and (b) any men with sufficient wealth to maintain warhorses whose services were not at the time required in the cavalry because the ‘establishment’ (of 1,000 or whatever) could

24

Athenian Democratic Origins

be made up without them, and who were consequently serving ‘ er t as hoplites. And we must add (a) oi‘ yp on kat
alogon [‘those above/beyond the katalogos’] and (b) those incapacitated ’ ynatoi)—the lame and the blind and so forth—who had (ad
sufficient property to make them financially capable of hoplite service.84 3. The Hippeis—more strictly, oi‘ i‘ pp
ada telo &ynteB [‘those assessed as Hippeis’]—will have consisted of the cavalry actually serving, minus the Pentakosiomedimnoi, and plus (a) former cavalrymen, whether now serving as hoplites or on the retired list, (b) any of those men already mentioned who were rich enough to be cavalrymen but were serving as hoplites because they were ‘surplus to establishment’, and (c) the incap’ ynatoi) of sufficient wealth to have served in the acitated (ad
cavalry, who were not equal to the physical exertion involved and who either fought with the hoplites or were excused military service altogether. 4. The Pentakosiomedimnoi were the top layer of the Hippeis. The nature of the qualification for this class cannot be determined until we have considered the position under Solon. The census classes, so defined, would remain important only so long as they were actually used for the purpose for which they were designed: qualification for office, provision of a surety or of a dowry for a thetic epikleros or the like, and perhaps classification for the levying of taxes.85 Apart from this, I cannot see that there will have been any occasion for anyone to refer to ‘Zeugitai’ or to ‘Hippeis’ (in the sense of oi‘ i‘ pp
ada telo &ynteB [‘those assessed as Hippeis’]) or to ‘Pentakosiomedimnoi’. In military contexts the terms used would be ‘hoplites’ and ‘cavalry’, and for the ‘hoplite class’ and ‘cavalry class’ together there existed another expression, oi‘ o
‘ pla parex
omenoi [‘those providing (their own) arms’]. Characteristically, it is in Thucydides, less averse than most historians from using technical terms, that we find our one reference to those below the hoplite class as ‘Thetes’. *** ’ katal. include metic [Ste. Croix here asked himself in a note ‘did oi‘ ek hoplites?’.] 85 See pp. 7–8 above and section (viii) below. 84

Tele¯ and Military Categories

25

Let us now try to see what can be said about the qualifications for membership of the military classes and the tele¯ in the fifth and fourth centuries. No direct information at all survives, and there is of course no earlier information of any kind except for the tele¯ in the time of Solon himself. At this stage I am going to make some tentative suggestions to which I shall return at a later stage. For the present we can mainly ignore Pentakosiomedimnoi and Thetes and concentrate upon cavalry/Hippeis and hoplites/Zeugitai. I begin with the cavalry and the telos of Hippeis. In the first place, we never have a hint in any source that the word i‘ ppe&iB can mean two quite different things: the telos of Hippeis, and the actual serving cavalry. Xenophon, our principal authority, can occasionally speak of t o i‘ ppik
on86 but he normally calls the cavalry 87 & of Aristophanes’ play of that name are i‘ ppe&iB. The ‘IppZB certainly the state cavalry.88 The telos of Hippeis evidently played little or no part in anyone’s consciousness. That becoming a cavalryman, at any rate in the fourth century, depended in any sense upon membership of a telos of i‘ ppe&iB or i‘ pp
ada telo &ynteB, or for that matter upon the possession of any time¯ma expressed in quantitative terms, seems to me entirely excluded by the negative evidence, above all that of Xenophon, Hipparch. I 9–12; IX 5, and Aristotle, Ath. Pol. 49.2. It is worth looking at this evidence closely. (a) Xenophon (Hipparch. I 9) simply advises his Hipparch toyB men to
inyn i‘ pp
eaB . . . kauist
anai . . . kat a t on n
omon toyB ’ ei’ s
agonta ei’ B dynatvt
atoyB kai xr
Zmasi kai s
vmasin Z ’ pe
iuonta [‘to raise . . . his cavalrymen . . . according dikast
Zrion Z to the law, those best qualified by wealth and strength, either by bringing them before the court or by persuasion’]. Had there been any specific qualification, in terms of money or of Solonian ‘measures’, it would surely have been mentioned here. (b) According to Aristotle (Ath. Pol. 49.2), anyone newly placed on the list (kat
alogoB) of cavalrymen compiled each year by the Katalogeis was excused if he was prepared to take an oath ’ omnysuai) m ’ tfi Z& oys
’ ifia,89 (ej
Z d
ynasuai tfiv& s
vmati i‘ ppe
yein Z 86

e.g. in de Vect. II 5; Hipparch. I 2, 3 etc. As in Hipparch. I 2, 5, 6 etc. Note lines 242 ff., where Simon and Panaitios are the hipparchs; cf. Schol. ad 242. 89 Cf. Xen., Hipparch. IX 5. 87 88

26

Athenian Democratic Origins

i.e. that he could not serve either because he was not physically fit to do so or because he had insufficient property. Had there been a fixed quantitative qualification, the class of each man would have ’ been settled automatically, and the ejvmos
ia [‘oath-taking’] would either have referred specifically to the qualifying time¯ma or would have been confined to physical incapacity. The cavalryman became such, therefore, simply by being financially able to serve in the cavalry. He volunteered for it, or was persuaded or compelled to serve,90 or obtained exemption as an ’ ynatoB [‘incapacitated’]; and this was what also made him a ad
member of the census class of the Hippeis. It may appear less surprising that there was no quantitative time¯ma possession of which involved cavalry service when we remember that there were no fixed qualifications which made men liable for the performance of liturgies either.91 Maintaining a warhorse was a kind of liturgy—and quite an expensive one. I turn now to the hoplites and Zeugitai. I believe that a man had himself registered in the hoplite katalogos if he could afford to provide himself with arms and armour and was financially able to bear the burden of going on campaign when required (see p. 17 above). It is true that some men shrank from serving in the cavalry;92 but I think everyone who could qualify as a hoplite would have wanted to have himself put on the katalogos, because of the social cachet he would thus obtain; and if there was anyone conspicuously able to serve who did not apply for registration, the Taxiarch of his tribe would no doubt enter his name, and social pressure to accept the fait accompli would be irresistible. I believe, then, that there was no fixed quantitative time¯ma possession of which made a man a hoplite. In favour of this view I would advance the following considerations:93 90

See esp. Xen., Hipparch. I 9–12. This seems quite certain and is now, I think, generally accepted, in spite of the opinion of Busolt-Swoboda, GS II2 271 n. 1. [J. K. Davies (personal communication) endorses Ste. Croix’s view.] 92 See e.g. Xen., Hipparch. I 9–12; IX 5. Contrast Lys. XVI 13; but the statement that service in the infantry was more dangerous on the occasion in question than cavalry service may be thought to receive some confirmation from Lys. XIV, esp. 7, 14. 93 I refuse to draw any conclusions from the conflicting statements made about the property of one distinguished hoplite, Socrates: contrast Xen., Oec. II 3, with Dem. Phal., FGrH II B 228 F 43.9 (ap. Plut., Arist. 1.9), and other sources discussed by Bo¨ckh, SA I3 142–4. 91

Tele¯ and Military Categories

27

(1) In the revolutionary movement of 412–11 we never hear either of the tele¯ or of a monetary or other qualification for the o
‘ pla parex
omenoi [‘those providing (their own) arms’]. This is perhaps understandable in the early stages of the movement, because the leading figures wanted a politeuma of less than 5,000,94 which was a good deal smaller than the hoplite class; but one might have expected a quantitative time¯ma to be fixed in the autumn of 411 for ‘the Five Thousand’ who were defined as the hoplite class.95 (2) Even more striking is the ‘constitution of Draco’ in Arist., Ath. Pol. 4, which I take to have originated in the fourth century.96 Here we do find precise property qualifications for archons and treasurers (1,000 drachmae) and for generals and hipparchs (10,000 drachmae), but none for the o
‘ pla parex
omenoi who alone are to enjoy full citizen rights and fill the minor offices. Why not, since this is by far the most important category—unless it was deliberately intended (in accordance with current Athenian practice, as I would say) not to have a fixed financial qualification for the o
‘ pla parex
omenoi at all? (3) Certainly in some states there were quantitative political time¯mata: see e.g. Arist., Pol. IV 1297b 1–6; V 1306b9–16; Hell. Oxy. XI 2 (the cities of the Boeotian League); Plato, Rep. VIII 550C, 553A; cf. Laws V 744B–5A. But I do not remember coming across a clear example of the fixing of a monetary figure as a hoplite (or cavalry) qualification, although of course this may have happened in some states.97 (4) If my theory is right, the situation would be much more flexible on the death of one of the poorer hoplites who left more than one son: obviously a fixed monetary time¯ma creates difficulties in such a situation. Even before the end of the fifth century the census classes as such had evidently become unimportant. Their principal function had always been to determine eligibility for political office, and in the half-century between 508/7 and 457/6 most offices were 94

Thuc. VIII 65.3. Thuc. VIII 97.1. They were probably now about twice that number, if we can place any reliance upon Ps.-Lys. XX 13. 96 See n. 10 above. 97 [In a marginal note, Ste. Croix expresses his aspiration to add ‘something on Rome’ here.] 95

28

Athenian Democratic Origins

probably opened to all classes which could possibly aspire to them—even the archonship to the Zeugitai in 457/6.98 (I cannot believe that any Thes would have thought of presenting himself as a candidate for an archonship for at least a generation after that date.) By the mid-fifth century, therefore, it is likely that candidates for office would always be of at least the required status, and hence that membership of a telos would be something that rarely entered anyone’s mind. Occasionally, as when allotments of land were being distributed in a colony or cleruchy, it might be convenient to make use of one or more of the census classes, because the four classes did, after all, include all citizens, and for none of them was there a precisely equivalent expression.

(IV) THE ORIGINAL QUALIFICATIONS O F T H E T E L E¯ It is unnecessary to set out at length all the numerous sources which purport to give information about the original qualifications of the four tele¯, since the later ones mainly copy from the earlier. Full details may be found in the standard works.99 I shall discuss only the more important sources here. 1. The earliest explicit account we have is that in Aristotle, Ath. Pol. 7.4 (cf. Pol. II 1274a15–21). With the exception of Poll. VIII 130, which will be mentioned below and separately examined later, all subsequent versions are directly or indirectly derived either from this passage or from some independent representative of the same tradition,100 and the variants they introduce are due to misunderstanding or error. Aristotle’s statements can be summarised as follows. The Pentakosiomedimnos was a man who derived from ’ tZB ’ iaB in the sense of gZB) & oi’ ke
iaB, sc. gZB, & or oys
& his own land (ek 500 measures, dry and liquid together (m
etra t a syn
amwv jZr a ‘ a)—annually, of course. The Hippeus was a man who had kai ygr
300 such measures. An alternative theory, held by ‘some people’, 98

Other offices as a qualification for which Solon may have prescribed membership of one of the two highest tele¯ will surely have been opened to at least the Zeugitai not later than the archonship—except perhaps financial ones. 99 e.g. Bo¨ckh, SA I3 578 ff., esp. 580a; Kahrstedt, SSA 250 ff.; Busolt-Swoboda, GS II 821–4, 836–8; Sandys, ACA2 26–30 nn. 100 I say ‘the same tradition’ to allow for the possibility that Aristotle’s statement is taken from an early Atthidographer. Cf. pp. 286–307 below.

The Original Qualifications of the Tele¯

29

that it was ability to i‘ ppotrowe&in [‘keep/rear a horse’] which made a man a Hippeus, is mentioned only to be rejected in favour of qualification in terms of ‘measures’. (I reserve for later discussion101 Aristotle’s reference to the statue and epigram of Anthe’
mion, advanced by the enioi [‘certain people’] as evidence for the theory Aristotle rejects.) A Zeugites was a man who had 200 measures; and ‘the others’, i.e. those who had fewer than 200 measures, or none at all, belonged to the class of Thetes. 2. The next source in order of value is Plutarch, Solon 18.1–2. Here we find the same account,102 except in regard to the Hippeis, ’ who according to Plutarch were toyB ‘
ippon tr
ewein dynam
enoyB Z m
etra poie&in triak
osia [‘those able to keep/rear a horse or produce 300 measures’]. (It is not clear whether Plutarch thought that a man might qualify as a Hippeus in either of these ways, or whether he was uncertain which alternative was right.) 3. Out of many passages in late authors, mainly lexicographers,103 only Pollux VIII 130–1104 (partly repeated in almost exactly the same words by Schol. Plat. Rep. VIII 550C) need be mentioned separately. Pollux gives the same qualifications as Aristotle and Plato for Pentakosiomedimnoi and Zeugitai. Of the Hippeis he says that they seem to have got their name from their ability to ’ ioyn de m
etra triak
maintain horses, epo
osia [‘and they produced 300 measures’]—in other words, he is on the side of Aristotle 101

See the Appendix to this chapter. ’ jZro&iB omo ‘ &y kai ygro& ‘ The m
etra [‘measures’] are en iB [‘in dry and liquid alike’]. 103 & Including Harp., Hesych., Phot., Suid. and Etym. Magn., s.v. QZteB kai uZtik
on (etc.) and i‘ pp
aB, i‘ pp
ada, i‘ ppe&iB; Harp., Phot. and Suid. ’ timZm
s.v. pentakosiom
edimnon; Hesych. and Suid. s.v. ek atvn; Hesych. s.v. zeyg
ision; and see Bekker, Anecd. Gr. I 260.33, 261.29, 264.19, 267.13, 298.20. 104
Poll. begins, tim
Zmata d’ Zn t
ettara, pentakosiomed
imnvn i‘ pp
evn ’ to &y pentak
‘ a poie&in zeygit vn & uZt vn. & oi‘ men ek osia m
etra jZr a kai ygr ’ Zliskon d’ ei’ B t klZu
enteB
an
o dZm
osion t
alanton
oi‘ de t Zn i‘ pp
ada telo &ynteB ’ men to &y d
ynasuai tr
ewein ‘
ippoyB keklZsuai ’ ioyn de m
etra & ek doko &ysin, epo
’ Zliskon de Zmit
‘ ’ o triak
osia, an
alanton. oi‘ de t o zeyg
Zsion telo &ynteB ap ’ Zliskon de mn &aB d
eka
oi‘ de t diakos
ivn m
etrvn katel
egonto, an
o uZtik on
’ ’ Zn Zrxon, ’ e an
’ Zliskon oyd
’ en. [‘There were four time¯mata, oydem
ian arx oyd Pentakosiomedimnoi, Hippeis, Zeugites and Thetes. The Pentakosiomedimnoi were so named because they produced 500 measures, dry and liquid together; and they paid a talent into the public treasury. Those assessed as Hippeis seem to get their name from their ability to rear/keep horses; and they paid half a talent. Those assessed as Zeugites were enrolled from 200 measures (i.e. from those able to produce etc.), and they paid 10 minae. Those assessed as thetes held no office, nor did they pay anything.’] 102

30

Athenian Democratic Origins

against the e’
nioi [‘certain people’] controverted in Ath. Pol. 7.4. There follows a reference to the epigram of Anthemion, which I will discuss with the corresponding passage in the Ath. Pol.105 Pollux also has one set of statements which is entirely independent of the main tradition: he declares that the Pentakosiomedimnoi ‘paid a talent into the treasury’, the Hippeis half a talent, the Zeugitai ten minae and the Thetes nothing. I shall examine these statements later, merely remarking here that they certainly cannot be accepted as they stand, but that they may preserve in a distorted form some valid information about the use of the tele¯ as a rough classification for the levying of taxes. 4. Only one other passage need be mentioned here: Ps.-Dem. XLIII 54, which (as we have already noticed) specifies the value of the dowry to be given to an epikleros belonging to the thetic class by a next of kin who does not marry her himself, the figures being 500 drachmae for a Pentakosiomedimnos, 300 for a Hippeus and 150 for a Zeugites. It is the zeugite figure which is interesting here, for there is an exact correspondence between the ‘measures’ required (according to Aristotle’s account) as a qualification for membership of the two top classes and the amount of the dowry in drachmae, and the question is whether the correspondence holds for the Zeugites also. Bo¨ckh,106 writing before the discovery of the London papyrus of the Ath. Pol., based on this passage his conclusion that the zeugite qualification was 150 and not 200 measures, and the same view has even been upheld by several writers to whom the Ath. Pol. was available,107 usually with the addition that the figure was raised after Solon’s time from 150 to 200 measures (and sometimes even with a date for this alteration!). At this point we need only note that the Demosthenic law does not even purport to give the qualifications of the three highest classes. One major question remains: how far have we reason to trust the account of Solon’s political classification given by Aristotle? I discuss in a later chapter of this book the whole problem of the reliability of the Ath. Pol., and at this point I need only make some brief observations. There are two factors which may make us feel uncertain about the value of the details Aristotle gives. First, we do not know how 105

106 See the Appendix to this chapter. SA I3 581. Including Cavaignac, Beloch, Lehmann-Haupt and Busolt-Swoboda: see GS II 822 n. 1. 107

The Original Qualifications of the Tele¯

31

far the laws contained on the actual Axones of Solon could still be read in the late fourth century, and how far Aristotle (and the Atthidographers) had to rely upon copies—which, of course, may or may not have been reliable, and may or may not have incorporated later alterations and additions to the original Solonian corpus.108 And secondly, the qualifications of the census classes were undoubtedly part of Solon’s constitution (polite
ia), rather than of his laws (n
omoi) in the narrower sense, and therefore ought not to be conceived as necessarily forming part of the nomoi on the Axones.109 The fundamental definition of the tele¯ probably did not appear on the Axones at all. We must scrutinise Aristotle’s account carefully (though certainly not contemptuously) and not refuse in advance to admit that Aristotle may be at least partly wrong. That Aristotle may indeed have been wrong in some important particulars, because he did not have complete and reliable sources of information, is strongly suggested by two arguments. First, the description of the qualifications of the top three classes as so many ‘ a [‘dry and liquid together’] measures t a syn
amwv jZr a kai ygr
makes Solon’s provisions ridiculous, as I shall proceed to show. A perfectly sensible alternative can be found, at any rate for the Pentakosiomedimnoi. We must then suppose that Aristotle either drew mistaken inferences from the Solonian nomoi, or was misled by a tradition already established, perhaps, by the Atthidographers. Secondly, the very way in which Aristotle states the qualification of the Hippeis in Ath. Pol. 7.4 is a clear indication of a lack of certain knowledge. It seems to me virtually impossible that when Aristotle was defining the four tele¯ in Ath. Pol. 7.4 he could even have thought he had before him the actual text of Solon’s provision. The Hippeis, he says, were toyB triak
osia ‘ d’ e’
nio
i wasi toyB i‘ ppotrowe&in dynam
enoyB [m
etra] poio &yntaB, vB [‘those who produced 300 [measures], or as some people say, those able to keep/rear horses’]; and he then proceeds to give the ’
rather feeble argument of the enioi, based on the name of the telos 108

See pp. 317–22 below. See pp. 310–17 below. Another explanation is offered by some scholars: that Solon’s laws contained no provisions regarding the qualifications of the census classes because Solon found them existing and left them unchanged—see, e.g., Busolt, GG II2 47 n. 2; Busolt-Swoboda, GS II 820–1 (esp. 821 n. 1), 836–7. I find this impossible to believe, for the reasons explained on pp. 48–9 and 55 below. However, I should like to point out that my reconstruction of the qualifications of the tele¯ would scarcely be affected if they were in fact pre-Solonian in origin. 109

32

Athenian Democratic Origins

and the statue and epigram of Anthemion.110 He then concludes, ’ ’ & oy’ m Zn all’ e ylog
vteron to&iB m
etroiB difi ZrZsuai kau
aper toyB pentakosiomed
imnoyB [‘however it is more probable that they were defined by measures, just as in the case of the Pentakosiome’ dimnoi’]. There is no doubt what Aristotle means by e ylog
vteron: more reasonable, more probable, more credible.111 Whether we translate (with Mathieu) ‘il est plus logique que’, or (with Kenyon) ‘it seems reasonable to suppose that’, or (with von Fritz and Kapp) ‘it is more probable that’, or in some other way (‘it was more reasonable that they should be classified according to measures’?), the natural supposition is that Aristotle had no hard evidence, any more than the people against whom he was arguing. If he had had the text of the law, he would not have needed to waste several lines on speculations he knew for certain to be wrong, only to end up with a feeble appeal to reasonableness or probability: he could have quoted the actual words—‘But Solon himself says . . . ’—and that would have settled the matter. Aristotle, then, appears not to have known for certain what the original qualification of the Hippeis was; and that makes us doubtful whether he necessarily had reliable information about the qualifications of the other classes.

(V) THE ORIGINAL QUALIFICATIONS OF THE PENTAKOSIOMEDIMNOI According to Aristotle, Ath. Pol. 7.4, e’
dei . . . tele&in pentakosio‘ an ’ ek ’ tZB & oi’ ke
iaB poifi Z& pentak
m
edimnon . . . oB osia m
etra t a ‘ syn
amwv jZr a kai ygr
a [‘a man had to be classed as a Pentakosiomedimnos if he produced from his own land 500 measures, dry and liquid together’]. Two main problems arise here: (1) Is ‘ a a correct descrippentak
osia m
etra t a syn
amwv jZr a kai ygr
tion of the method of assessment? (2) Is it true that the only income ’ tZB & oi’ ke
iaB [‘from his own land’], so which counted was that ek that only the produce of land counted, and not income derived from manufacture or trade or personal skills? We shall first consider the former question (in paragraphs 1 to 7 below) and then turn to the latter (paragraphs 8 to 12). 110

On which see the Appendix to this chapter. Cf. de Gener. I 18, 725a24; EN I 13, 1102b2; de An. I 4, 408a34; Poet. 24, 1460a34; de Caelo II 4, 286b34; 7, 289a22. 111

The Original Qualifications of the Pentakosiomedimnoi

33

‘ a [‘dry and 1. (a) The 500 measures t a syn
amwv jZr a kai ygr
liquid together’] raise problems which very few scholars have honestly faced. Aristotle evidently thought that one medimnos of wheat, one medimnos of barley, one metretes of oil and one metretes of wine each counted the same. But they certainly did not have the same value. Throughout most of Greek and Roman history, from at least the fourth century onwards (and doubtless earlier), the price of wheat was ordinarily somewhere in the neighbourhood of twice that of barley;112 and since wheat was always preferred for human consumption, and the soil of Attica is far better suited to barley,113 the value of wheat in Solon’s Attica is not likely to have been much less than twice that of barley. We know nothing of the prices of oil and wine at this early date, and both of course vary greatly in quality as well as quantity from place to place and year to year. But in most years it would surely have been possible to produce considerably more metretai of wine than medimnoi of wheat or even barley from a given piece of land,114 and the price of a metretes of wine of average quality will scarcely have been less than that of a medimnos of barley and may have been distinctly higher.115 On the other hand, a given area of land would probably have produced fewer metretai of olive oil than medimnoi of corn, barley anyway,116 although a metretes of oil would be likely to fetch appreciably more than a medimnos of barley, or even wheat.117 Many attempts have been made by misguided scholars to calculate the respective areas of land which a man would have to own in order to qualify as a Pentakosiomedimnos, Hippeus or Zeugites.118 Commenting on these, Glotz and 112 See Jarde´, CAG I 164–83; F. Heichelheim in RE Suppl. VI, cols 886–90; Econ. Survey of Anc. Rome (ed. Tenney Frank), I 195–7, 402–3; II 310–12; III 264–9; IV 347–8, 383–6; V 318. In Revelation VI 6, wheat costs three times as much as barley. 113 See Theophr., HP VIII 8.2, where Attica is actually described as kriuow
oroB ’ istZ [‘excellent barley-bearing land’], a text of which too little notice has been ar
taken. 114 This invalidates the arguments of Kahrstedt, SSA 251 n. 2. 115 See Endnote vii (p. 66 below). 116 See Jarde´, CAG I 186–7. 117 The one piece of evidence I know from Athens which gives corn and oil prices simultaneously is an early fourth-century sacrifice tariff: IG II2 1356. 7–8, 13–14, 17–18, 21–2. The values work out at 3 dr. per med. for wheat and 12 dr. per metr. for oil. See also Jarde´, CAG I 185–6; W. K. Pritchett in Hesp. 25 (1956), at p. 184. 118 e.g. B. Bu¨chsenschu¨tz, Besitz u. Erwerb im griech. Alt. (1869), 54–6; De Sanctis, A2 235–6; Beloch, GG II2 I 302–3; Ed. Meyer, GdA III2 605–6. Contrast

34

Athenian Democratic Origins

Cohen119 have wisely remarked, ‘Au fond, ces calculs, qui auraient un grand inte´reˆt si les bases en e´taient suffisantes, ne me`nent pas a` grand’ chose, parce que le remplacement des matie`res se`ches par des liquides en bouleverse les donne´es et se preˆte a` toutes les combinaisons’ [‘in fact, these calculations, which would be of great interest if they rested on sufficient foundations, are of no significance, since the replacement of dry material by liquids turns the premises upside down and allows innumerable combinations’]. The figures Glotz himself gives elsewhere120 (allowing wide margins) for vineyards, land under corn, and mixed estates are probably about as near as we can hope to get. Jarde´121 has independently tried to give an estimate of the yields obtained from a hectare of land in ancient Greece under wheat, barley, wine and oil respectively, and the comparative money values of these crops at various times between the fourth and second centuries B.C. On the whole this is the most useful guide available, except that Jarde´ has unfortunately been guilty of an oversight all too common among modern scholars: he has failed to allow for alternate years of fallow for land devoted to the production of cereals.122 This of course has made his figures for barley and wheat appear twice as great as they should be, compared with the figures for olives and vines. (b) The name Pentakosiomedimnos is surprising if metretes and medimnos were automatically equated—why was it not Pentakosiometros? Some have supposed that olive cultivation was as yet little developed,123 but this does not entirely dispose of the difficulty which arises from the application of the name Pentakosiomedimnos to a man who produced 500 measures dry or liquid. (c) No provision is made under this system for the man who devoted any appreciable part of his land to producing things which could not be measured in medimnoi or metretai, or not on the same basis: not only the man who disposed of his figs, nuts, olives or grapes from the tree (he could hardly be allowed to reckon this the comments of K. H. Waters, in JHS 80 (1960), at p. 183, and A. French, in Historia 10 (1961) 510–12. And see the next note. 119

HG I 415 n. 151. See also A. French, loc. cit. AGW 246–7 (¼TGA 296–7). CAG I 186–7, esp. 186 nn. 3–4. 122 This was noticed by C. M. A. Van den Oudenrijn, in Mnemos., 4e se´r., 5 (1952) 19–27, at pp. 22–3. 123 e.g. E. Cavaignac, Population et capital dans le monde me´dit. ant. (1923), 51. 120 121

The Original Qualifications of the Pentakosiomedimnoi

35

kind of produce according to the number of medimnoi of fruit!), but also the man who pastured horses, cattle, sheep, goats or swine. Yet as early as Solon’s day we hear of a rich Athenian, the first Alcmaeon, who won an Olympic victory with a four-horse chariot.124 The breeding of horses, as we saw earlier,125 was the mark of ’ vn a very rich man (at Athens Alcmaeon, the elder Miltiades e oi’ k
iZB teurippotr
owoy [‘being of a house which could rear chariot-teams’],126 the elder Cimon,127 and later Alcibiades),128 and it must have involved using a certain amount of precious land as pasture. 2. In a brief article which appeared just over thirty years ago Miss K. M. T. Chrimes,129 attacking what she called ‘the problem of the ‘‘wet and dry measures’’ ’, began with a shrewd exposure of the difficulties of the Aristotelian account. As she rightly observes, ‘the assumption that a landowner might produce his measures either in barley only, or in wine only, or in any combination of these commodities that he chose, results in an impossible scheme.’ She then proceeds to summarise Jarde´’s figures: ‘A given area in Attica’, she says, ‘would yield about twice as much wine as barley, while the ratio of olive-oil to barley from the same area would be about one-third.’ That is a fair summary of what Jarde´ says; but, for the reasons indicated in para. 1 (a) above we must halve the quantity of barley, and see the ratio of production of wine to barley as of the order of 4 : 1, and of barley to olive oil 4 : 3 or a little more—both these figures, of course, being only the loosest of approximations. When Miss Chrimes adds, ‘practically no wheat was ever grown in Attica’, she is exaggerating.130 Miss Chrimes goes on to infer (evidently adopting, it must be admitted, rather extreme figures) that ‘a man could be either Pentakosiomedimnos or Thes with the same plot, according to what he chose to cultivate’ (if we take Aristotle’s statement literally, that is); and she expresses the opinion that ‘it would not only have been an illogical, irritating, and bewildering policy, but a suicidal one, to admit the wine-growers to high political privileges while excluding the 124 126 127 128 129 130

125 Hdts VI 125.5. See above, p. 16. Hdts VI 35.1; 36.1; 103.2. Hdts VI 103.2–3. Thuc. VI 16.2; cf. 12.2. ‘On Solon’s Property Classes’, in CR 46 (1932) 2–4. See Endnote viii (p. 67 below).

36

Athenian Democratic Origins

corn-growers, on whom the very life of the country depended’. Although all this is also exaggerated, there is certainly something in it. I might add that another odd element is introduced if we admit that a given piece of land would have been likely to produce fewer metretai of olive oil than either metretai of wine or medimnoi of corn. Solon’s prohibition of the export of natural products other than olive oil131 suggests that he wanted to encourage the production of oil. Why then should he in effect rate the possessor of a given plot of land at a lower level if he grew olives than if he cultivated vines or cereal crops? Miss Chrimes’s positive proposals have found no favour at all, as far as I know; and this is not surprising, for they presuppose an elaborate and quite unnecessary survey (on principles never known to have been adopted in ancient Greece)132 of all the cultivated land of Attica, the ultimate object of this survey being ‘to express the area of any holding in terms of the average yield of barley from a plot of that size, even though only a part of it, or perhaps none of it, was actually used for growing barley’.133 However, even here there is a step in the right direction, as will become apparent in due course. 3. George Thomson, in an article published in 1953,134 made a suggestion in two sentences which puts us on the right track; but he did not develop it or provide any direct evidence in its favour, and it seems to have been largely overlooked.135 In Solon’s classification, says Thomson, ‘the annual yield of corn, oil and wine was assessed in terms of corn.’136 For ‘corn’ we must of course read ‘barley’. If, then, wheat was reckoned twice as valuable as barley, and the value of a metretes of oil was, shall we say, twice as high again, a man could qualify as a Pentakosiomedimnos if he produced 500 medimnoi of barley or 250 medimnoi of wheat or 125 131

Plut., Sol. 24.1–2. Neither Jarde´, CAG I 314, to which Miss Chrimes refers, nor Jarde´’s article, ‘PentZkont
axoyB’, in REA 12 (1910) 373–6, provides any foundation that I can see for her statement that ‘the plot of land in fourth-century Attica was expressed in terms of the amount of barley required to sow it’. 133 Op. cit. (n. 129 above) 4. Cf. id. 3: ‘Any given area of Attic land would become translatable in terms of the amount of barley which it would produce.’ 134 ‘On Greek Land-Tenure’, in Stud. Pres. to D. M. Robinson II (1953) 840–50, at p. 848. 135 I am grateful to Miss L. H. Jeffery for drawing my attention to this article. 136 Thomson refers to the situation in modern Syria, where, he says, ‘a camel’s load of corn’ serves the same purpose as a Solonian ‘measure’. 132

The Original Qualifications of the Pentakosiomedimnoi

37

metretai of oil—or, say, 200 medimnoi of barley, 50 medimnoi of wheat and 50 metretai of oil all told. (Of course I do not mean to imply that there was any fixed tariff.) The fact that a given area of land would yield fewer medimnoi of wheat than of barley, and probably still fewer metretai of oil, would be compensated for by the relatively higher value of the wheat and oil, and a reasonably satisfactory situation would be established. This explanation also accounts for the name Pentakosiomedimnoi. 4. An essential point is that valuation is terms of barley and not money, or even silver by weight, would be a very natural phenomenon in early sixth-century Attica. The drastic lowering of the date of the first introduction of coinage, in Lydia and Ionia, to somewhere about the middle of the second half of the seventh century,137 has had enduring repercussions. As early as 1950 W. L. Brown138 showed that an eighth or seventh century Pheidon cannot have coined at Aegina, and that the first Corinthian coins are not likely to be earlier than the second quarter of the sixth century. In 1956 C. M. Kraay139 gave a masterly demonstration, by a series of converging arguments, that we must now date the first Athenian coins, the ‘Wappenmu¨nzen’, to c. 575–525, and that the earliest ‘owls’ do not begin until the 520s. The old notion that the earliest coins of Aegina must be earlier than those of Athens or Corinth may still be true, but there is no real evidence. E. S. G. Robinson in 1951 was prepared to treat the earliest Aeginetan coinage as the first in the Greek homeland, but would not date it before the last quarter of the seventh century. Kraay now tells me that there is no evidence at all to show that Aegina 137 E. S. G. Robinson, ‘The Coins from the Ephesian Artemision Reconsidered’, in JHS 71 (1951) 156–67, with P. Jacobsthal, ‘The Date of the Ephesian Foundation-Deposit’, ibid. 85–95. 138 ‘Pheidon’s Alleged Aeginetan Coinage’, in NC Ser. VI 10 (1950) 177–204, rightly described as ‘an excellent article’ by Robinson, op. cit. 166 n. 65. 139 ‘The Archaic Owls of Athens: Classification and Chronology’, in NC Ser. VI 16 (1956) 43–68, recently endorsed by E.S.G. Robinson, in NC Ser. VII 1 (1961) 107 ff., at p. 109. It is a great pity that Hammond, in his HG, esp. p. 659, has paid no attention to the compelling arguments of Robinson, Brown and Kraay. (In this he was apparently influenced by C. Seltman, who in the second edition of his Greek Coins [1955] similarly refused to take account of the new evidence.) I may add that the conclusions Hammond seeks to draw from the Perachora drachma inscription, which he dates before the 640s, are quite unjustified: according to the leading authority on early Greek scripts, the lettering of that inscription ‘appears to be no earlier than 600 B.C. and possibly as late as 550 B.C.’ (Miss L. H. Jeffery, ap. R. M. Cook, in Historia 7 [1958] 258 n. 12).

38

Athenian Democratic Origins

began to coin much before Athens and Corinth, and that although the first ‘turtles’ may be late seventh century they are just as likely to be c. 575 or even a little later.140 It is possible, therefore, that coinage was virtually unknown in Solon’s Attica, except to those who had travelled to western Asia Minor. At the date of Solon’s legislation, then, neither Athens nor any other Greek state on the west side of the Aegean was coining money, with the possible exception of Aegina, and in Attica of the 590s coinage could have been little used for everyday transactions, if at all. Barley had been employed for centuries in the Mesopotamian area and elsewhere as a standard of value, alongside silver,141 and since at this time it was the staple crop in Attica it was quite natural that the produce of Athenian farmers should, for the purposes of the Solonian tele¯, be assessed in terms of barley. The ox was hardly a possible alternative, especially since Solon was legislating in terms of annual produce.142 All this may incline us to feel sceptical when we are presented with alleged ‘laws of Solon’ in which values are given in drachmae. Of course, a ‘drachma’ was a weight before it was ever a coin, and it is quite possible that Solon may have expressed some values in terms of silver by weight even if coined silver hardly circulated in Attica in his day. To say that any ‘Solonian’ law which evaluates in terms of drachmae must be a later revision or invention would be to go too far. But where Solon did not state values in drachmae, his omission would sometimes be repaired later, after the use of coinage had become general, and his laws would then surely be quoted, still on occasion as ‘laws of Solon’, with the drachma values inserted. This may even be true where a Solonian Axon is 140 [C. J. Howgego tells us that the statements in this paragraph are still valid: cf. pp. 414–16 below.] 141 [See A.L. Slotsky, The Bourse of Babylon: Market Quotations in the Astronomical Diaries of Babylon (1997), 24–8.] 142 Poll. IX 61 refers to a Draconian law in which a 20-ox fine is prescribed; but, as Miss L. H. Jeffery has pointed out to me, the homicide laws were in a sense sacred laws, to whose conservatism on this point another witness exists in the latter part of ’ tfi Z& par the same passage (IX 61): kai en a DZl
ioiB uevr
ifia t on k
Zryka kZr
yttein ‘ ote dvre
’ v was
in, op
a tini d
idotai, o‘
ti b
oeB toso &ytoi dou
Zsontai ayt fi & , kai d
idosuai aB ’Attik
aB [‘and in the Delian theoria, they say that kau’ e
‘ kaston bo &yn d
yo draxm the herald announces, when a gift is made to anyone, that so many oxen will be given to him, and two Attic drachmae are given for each ox’]. It is absurd to argue, as for example Gilbert has done (CASA 122 n. 1), that ei’ kos
aboion [‘20 oxen’] was ‘only an antiquated expression for 40 drachmae’ in Draco’s day.

The Original Qualifications of the Pentakosiomedimnoi

39

quoted by number, as in Plut., Sol. 23.3–4. By the fourth century even well informed Athenians like Androtion143 and Aristotle144 believed that Athens itself had coined money even before Solon’s day. 5. The theory of a ‘barley standard’ for the assessment of the qualifications of the Pentakosiomedimnoi (for we are not concerned at the moment with the other classes) becomes very much stronger when we take notice of a law of Solon which has not previously, I think, been utilised in this context. An Athenian law referred to in Isaeus and other sources145 provided that a woman146 might not symb
allein p
era med
imnoy kriuvn, & which I take to mean, ‘enter into any kind of contract involving the worth of more than a medimnos of barley’.147 This law must date from before the time at which coinage came into common use at Athens, or it would surely have expressed the limiting value in drachmae; and there is a high probability that it is Solonian. Since in this law ‘a medimnos of barley’ must mean ‘a medimnos of barley or its equivalent in value’, we are justified in supposing that it can mean this, in suitable cases, in other laws of Solon—in particular the law regarding the qualifications of the tele¯, where this interpretation makes very much better sense than Aristotle’s. 6. A further argument in favour of accepting the medimnos as, so to speak, the unit of a ‘barley standard’ is provided by the wording of Plut., Sol. 23.3: ei’ B m
en ge t a tim
Zmata tvn & uysivn & 148 143

144 FGrH III B 324 F 34, ap. Plut., Sol. 15.3–4. Ath. Pol. 10. Isae. X 10; Ar., Eccl. 1024–5, with Schol.; Dio Chrys. LXXIV 9; Harp. (Phot. and Suid.), s.v. o‘
ti paid
i. 146 And just possibly also a minor, as Harp. etc. believed: see Wyse, SI 659–60. But most other recent writers have thought that Harp. made a false inference from the ambiguous wording of the passage in Isaeus: o‘ g ar n
omoB diarr
ZdZn kvl
yei ’ inai symb
paidi m Z eje& allein mZde gynaiki p
era med
imnoy kriu vn & [‘for the law expressly forbids a child from entering into any kind of contract [sc. under any circumstances], and forbids a woman from doing so when it involves the worth of more than a medimnos of barley’]—see e.g. Lipsius, ARR 502 n. 12, who appeals to Dio’s version. 147 LSJ9, s.v. symb
allv I 8, give the Isaeus passage with the interpretation ‘advance, lend’; but this is surely to narrow the scope of the provision without any justification. The word symb
allein has a very wide range of meanings, like the corresponding noun, symb
olaion—LSJ9, s.v., I, give sufficient examples (add Isocr. XVIII 18, where the word might be translated ‘dealing’). See also my remarks on this word in my NJAE I, at pp. 101–4, esp. 101 n. 6. Closely parallel are synall
attein and syn
allagma, and it is synall
attein which Dio uses in his version of the law in question. 148 ’ vn, On Wilcken’s emendation, oysi & see p. 45 below. 145

40

Athenian Democratic Origins

’ i med
imnoy, ‘in the valuation log
izetai pr
obaton kai draxm Zn ant of sacrifices he reckons a sheep (or goat) and a drachma as the equivalent(s)149 of a medimnos’—not ‘a sheep and a medimnos as the equivalent(s) of a drachma’. The basic unit is the medimnos. The drachma150 may not have been added to the law until a time when coinage had become an established medium of exchange, and fines, rewards and values in written Codes were altered accordingly—if Solon himself had used the expression, would he not have ’ yroy [‘drachma of silver’]? spoken of a draxm Zn arg
7. The ‘barley standard’ theory even enables us to explain the difficulty stated in para. 1 (c) above, about the man who sold his fruit from the tree, or possessed an appreciable quantity of pasture land. The former case needs no explanation. In the latter, the ’ tZB & oi’ ke
iaB [‘from his own land’] would presumably ‘income’ ek be the foals, calves, lambs, kids, piglets born in the year, valued in medimnoi of barley. 8. It seems, then, that Aristotle was wrong in his statement of the original qualification of the Pentakosiomedimnoi: m
etra t a ‘ a [‘measures both dry and liquid together’]. syn
amwv jZr a kai ygr
This need not surprise us, as we have already seen reason to conclude that he did not have the text of the law defining the tele¯. But was he also wrong about the necessity for the income of the Pentakosiomedimnoi (as of the other classes) to be derived ‘from their own land’? This question is of some importance for our understanding of the nature of Solon’s constitution. We are not concerned at this stage with the question whether the qualification of the Pentakosiomedimnoi (as of the other census classes) was later changed, either so as to include other income than that from land, or into its equivalent in money capital. All we have to decide now is whether the original qualification was or was not purely in terms of landed produce. Some, including Eduard Meyer,151 have The question is whether pr
obaton kai draxm
Zn means ‘a sheep and a drachma’ or ‘a sheep or a drachma’. 150 ‘If it does not stand here for sacred spits’, as Miss Jeffery remarks in a letter to me. 151 GdA III2 607: ‘Die reichen Fabrikanten und Kaufleute, die keinen Grundbesitz hatten . . . , sind offenbar nach ihrem Einkommen den Klassen zugewiesen worden, wobei man nach damaligem Marktpreis (Plut. Sol. 23) den Scheffel zu 1 Drachme rechnete’. Other scholars have held much the same view, e.g. Bruno Keil, SVAVA 68–70 (cf. n. 189 below). Among more recent writers, K. Freeman, The Work and Life of Solon (1926), 58–9, gives a very unsatisfactory discussion, which can best be summed up in her own words: ‘It is safest to hold to the tradition 149

The Original Qualifications of the Pentakosiomedimnoi

41

thought it ‘obvious’ that other income was included from the first in the qualification of each of the classes, a drachma being taken as the equivalent of a medimnos. Actual evidence in favour of this view is entirely lacking, unless we accept (as I of course do not) ’ vn Wilcken’s emendation, oysi & [‘property’] for uysivn & [‘sacrifices’], in Plut., Sol. 23.3, which I shall discuss presently. Some modern scholars who have taken it for granted that Solon introduced classification by money income have drawn far-reaching conclusions from their assumption. Heichelheim,152 for instance, can say that Solon’s law ‘according to which the membership of Athenians in the four new political classes of the polis was not only to be decided by their wheat harvest,153 as had been the case before Solon, but in addition by their adaerated income from cattle, olive oil, and money, must have brought landowners of the middle class, rich merchants, and craftsmen into the higher classes.’ 9. There are two arguments which seem to me to rule out completely any theory which supposes that annual produce in kind could ever have been equated, for census purposes, with its approximate equivalent in money income. (a) An annual income in medimnoi/metretai stands for (and guarantees the possession of) capital, in land, whereas money income need not—and, if it is earned income, will not—stand for any capital at all. Therefore money income cannot be directly equated with income in kind from land for assessment purposes. This argument alone is decisive. It would be ridiculous to equate a man who received 500 or more measures of barley (or their equivalent) from his own land, and who would necessarily be quite a big landowner by Athenian standards, with a man who merely earned that Solon assessed the classes in terms of produce from land; and to conjecture that he did not expressly insist on a land-qualification, so that the way was left open for the substitution of money-income for produce-income.’ J. Johnston, in JHS 54 (1934), at pp. 180–1, and K. H. Waters, in JHS 80 (1960), at pp. 182–3, assume that Solon allowed reckoning in money income. Waters is criticised, but inadequately, by A. French, in Historia 10 (1961) 510–12. Hammond, HG 160, says cryptically that Solon’s census was based ‘on annual income assessed in terms of agricultural produce—cereals, oil, and wine—with which other forms of income were integrated’. 152 AEH I 285. It is interesting to find ‘landowners of the middle class, rich merchants, and craftsmen’ substituted for ‘die kleinen und mittleren Bauern’ of the original German edition (WA I 286). 153 If this is still intended to represent ‘Getreideproduktion’, as in the original German edition, it is a mistranslation: for ‘wheat’ read ‘cereal’.

42

Athenian Democratic Origins

500 drachmae or even a great deal more by his own skill and might not have a penny of capital. Assessment in terms of income in kind from land was a simple but effective way of assessing land ownership. Valuation of land, at a time when it was rarely bought and sold and had nothing that could legitimately be called a ‘market value’ in the modern sense, would scarcely have been practicable, except perhaps by taking the value of the annual produce and multiplying it by some arbitrary figure (10 or 12, for instance), so as to assign to it a fictitious capital value in silver—a pointless activity, given the economic conditions of Solon’s time, and not one which I can imagine Solon wishing to carry out. The value of a man’s land could most easily be assessed in terms of its annual produce, calculated according to what I have called a ‘barley standard’. The Pentakosiomedimnos, in this sense, would be a man with plenty of capital in the safest of forms, land—and this was what the state wanted above all: the main reason why the treasurers of Athena had to be drawn from the highest property class was that the state had to be sure of being able to recoup itself out of the treasurer’s own property if he embezzled any of the valuables that passed through his hands. (b) It would be impossible to assess manufacturers, merchants and ship-owners fairly on an income basis, for two different reasons. First, in the still non-monetary economy of early sixthcentury Attica a certain amount of trade might take the form of sales at ‘prices’ expressed in quantities of barley, or silver by weight, but much would be pure barter: this does not permit the ‘profit’ on even a single transaction to be assessed in quantitative terms, much less on a year’s trading. An Athenian merchant of Solon’s day (or, for that matter, the fifth or fourth century) could have no conception of an ‘annual net profit’ in the modern sense, or anything like it. And secondly, the accounting system of antiquity, even in the heyday of Greek and Roman civilisation, was nothing like as efficient as modern accounting, and in particular could not distinguish properly between capital and income. (I have described the principal features of ancient accounting elsewhere154 and need say no more about the subject here.) A manufacturer or trader, even when the use of money became general, would simply 154 ‘Greek and Roman Accounting’, in Stud. in the Hist. of Accounting, ed. A. C. Littleton and B. S. Yamey (1956) 14–74.

The Original Qualifications of the Pentakosiomedimnoi

43

not know what his ‘income’ or his ‘profits’ expressed in terms of drachmae were. This is one of the basic facts about the economy of the Greek world (and the Roman world) which many modern historians have entirely overlooked, because they persist, quite unconsciously, in conceiving the ancient economic systems in terms taken over directly from the modern or the medieval world. No characteristic of the economy of modern or even medieval Europe can safely be assumed to have been present in that of ancient Greece until actual evidence of its existence there has been found. 10. By the end of the fifth century, as we know from the Erechtheum accounts,155 wage rates of one drachma per day were common. The daily pay of sailors in the fleet was also between one drachma per day (as in the Potidaean and Sicilian campaigns)156 and half a drachma (as in the Ionian war),157 and the daily pay of dicasts was half a drachma from 425 onwards.158 A penniless man, therefore, might earn upwards of 300 drachmae a year, and qualify as a Hippeus, if in the Solonian scheme money had been allowed to count equally with landed produce at the rate of one drachma per medimnos, or if that equivalence was subsequently introduced—unless of course the qualifications were substantially raised before the second half of the fifth century. 11. We must consequently reject any theory which could establish monetary equivalents for the Solonian ‘measures’, in Solon’s day or at any subsequent time. As the name ‘Pentakosiomedimnos’ proves, Solon chose to assess at any rate his highest class in terms not of capital but of a kind of income which accurately reflected the ownership of ‘capital’, in land. Manufacturing and trading profits, as I have shown, were not ascertainable. And earned income, even in a theoretically assessable form such as a regular wage,159 was not comparable in principle, for the reason that it did not represent any ‘capital’. Nor was capital wealth other than land quantitatively comparable in practice: in early sixth-century Attica there can 155

IG I3 474–9; L. D. Caskey in The Erechtheum (1927), ch. iv. See R. H. Randall, ‘The Erechtheum Workmen’, in AJA 57 (1953) 199–210. 156 Thuc. III 17.4; VI 8.1; 31.3. 157 Thuc. VIII 29.1; 45.2; Xen., Hell. I v 4–7. 158 Ar., Eq. 51 etc., with Schol. Vesp. 88. 159 In fact very few Athenian citizens (and those mainly unskilled men) would work for a regular wage.

44

Athenian Democratic Origins

have been no ‘market’ which would enable any asset whatever to be assigned a capital value.160 The approximate value of ordinary everyday commodities would be known, in terms of each other and of barley and probably silver, but major items such as workshops, mines and merchant ships, and above all land,161 could hardly have been thought of as having a ‘market value’—and could not, therefore, be equated in an assessment. Official assessments of income in terms of money were never made in antiquity, pre-Classical, Greek or even Roman, just as taxes were never levied upon income. We hear of a great variety of taxes in the ancient world, especially from Ptolemaic and Roman Egypt, but I have never been able to discover anything in the shape of an income tax, except of course for one type of tax which was indeed widely prevalent (though not, as far as our information goes, in the free Greek cities, except under tyrants), namely, a proportion of agricultural produce in kind—a ‘tithe’ in the loose sense. Such a tithe could be levied, because it was a direct proportion of the crop and did not require to be expressed in money. An income tax of any other sort was impossible, precisely because people did not reckon their incomes in money, and indeed could hardly have done so without a complete change in methods of reckoning and accounting, including a much clearer distinction than was ever made in antiquity between capital and income. As I have said earlier in my paper on the fourth-century eisphora,162 ‘It is unlikely that any Greek ever had occasion to reckon his total income in money, and there is no evidence that estimates of wealth, official or unofficial, were ever made in such terms. In Plato’s Laws (XII 955DE) the two alternative forms of eisphora are on total capital and on agricultural produce. At Athens it was the possession of 300 drachmae capital which disqualified a man from receiving a state pension as ’ ynatoB (Arist., Ath. Pol. 49.4).’ If the qualifications for the an ad
160 Here I should like to mention an interesting book which I think has received much less than its due share of attention from Greek historians: Trade and Market in the Early Empires, ed. Karl Polanyi and others (1957). Although, as I made clear in my review in Econ. Hist. Rev. Ser. II 12 (1960) 510–11, some of the opinions expressed in this book about the economy of ancient Greece rest upon insufficient first-hand knowledge of the sources, and there are serious errors, yet some of the basic ideas are good, and Greek historians might apply them with profit—in particular, the warning that a full and free ‘market economy’, in the modern sense, cannot be taken for granted until evidence of it appears. 161 162 (no note). DTAEFC 41.

The Original Qualifications of the Pentakosiomedimnoi

45

Athenian tele¯ were ever turned into their monetary equivalents in later times (a possibility we shall consider later), they must at the same time have been capitalised. 12. It is convenient to deal here with the brilliant, and palaeographically plausible, but nevertheless definitely wrong, emendation of Plut., Sol. 23.3, by Wilcken,163 which has received a good & deal of attention in recent years.164 With the alteration of uysivn ’ vn, ’ vn to oysi & the text reads ei’ B m
en ge t a tim
Zmata tvn & oysi & ’ i med
imnoy, so that Plutarch log
izetai pr
obaton kai draxm Zn ant is not speaking about a sacrifice tariff at all, but is saying that for the purpose of fixing the assessments of property the Solonian law ‘reckons a sheep (or goat) and a drachma instead of a medimnos’. It is worth mentioning that the passage goes on at once to say that an Isthmian victor got 100 dr. and an Olympian victor 500 dr., and that a man who brought in a wolf got 5 dr., a wolf cub 1 dr.; and Plutarch adds that according to Demetrius of Phaleron 5 dr. was the price (tim
Z) of an ox, 1 dr. of a sheep. It is also material that Plutarch proceeds to say that the tima
i which Solon fixes in his Sixteenth Axon are for choice sacrificial animals, naturally many times as high as normal, but even then low compared with prices in Plutarch’s own day.165 I may remark that if Wilcken’s emendation were justified, then Aristotle—not to mention Plutarch—would have failed to give us a not unimportant piece of information; for it makes a real difference to our picture of Solon’s constitution whether he made his three top classes, or at any rate the highest of all, depend entirely upon the ownership of land or allowed other forms of wealth to count equally. To acquit Aristotle and Plutarch of this charge it would be necessary to suppose that the rule equating sheep, drachma and 163

‘Zu Solons Schatzungsklassen’, in Hermes 63 (1928) 236–8. Wilcken’s emendation has been accepted by several scholars, including Heichelheim, AEH I 253–4 (¼WA I 255–6), Bengtson, GG2 121 n. 3, and (with quite a different interpretation) by J. H. Thiel, in Mnemos., 4e se´r. 3 (1950) 1–11; cf. Van Groningen as cited in n. 167 below. Wilcken and Thiel have been criticised by C. M. A. Van den Oudenrijn, in Mnemos., 4e Se´r. 5 (1952) 19–27; and Thiel has accepted his criticisms in ibid. 27. Cf. also E. Will, in Annales 9 (1954) 21; Kahrstedt, SSA 255–6. 165 For a useful collection of all the relevant material on prices, see W. K. Pritchett, in Hesp. 25 (1956) 255–8 (oxen), 259–60 (sheep): the best attested prices for animals, which come from the fourth century, are around 70–90 dr. for oxen and 12–15 dr. for sheep, all sacrificial animals. [F. van Straten, in T. Linders and G. Nordquist (eds.) Gifts to the Gods (1987), 159–70.] 164

46

Athenian Democratic Origins

medimnos for property assessments was a post-Solonian addition to the original law. It will be immediately obvious that Wilcken’s emendation is entirely incompatible with what I have been saying above; and if I have proved that there can never have been an assessment in terms of money income, then Wilcken’s theory falls to the ground.166 I do not think I need discuss it further. Wilcken’s position has been reformulated by van Groningen167 in a form which does not leave it open to quite such fatal objections as those set out above; but even so it seems to me impossible. As I do not think many people will be attracted by van Groningen’s hypothesis, I have relegated my discussion of it to an Endnote.168

(VI) THE ORIGINAL QUALIFICATIONS OF THE HIPPEIS AND ZEUGITAI 1. We have seen that Aristotle (Ath. Pol. 7.4) seems not to have known for certain what the original qualification of the Hippeis was, and that he is very probably wrong about that of the Pentakosiomedimnoi. Now when he describes the qualification of the Zeugitai169 he employs the same phrase he has just used of the Pentakosiomedimnoi, for toyB diak
osia t a syn
amwv poio &yntaB [‘those producing 200 together’] can stand for nothing but toyB ‘ a poio &yntaB [‘those diak
osia m
etra t a syn
amwv jZr a kai ygr producing 200 measures, dry and liquid together’]. And if we believe that Solon’s ‘measures’ represented a ‘barley standard’,170 we cannot accept Aristotle’s formula, which we shall then treat as evidence that he did not have the law giving the qualification of the Zeugitai either. And we cannot say here, as we could in the case of the Pentakosiomedimnoi, that the very name of the telos shows the number of measures at any rate to be right. The suspicion will then 166

I see no objection to uysi vn. & Cf. Van den Oudenrijn, op. cit. (n. 164 above)

24–7. B. A. Van Groningen, Aristote, Le second Livre de l’E´conomique (1933) 51 n. 2. Endnote ix (p. 67 below). 169 No source of any value, as far as I know, apart from Aristotle and Plutarch (Sol. 18.1), and Poll. VIII 132 (see n. 195 below), purports to give the original qualification of the Zeugitai. The only other conceivably relevant text is the law cited in Ps.-Dem. XLIII 54; and this law says nothing about the qualifications of the classes. 170 See pp. 36–40 above. 167 168

The Original Qualifications of the Hippeis and Zeugitai

47

grow upon us that Aristotle may have been basing his conclusions upon some law or laws resembling that quoted in Ps.-Dem. XLIII 54, which gave figures of 500, 300 and (instead of 150, as in Ps.-Dem. XLIII 54) 200 drachmae for something that had to be done by members of the respective tele¯. Aristotle had copies, whether accurate or not, of some obsolete laws which were, or were believed to be, laws of Solon, and some of these were certainly concerned with financial matters: Aristotle (Ath. Pol. 8.3) actually quotes from them the expressions toyB naykr
aroyB ei’ spr
attein ’ ’ to &y naykrariko &y [‘the naukraroi shall exact’] and anal
iskein ek ’ argyr
ioy [‘pay out from the naukraric silver’]. If in such a law a figure of 500 drachmae for a Pentakosiomedimnos occurred in conjunction with 300 for a Hippeus and 200 for a Zeugites, Aristotle may well have jumped to the not unreasonable conclusion that the member of each class (and not only the Pentakosiomedimnos) was being made liable in proportion to the value of his qualifying assessment. Or, even more plausibly, some such false inference had already been drawn by one of the early Atthidographers and had become generally accepted, so that Aristotle was simply repeating what everyone believed. 2. Quite apart from the fact that we have reason to challenge Aristotle’s authority as a source in this case, I myself find it impossible to believe that the relationship between the minimum assessment of the Zeugites and the Hippeus can have been 200 : 300. There are two separate arguments here: (a) I cannot doubt, in view of the name of Solon’s second class, that they were men who at least might be expected to i‘ ppotrowe&in [‘rear/keep horses’], even if that was not itself the original ground of qualification (as I believe, with the e’
nioi contradicted by Aristotle, Ath. Pol. 7.4).171 But maintaining a warhorse was always regarded as a very expensive activity which only a really rich man could be expected to undertake.172 If, then, the possession of an annual income of 300 medimnoi put a man into the class of Hippeis, is it credible that two-thirds of that amount (two-thirds of the income of a really rich man) was the minimum amount required to qualify as a hoplite? To me this seems incredible. (b) The class of Hippeis, like that of the Pentakosiomedimnoi, must have been small in the early sixth century. It is impossible 171

See pp. 48 ff. below.

172

See pp. 16 and 35 above.

48

Athenian Democratic Origins

to estimate their numbers, but I think most people would be content with a maximum figure of two or three hundred. The zeugite class must have been very much larger: if the Zeugitai were the hoplites (a question we shall come to presently), they are likely to have numbered some three to five thousand,173 and therefore to have been at least ten times as numerous as the two higher classes together. Is it possible to believe the demographic curve dropped so sharply at a point corresponding to the possession of 300 ‘measures’ that several thousand Athenians had incomes in the narrow range between 200 and 300 ‘measures’, but that there were only a few hundred who had over 300 ‘measures’—especially when a significant number (enough for a whole telos to themselves) certainly had more than 500 ‘measures’? Again, I cannot believe it. 3. The theory which in my opinion fits better than any other all the historical facts, especially the puzzling silence of the sources regarding any change in the qualifications of the census classes, is that the qualifications of Hippeis and Zeugitai were always exactly what we have seen them to have been in the late fifth and fourth centuries: the Hippeis were, broadly speaking, those who could afford to maintain horses for military use, and the Zeugitai were the hoplite class,174 and there was no quantitative time¯ma in either case. Again, the relationship between the census classes and the military categories has generally been conceived the wrong way round, for Solon’s day as for the later period: a man did not become a cavalryman (or ‘mounted infantryman’) or a hoplite because he was qualified as a member of one of the middle two census classes according to a fixed numerical assessment (in ‘measures’ or money); it was being a cavalryman or a hoplite which made him a Hippeus or a Zeugites. According to my interpretation, Hippeis and Zeugitai (and therefore Thetes) corresponded with military and social groups that already existed in Solon’s day, and only 173 As Jones (AD 8) says, our earliest figures, which are for a century after Solon’s time, suggest a total citizen population of c. 30,000, divided 1 : 2 between hoplites and those below hoplite status. Our earliest hoplite figures, for what they are worth, are the ?9,000 for Marathon, in Nepos, Milt. 5.1; cf. Paus. X 20.2 and Justin II 9.9; and the 8,000 for Plataea in Hdts IX 28.6. We must of course allow for a considerable growth of population during the sixth century, and for an increase in general wealth which would cause men below hoplite status to rise into the hoplite class. 174 I have attempted more precise definitions above, pp. 23–4.

The Original Qualifications of the Hippeis and Zeugitai

49

Solon’s Pentakosiomedimnoi (really the top layer of the Hippeis) were an entirely new creation, and an artificial one. This fact is reflected in the nomenclature of the census classes: the Pentakosiomedimnoi had a name of a different kind from the rest. I cannot see that any formal minimum qualification in terms of property would be necessary or even desirable for either military category, and I suggest that to fix one at all would have had obvious disadvantages. The most Solon could hope to do was to get the qualification ‘right’ in each case, in the sense that those who possessed it would coincide with those who were financially able to perform the duties associated with membership of the classes concerned. If the qualification was not ‘right’ in this sense, either the military category concerned would be unnecessarily narrowed or men would be forced into it who were not really capable of sustaining its burdens. (Aristotle—who does not specifically connect Solon’s Hippeis with cavalry and Zeugitai with hoplites— may conceivably have supposed that the time¯mata he represents Solon as fixing, 300 and 200 ‘measures’, represented roughly the amounts of property needed to enable men to serve as cavalry and hoplites respectively.) Now a quantitative time¯ma for the exercise of political rights in an oligarchy, or as a condition of eligibility for certain magistracies in a democracy or oligarchy, may be appropriate enough; but to fix such a time¯ma seems to me absurd if what is wanted—as in our case—is the largest possible number of men who can be made liable for military service. Since membership of each telos was a privilege from the first, as giving access to office, Solon could reasonably expect no lack of suitable volunteers; and, as I argued earlier, membership of the hoplite class in particular was something that everyone would covet. 4. No one can doubt the military significance of the word i‘ ppe
yB, but there is a dispute about zeyg
itZB. Some,175 even in recent times, have adhered to the older view, that the word is derived from ze &ygoB in the sense of a ‘yoke of oxen’, and that the zeyg
itZB was the ‘teamster’, the man able to keep a yoke of oxen.176 175 e.g. Busolt, GG II2 182 n. 3; Busolt-Swoboda, GS II 823 and nn. 1 & 2 (and see 820 n. 2); Gilbert, CASA 116 and n. 2; De Sanctis, A2 231 n. 1; Hignett, HAC 101 n. 1. 176 ’ eloyn [‘and Cf. Poll. VIII 132: kai zeyg
Zsi
on ti t
eloB oi‘ zeygotrowo &ynteB et
the men who could rear a yoke (sc. of oxen) were assessed in the so-called zeugite class’].

50

Athenian Democratic Origins

Others177 have thought that zeyg
itZB here must bear the meaning it sometimes does in later times:178 the man in the battle-line, the ze &ygoB (though zyg
on, zyg
oB, is much more common in this sense). Although I do not regard the question as finally settled, the linguistic evidence so far produced179 is very much in favour of the second interpretation, for, as Beloch put it,180 ‘zeyg
itZB ist der Ochse, der in Joch geht, nicht dessen Herr’ [‘zeyg
itZB is the ox, which goes under the yoke, not its owner’], and if this is true in all contexts it might be held to settle the matter. I have not yet found a single text in which the word zeyg
itZB is used in the sense of ‘teamster’. I very much doubt myself if the possession of a yoke of oxen would necessarily make a man able to perform hoplite service, even in the early seventh century, when, probably, Athens began to create a hoplite army.181 And it is worth remarking that no ancient source suggests that the legal qualification either of hoplite or of Zeugites was possession of a yoke of oxen. 5. Analogy with the Hippeis, for what it is worth,182 suggests a military complexion for the Zeugitai too. And there are further arguments for treating the Zeugitai as the ‘hoplite class’, in the sense in which I have already defined that term. First, we have seen that this was the situation at any rate in the late fifth and fourth centuries, at which time there was no quantitative qualification making a man liable to hoplite service. And I find it hard to see how the Zeugitai could have become the hoplite class if they were not that from the beginning under Solon. And secondly, the hop177 e.g. Cichorius (1894), Lehmann-Haupt (1914) and Beloch (1924): see BusoltSwoboda, loc. cit. (n. 175 above); and see also Andrewes, GT 87; Bengtson, GG2 108 and n. 2. It is interesting to note that Ed. Meyer, who in the first edition of his GdA (II 653) had adopted the older theory, was convinced by Cichorius, and in GdA III2 605 and FAG II 523 accepted the military explanation. 178 e.g. in Plut., Pelop. 23.4. 179 Callim., Hymn. II ad Apoll. 48; Diod. XVII 71.2. 180 GG I2 i 303 n. 1. 181 For the date of introduction of hoplite fighting, see H. L. Lorimer, ‘The Hoplite Phalanx’, in BSA 42 (1947) 76–138; cf. S. Benton, in BSA 48 (1953) 338–9; J. M. Cook and J. Boardman, in JHS 74 (1954) 153 and pl. VIII 5–6. [K. A. Raaflaub, ‘Soldiers, citizens, and the evolution of the early Greek polis’, in Mitchell/Rhodes, DPAG, 49–59; P. Cartledge, Spartan Reflections (2001), 153–68.] 182 The analogy is by no means complete, for although the name Hippeis implies that members of this class would own horses, it is doubtful whether there can have been any duty to keep a warhorse until a cavalry force was created in the mid-fifth century, whereas liability for conscription (of both classes) as hoplites must have existed even before Solon’s time.

Were the Qualifications of the Tele¯ Changed?

51

lites in Solon’s time must already have been a distinguishable group within the state. They were an obvious body to designate for the lower political offices, which did not need to be reserved for Pentakosiomedimnoi or Hippeis. Why should Solon have wished to create another group—which, as many of those who do not accept the complete identification of Zeugitai with the hoplite class have admitted,183 cannot have been very different from the hoplites?184 6. There is another important conclusion we must not fail to draw: that whereas the Pentakosiomedimnoi were qualified as such in terms of landed property alone (I say ‘as such’ because most if not all of them will have been independently qualified as Hippeis), the Hippeis and Zeugitai will have been qualified, through their status as cavalry or hoplites, by their total wealth, in whatever form. Doubtless the great majority of the Zeugitai and all the Hippeis will have been landowners above all; but the great merit of the absence of any specific time¯ma for cavalry or hoplite status was that a man need not rely on his landed property alone, and men might serve in the army (and hold the offices appropriate to their census class) whose purely landed wealth would not by itself have been sufficient to give them the requisite financial status. Metics could not own land (although they could lease it),185 but they could serve as hoplites, and did so in large numbers by the time of the Peloponnesian war186 and probably a good deal earlier.

(VII) WERE THE QUALIFICATIONS O F T H E T E L E¯ C H A N G E D ? We have already seen that there is no evidence whatever for fixed quantitative qualifications, either for the tele¯ in the fifth and fourth centuries or for the military classes (the cavalry and hoplites) at any time, and that there is also no evidence for any change in the qualifications of either category. It is this last point I now wish to emphasise, and particularly the absence of any hint of a change in 183 See e.g. Andrewes, GT 37, 88, where the Zeugitai are ‘roughly’ the hoplites or hoplite class; Hignett, HAC 101 (see n. 74 above). 184 And see para 3, p. 48 above. 185 See Lys. VII 9–10. 186 See Thuc. II 13.7; 31.2; cf. Xen., de Vect. II 2–5.

52

Athenian Democratic Origins

the qualifications of the census classes.187 I would point out that if there was, as the majority of scholars have supposed, a change from assessment by ‘measures’ to assessment in monetary terms, then— especially if this happened before the fifth century, at a time when telos-membership must have been much more important than it was later—the change represented a major alteration of the political system, for it destroyed the exclusive right of landowners to hold office. If the change occurred after Cleisthenes’ time, and especially after 458/7, it would be less significant, but we should have all the more right to expect some reference to it to survive. Two different attempts have been made to explain the puzzling absence of evidence of that change in the qualifications of the tele¯ which most scholars have taken for granted; but both must be rejected. (a) Some historians, Eduard Meyer for example,188 and Bruno Keil,189 have used this absence of evidence as an argument for the hypothesis that Solon’s classes must have been defined in terms of money income from the first; but this is an impossible position, as I have already shown. (b) At the other extreme, Kahrstedt190 has boldly denied that the qualifications were ever changed from what they were in Solon’s time, as defined by Aristotle, Ath. Pol. 7.4; but Kahrstedt’s whole view depends upon entirely dissociating the tele¯ from the military classes in the fifth and fourth centuries, and this also we have seen to be an untenable position. I can now add another argument against Kahrstedt’s theory: the fact that Aristotle never appeals to the situation in his own day in support of his statements in Ath. Pol. 7.4 about the qualifications of the tele¯, although at least one of 187 As I have shown, Plut. Sol. 23.3 cannot be read as emended by Wilcken and then brought forward as evidence of a later amendment of Solon’s law, any more than it can be held to represent the original law. 188 See pp. 40–1 and n. 151 above. 189 SVAVA 68–70: ‘So folgt, dass die Klasseneinteilung von Solon nicht nach dem Ertrag des Bodens normiert worden ist, sondern dass das ganze Vermo¨gen oder richtiger der Nutzwert des Vermo¨gens der Einteilung zu Grunde gelegt war . . . Wenn erst in spa¨terer Zeit die Umwandlung des Censustarifes aus Produktion-zu ¨ berlieferung wirklich keine Spur von dieser Geldsa¨tzen erfolgt wa¨re, sollte die U einschneidenden, demokratischen Massnahme bewahrt haben?’. 190 SSA 249–55; cf. UMA 20. According to Hignett, HAC 226, ‘this hypothesis is refuted by the facts, which show that Athenian citizens in the second half of the fifth century were still grouped in the property classes to which they belonged.’ I cannot see that this affords any answer to Kahrstedt: see his SSA 252–5.

Were the Qualifications of the Tele¯ Changed?

53

them was admittedly controversial. When Aristotle says e’
dei de tele&in . . . i‘ pp
ada de toyB triak
osia [sc. m
etra t a syn
amwv jZr a ‘ a] poio &yntaB [‘those who produced 300 [sc. measures, dry kai ygr and liquid together] were classed as Hippeis’], why did he not add some such phrase as v
‘ sper kai n &yn [‘as is still the case’], either ‘ d’ e’
nio
i wasi toyB i‘ ppotrowe&in dynam
enoyB before continuing vB [‘or as some people say, those able to rear/keep horses’] or when he ’ ’ came back to give his opinion that oy’ m Zn all’ e ylog
vteron to&iB & m
etroiB difi ZrZsuai kau
aper to &yB pentakosiomed
imnoyB [‘however it is more probable that they were defined by measures, just as in the case of the Pentakosiomedimnoi’]? It is easy, on my theory, to see why Aristotle did not want to mention contemporary ‘ d’ enio
’
i wasi toyB practice: it would have meant saying, vB
‘ sper kai n &yn [‘or as some people say, i‘ ppotrowe&in dynam
enoyB, v those able to rear/keep horses, as is still the case’]; and if you have ’ nothing better than a e ylog
vteron [‘more probable’] to finish up with, it is unwise to make any unnecessary concession to your opponents. Yet another argument against Kahrstedt’s position can perhaps be drawn from certain passages in Aristotle’s Politics; but this I shall relegate to an Endnote.191 The majority of modern scholars have had no explanation at all to offer of the silence of all our sources regarding any change in the qualifications of the census classes; but this very absence of evidence has driven most of them to date the change, which they take for granted, as early as possible,192 and in no case later than about the mid-fifth century.193 It is only Hignett, I think, who is prepared to go as late at that. After first refusing to commit 191

Endnote x (p. 68 below). An interpretation of Arist., Ath. Pol. 13.2, which is certainly over-confident and may be entirely unjustified, and which had led Keil (SVAVA 60 ff., esp. 68–70) to date the supposed classification by money income right back to Solon himself, induced Glotz and Cohen (HG I 443) to imagine a change over to classification by money income before 581. W. Aly, in RE, 2te Reihe, III (1927) col. 972, s.v. ‘Solon (1)’, merely says, ‘Spa¨ter, wahrscheinlich sehr bald, hat man die Zensussa¨tze in Geld umgesetzt’; but it is not clear whether he thinks they were now assessed in terms of income or capital. 193 Bo¨ckh (SA I3 589), it is true, seems to have thought of the Solonian system of assessment in terms of land only as existing right down to the Peloponnesian war, when ‘Vermo¨gensteuern’ were introduced owing to the impossibility of continuing to tax the landowners alone—but Bo¨ckh was confused by his own misconception of the Solonian census classes as above all the basis for an elaborate system of taxation on a sliding scale (cf. p. 57 below). 192

54

Athenian Democratic Origins

himself,194 he eventually seems to come down in favour of a change over to assessment in terms of income in money not later than 462.195 The census classes, however, can never have been assessed in terms of money income,196 and some scholars—De Sanctis, Busolt and Beloch, for example197—who have accepted a change after Solon’s time to a method of assessment in terms of money have at least admitted the necessity for a capitalisation, with an assessment in money capital replacing the Solonian method of assessment in terms of income in kind. This is an improvement in principle, but it will not do. I have shown that Hippeis and Zeugitai were not qualified by ‘measures’ at all, either in Solon’s original constitution or in the fifth and fourth centuries, but that a man’s position as a member of one of these classes was determined by his military status, which in turn depended solely on his ability to perform cavalry or hoplite service. There will never, therefore, have been any necessity for a change in the legal qualifications of these two classes. Only the Pentakosiomedimnoi remain. Although they were the highest class, they must also have been the smallest, and since they are not to be identified with any particular section of the community having social or military significance, but were really little more than the top layer of the Hippeis, and by the fifth century did not have exclusive access to any office of major political importance, the nature of their qualification in later times is of little interest. In the absence of any evidence we can do no more than guess. My own guess is that the qualification of the Pentakosiomedimnoi was never legally altered, but that by degrees, as the economy of Athens became increasingly a monetary one, the fact 194 HAC 143: ‘But though it is usually assumed that at some date between Solon and the fourth century income from movable as well as real property was included in the calculation of a man’s telos, the evidence on which this assumption is based is inconclusive, and even if the alleged innovation is accepted as a fact, it is more likely to have been introduced by Perikles than by Kleisthenes.’ 195 HAC 225–6, ending, ‘It is, however, possible that the older classification, which restricted office to holders of land in Attica, was successfully maintained by the conservatives till 462.’ Cf. 226, where Hignett seems to accept a change by the fifth century to ‘possession of an income of 500 drachmas’ as the qualification of a Pentakosiomedimnos. 196 See pp. 41–5 above. 197 De Sanctis, A2 236–9, also Busolt, GG II2 269 and Busolt-Swoboda, GS II 837 (probably under Cleisthenes); Beloch, GG II2 i 89 and n. 2 (not later than the period of the Persian wars).

Were the Qualifications of the Tele¯ Changed?

55

that the 500 ‘measures’ had originally been a ‘barley standard’ was gradually forgotten, and medimnoi and metretai of all kinds came to be equated for this purpose, so that the Pentakosiome’ tZB & dimnos was now a man who, as Aristotle says, enjoyed ek ‘ a [‘from oi’ ke
iaB . . . pentak
osia m
etra t a syn
amwv jZr a kai ygr
his own land, . . . 500 measures, dry and liquid together’]. There can never have been any register of any of the four classes as such, and a man would simply announce his status as a member of the appropriate class on the rare occasions when he was applying for an office open only to one or more of the higher classes, or offering himself as surety, or giving a dowry to an epikleros as next of kin, or doing some other act to which his census class was relevant. Except in so far as he was, or had been, a cavalryman or a hoplite, or, though able-bodied, had never performed either service, no man would be, so to speak, labelled permanently and conspicuously as Pentakosiomedimnos, Hippeus, Zeugites or Thes: there was certainly no list on which he had to be so described. There is no difficulty in believing that the Pentakosiomedimnoi, and they alone, were always qualified in terms of landed produce. It was secure and available wealth, in the best possible form, namely considerable landed property, that was required of them, and for very good reasons. A Hippeus belonging to a noble family, who felt that his distinguished ancestry and his social position made it imperative that he be a breeder of horses, might well devote a large part of his land and of his substance to that end, and not leave himself very much over in the way of available assets. He could fill the most important offices of state, even eponymous archon, polemarch or archon basileus, but he would not be a very suitable treasurer of Athena, because if he converted the valuables of the goddess to his own use, there might be no easily realisable assets of his to seize and confiscate. Hence the necessity for aspiring treasurers to have landed property producing a large volume of crops annually: this kind of asset alone was the best form of security. It has not, I think, been realised that the very existence of a census class higher than that called the ‘Hippeis’ is a curious phenomenon which requires explanation; and there is no other explanation that I can see. Of course, one or two other offices, that of the Kolakretai for example, may have been limited to the Pentakosiomedimnoi, but we have evidence only for the treasurers

56

Athenian Democratic Origins

of Athena; and if there were any other such offices, I am convinced they will have been financial in character.198 The archonship, I believe, was probably open to the Hippeis from the first.199 If iteration of archonships was altogether excluded—so that being a Thesmothetes disqualified a man from ever being archon eponymous, for instance—it seems to me that there could hardly have been enough Pentakosiomedimnoi to fill all these posts, in addition to the treasurerships. And even if the various archonships were separately treated, so that a man could be eponymous archon and also Thesmothetes, or even polemarch, I do not think the Athenians would have wished to limit their highest political offices (as distinct from the purely financial ones) to a class narrower even than the Hippeis—they would have been depriving themselves quite unnecessarily of some of the potential candidates best qualified by upbringing and family connexions to provide the kind of leadership that was needed in that aristocratic age. If the stories we are told of Solon’s modest wealth are true,200 he is unlikely to have been a Pentakosiomedimnos; but I should be surprised if he was not a Hippeus.201

(VIII) POLLUX VIII 130 I have not yet dealt with the statement of Pollux VIII 130 (cf. Schol. Plat., Rep. VIII 550C) that (a) the Pentakosiomedimnoi ’ Zliskon ei’ B t an
o dZm
osion t
alanton [‘paid a talent into the demosion (public treasury)’], (b) oi‘ de t Zn i‘ pp
ada telo &ynteB [‘and those assessed as Hippeis’] half a talent, (c) oi‘ de t o zeyg
Zsion telo &ynteB [‘and those assessed as Zeugites’] ten minae, (d) oi‘ de t o uZtik on [sc. telo &ynteB] [‘and those [sc. assessed as] Thetes’] nothing.202 It is not immediately clear what this is sup198 It is a pity that we have no information about any qualifying status for Hellenotamiai. 199 I realise that by the late fourth century it could be said even by a well-informed Athenian, Demetrius of Phaleron (FGrH II B 228 F 43, ap. Plut., Arist. 1.2), that the archonship had once been open to the Pentakosiomedimnoi alone. But Demetrius’ statement is untrustworthy and inaccurate anyway: it is made in relation to Aristeides, whom he conceived as chosen by lot, after 479 (loc. cit., also F 44, ap. Plut., Arist. 5.9–10), whereas in fact Aristeides was elected, and his year was 489/8—I imagine Demetrius was writing from memory and confusing the year after Marathon with the year after Plataea. 200 [See Arist., Ath. Pol. 5.3, with Rhodes, CAAP 123.] 201 Especially if fr. 23 (West) is really his. 202 For the full text, see n. 104 above.

Pollux VIII 130

57

posed to mean: presumably the liability specified is that of each member of the telos concerned individually, but on what occasions and how often did he make the payment? In any event, the statement cannot be accepted in any literal sense. I suggest that we must either deny all authority to this text (as Lipsius and others have done, and as I did in my discussion of the fourth century eisphora)203 and suppose it to be due to some invention or misconception, or else, as I now incline to think, take it to be a misunderstanding by Pollux or his source of a genuine rule that when an eisphora was levied, all Pentakosiomedimnoi were to pay equally, at the rate fixed for the occasion (probably 1% would be the most likely),204 on a theoretical assessment of one talent, all Hippeis equally on half a talent and all Zeugitai on ten minae.205 I still believe, as I did in 1953, that it would be very wrong to adopt any form of Bo¨ckh’s theory,206 that the enigmatic statements of Pollux indicate the existence of a sliding scale of taxation, with the members of the three highest tele¯ paying at three different rates, but according to their individual assessments.207 The whole conception is much too advanced and intricate for sixth-century or even fifth/fourth-century Athens, and nowhere in the ancient world do we ever find direct taxation imposed on a detailed progressive scale in anything like the manner required by Bo¨ckh’s theory. I would now accept the passage in Pollux as probable evidence that the census classes were intended to be used as the basis of assessment for direct taxation, in such a way that the three top classes paid at the same rate (whatever it was on each occasion) but on a notional 203

204 DTAEFC 42–43. Cf. my DTAEFC 34, 35, 49–50. I wonder whether something like this was in the mind of Ed. Meyer when he said (GdA III2 607) ‘Spa¨ter scheinen die Einkommensa¨tze der einzelnen Klassen kapitalisiert und dabei herabgesetzt zu sein; eine durch ihre knappe Fassung fast unversta¨ndliche Notiz (Pollux VIII 130) scheint zu besagen, dass die Pentakosiomedimnen 1 Talent, die Ritter 3000, die Zeugiten 1000 Drachmen (¼ 10 Minen) versteuerten’. (The first part of this sentence, down to the semi-colon, can of course be ignored: it derives from Meyer’s erroneous theory that the Solonian classes must have been qualified from the first in terms of money income: cf. pp. 36 ff. above.) Cf. 608 n. 1; ‘Auch ist es ja unzweifelhaft, dass die t
elZ zugleich Steuerklassen waren’—a statement which is preceded by an explicit rejection of Bo¨ckh’s theory on the point. 206 SA I3 571–628. 207 This is what it comes to in practice, although Bo¨ckh conceived the assessment itself as varying according to each man’s class, so that the higher the class, the greater the proportion of his time¯ma on which a man paid tax. 205

58

Athenian Democratic Origins

assessment which was different for each class, so that if the rate of the levy was 1%, for example, each Pentakosiomedimnos would pay 60 drachmae, each Hippeus 30 drachmae and each Zeugites 10 drachmae. Whether tax was ever in fact levied according to this method is another matter: there is no evidence at all, and we cannot be sure that the system was actually put into operation. But at least, unlike the complexities which Bo¨ckh’s theory involves, such a system was an admirably simple one, well suited to the political conditions of early Athens, where there was nothing in the way of a civil service to administer an involved taxation scheme; and it would have had the great advantage of enabling the Athenians to know in advance roughly how much a given percentage levy would raise, provided only they knew the approximate numbers in each class. I would of course take it for granted that the Athenians never had any intention of levying eisphora regularly, but only in an emergency— almost certainly in wartime alone, as in the late fifth and fourth centuries.208 It is not possible to say with any confidence when this scheme of taxation was introduced—or, for that matter, when it was abandoned. It might conceivably have been Solon’s work: even before coined money came into common use, such a form of taxation, with payments in silver by weight, is not impossible, and a striking fiscal innovation such as the first introduction of a form of direct taxation is perhaps not uncharacteristic of that radical reformer, who was not afraid to cancel debts, forbid debt-slavery for the future, and make access to political office dependent upon property assessments. I suppose that a date in the second quarter of the sixth century, shortly before the tyranny of Peisistratus, is just possible, but during that disturbed period the introduction of such a major innovation seems to me unlikely. The Peisistratids introduced regular direct taxation of landed produce in kind,209 but this would certainly not have continued after their expulsion in 510. To my mind, perhaps the most likely time for the invention of the system of taxation I believe to lie behind the Pollux passage is 508/7, when Cleisthenes’ constitution was promulgated. I cannot believe, however, that this very crude system remained in use for very long 208

Cf. my DTAEFC. Thuc. VI 54.5; Arist., Ath. Pol. 16.4, 6; and the other sources cited by Sandys, ACA2 66 n. 209

Pollux VIII 130

59

during the fifth century. We have no details at all of the method of levying eisphora at Athens before the introduction of the symmory system in 378/7,210 but we know that a levy took place in the latter part of 428,211 and that there were at any rate two others during the Ionian war,212 and epigraphic evidence213 proves that an eisphora was at least a possibility which could be seriously contemplated in the third quarter of the fifth century.214 I think we must suppose that the primitive arrangements reflected in Poll. VIII 130 were replaced during the fifth century by a more sophisticated and more profitable215 method in which every man rich enough to be subject to eisphora was liable to pay tax (at whatever rate was decreed at each levy) upon the total value of his property, his time¯ma, as assessed by himself.216 The existence of the tribute may have made actual levies unnecessary until 428.217 210

Cf. my DTAEFC 44–5. Thuc. III 19.1. Cf. my DTAEFC 34 n. 20. 212 In my opinion, Lys. XXI 3 makes it likely that two eisphorae, and not more than two, were exacted between 411/10 (ibid. § 1) and the end of the war in 404; and I find it hard to believe that there were any levies between 425 (when the assessment of tribute due from the Empire was roughly trebled) and the Sicilian disaster in 413. Here I am in general agreement with K. J. Dover (in CQ 44 (1950) 44–60, at pp. 58–9) against Gomme, HCT II 278–9. Contrast, however, the references to pollai ei’ swora
i [‘many eisphorae’] in Lys. XII 20 and XXV 12, and to pollai kai meg
alai ei’ swora
i [‘many major eisphorae’] in Ant. II ii 12; and cf. Ps.-Lys. XX 23; Lys. VII 31; XVIII 7, 21; XIX 29, 57; XXX 26. 213 M/L 58 ¼ Fornara 119 (¼ IG I3 52 B) 17, 19; IG I3 41 c 38; IG I3 21 ¼ Fornara 92 h 56. 214 I think this is all we need infer from the evidence cited in n. 213 above. If an eisphora was actually levied before 428, then either Thucydides (III 19.1) is wrong or, against the natural interpretation of his words, we must suppose that when he ’ ’ speaks of the Athenians in 428 as esenegk
onteB t
ote pr vton & eswor an diak
osia t
alanta [‘contributing then for the first time an eisphora of 200 talents’] he means either (a) that this was the first eisphora during the war (cf. Gomme, HCT II 278), or (b) that this was the first eisphora which was as large as 200 talents. [The case for (b) is argued by J. G. Griffith, AJAH 2, 1977, 3–7.] 215 The two payments of eisphora made by the very rich speaker of Lys. XXI (see § 5) were of 3,000 and 4,000 dr. respectively. If each payment was not (as is quite possible) larger than the amount compulsorily due, and was made in virtue of a 1% eisphora (on a time¯ma of 50 tal. in the first case and 66 2/3 in the second), then the man was being made liable for over a talent (6,880 dr.) more than he would have had to pay in two levies at 1% under the old system, as I have conceived it. (I give these figures exempli gratia.) 216 See my DTAEFC 33–4. 217 In DTAEFC 45 I have remarked that Ar., Eq. 923–6, does not necessarily imply that the rich then paid at a higher rate. I would add that this passage is consistent with the use of either of the two systems of assessment contemplated above. I now feel certain that Ps.-Arist., Oec. II ii 5, 1347a 18–24 (cf. DTAEFC 44), 211

60

Athenian Democratic Origins

As I have indicated above, however, all such attempts to date the introduction or the abandonment of the system of taxation I have supposed to lie behind Poll. VIII 130 are conjectural.

(IX) MISCELLANEOUS 1. The earliest evidence known to me of the existence of a detailed time¯ma of each man’s property comes from the late 390s: in Isocr. XVII 49218 it is said of a man named Cittus that Pasion ‘returned him in his assessment(s)219 as a slave among his other servants’ ’ on de to &yton apegr
’ ’ to&iB tim
‘ do &ylon (t on ayt acato . . . en Zmasin vB ’
met a tvn & oi’ ketvn & tvn & allvn). No evidence is given in support of these statements, and my own impression is that Pasion probably insisted all along that Cittus was really a free man, probably his own freedman; but the words quoted, even if they are false, provide unimpeachable evidence that slaves could be (and therefore, presumably, had to be) included in a man’s time¯ma—and if slaves, then very probably (if not quite certainly) all other movable property, as well as land. However full of misrepresentations the speech may be, I do not see how the conclusion I have stated can be avoided. Unfortunately, it is not certain whether Pasion at the time in question was still a freedman, i.e. a species of metic, or had already become a citizen. Pasion died a citizen in 370/69,220 and it seems to me unlikely that he was a citizen as early as the 390s.221 If he was not then a citizen, his time¯ma cannot have had anything to do with registration in connexion with the census classes, which of course were relevant to citizens alone. And in any event, if my picture of the qualifications of the tele¯ in the fifth and fourth centuries is right, no special return of property would ever be necessary for the relates to the fourth-century cleruchy at Potidaea—which may have enjoyed a greater measure of autonomy than is usually attributed nowadays to fifth/fourthcentury cleruchies: see CQ 56 ¼ n.s. 13 (1963) 110–19, at p. 111 n. 1. 218 The outside dates for the speech seem to be 393 and 387. The events described in it will have occurred during the late 390s. 219 Plural for singular, I suspect. 220 Ps.-Dem. XLVI 13. 221 [vacat in Ste. Croix. This is surely correct: the decisive passage is Isoc. XVII ‘ yp ‘ er Pas
ivnoB a‘
panta kai l
egei kai pr
33.1, Pyu
odvron . . . oB attei, i.e. at the time of the speech Pasion still required a citizen prost
atZB: cf. e.g. Davies, APF 429–30. PJT]

Miscellaneous

61

purposes of the tele¯. In either case, Pasion’s time¯ma will have been that which he returned in order that his liability for eisphora might be assessed—we know that metics (probably only the richer metics) were indeed liable for eisphora by the early fourth century.222 We may for convenience say that there is evidence in Isocr. XVII 49 for what may be called ‘fiscal-time¯ma’ only, as distinct from ‘political-time¯ma’, any assessment required for deciding to which of the four tele¯ a citizen belonged. (I of course believe that in reality there was no such thing as ‘political-time¯ma’, no registration or return being necessary to establish a man’s telos.)223 2. The distinction between ‘fiscal-time¯ma’ and ‘politicaltime¯ma’ becomes important in connexion with Isaeus VII 39, which has been claimed by A. H. M. Jones as proof that ‘the Solonian classes were by the fourth century based on property, not income’.224 I cannot accept Jones’s inference from this passage. I have just shown that fiscal-time¯ma and political-time¯ma could be two quite different things; and what Pronapes (in Isae. VII 39) is accused of having assessed at a low figure is his fiscaltime¯ma. Indeed, the general theory I have offered fits the passage much better, for if telos-membership was not regulated according to a quantitative time¯ma at all, Pronapes’ action in giving himself a low fiscal-time¯ma and yet claiming to be qualified for offices ‘as if he were a Hippeus’ would be much more easily accomplished. (Not that I would pay much attention to unsupported oratorical allegations of this kind.) 3. Nor can I accept Jones’s suggestion225 that in at any rate the fourth century there was a qualification for zeugite/hoplite status of 2,000 drachmae, the minimum amount of capital demanded by 222

See Isocr. XVII 41; also Lys. XXII 13, with 5 (c. 385 B.C.). In case anyone still hankers after the view that citizens had to make returns of their property (‘political-time¯mata’), so that each man’s telos might be decided, I would remark that the fact (which seems certain from Isocr. XVII 49) that movable property (slaves anyway) had to be included in one’s fiscal-time¯ma affords no evidence whether property other than land would have had to be included in political-time¯mata, had such things existed. Registration for eisphora (‘fiscaltime¯ma’) may in any case have been something quite different from registration for the census classes (‘political-time¯ma’), and the fact that slaves formed part of the capital assessment on which eisphora was paid proves nothing whatever about the basis or assessment for the tele¯, even if such a thing existed. 224 AD 142 n. 50. By ‘property’, of course, capital is meant. 225 AD 142 n. 50; cf. 76 and 149 n. 3. 223

62

Athenian Democratic Origins

Antipater’s constitution in 322, at the end of the Lamian war, for enjoyment of full citizen rights.226 It is a priori very likely that Antipater would wish political rights to be confined to roughly the hoplite class; but of course 2,000 drachmae need not already have been the qualification, and on the interpretation of the evidence about tele¯ and military classes adopted in this essay it cannot have been. I suggest that the figure of 2,000 drachmae was simply Antipater’s attempt to fix a definite limit which would result in a body of full citizens consisting approximately of the o
‘ pla parex
omenoi [‘those providing (their own) arms’]—as it apparently did, for such hoplite figures as we have after the great plague at Athens227 are reasonably consistent with a total figure for citizen hoplites of around 9,000, which was also the size of Antipater’s pol
iteyma, according to both Diodorus and Plutarch.228 It is interesting—and it supports my general thesis—that we should hear of a fixed quantitative time¯ma only at the precise moment when, for the first time at Athens, a qualification was introduced for the exercise of basic political rights. For this purpose, in an oligarchy, a precise monetary time¯ma was natural and necessary: it was something quite different from fixing the lower limit of a hoplite class, which the state naturally wanted to be as large as possible. It is evident from many statements in Aristotle’s Politics229 that such quantitative time¯mata existed widely in the fourth century. I may mention that the language used in these and other passages230 suggests that most states allowed not only 226

Diod. XVIII 18.4–5. The same figure, 2,000 dr. (Alexandrian drachmae this time) appears at Cyrene as the qualification for the pol
iteyma, the body of citizens possessing full political rights, in the constitution set up under Ptolemy’s aegis at the end of the fourth century: SEG IX 1 § 1. 227 Thuc. IV 90.1, with 93.3 & 94.1 (B.C. 424); Ps.-Lys. XX 13 (411); Xen., Hell. IV ii 17 (394); Diod. XV 84.2 (362); XVI 37.3 (352); and esp. XVIII 10.2 & 11.3 (323–2): see Jones, AD 30–31 and 142 n. 42; 81 and 150 nn. 25–6. The figures for the Lamian war of 323–2 suggest, as Jones says, a grand total not far off 9,000 citizen hoplites—if we remember that some metics also served as hoplites (see n. 186 above). 228 Diod. XVIII 18.5 (22,000 disfranchised); Plut., Phoc. 28. 7 (12,000 disfranchised, which I would prefer). Cf. Jac. ii 377 n. 16, apparently equating (a) Zeugitai, (b) o‘
pla parex
omenoi, and (c) Antipater’s 9,000. 229 e.g. 1266b21–4; 1274a15–21; 1278a21–3; 1282a29–32; 1283a17–19; 1291a 39–41; 1292a39–b2; 1292b25–30; 1294b2–13; 1297b1–3; 1298a35–9; 1303a21–5; 1305b30–4; 1306b6–16; 1307a27–9; 1308a35–b10; 1317b22–3; 1318b27–32; 1320b 21–8; 1321a26–8. Cf. Plato, Rep. VIII 550C, 553A. 230 e.g. Hell. Oxy. XI 2.

Endnotes

63

land but also other forms of property to be included in a man’s time¯ma. 4. Jones231 says of Antipater’s 2,000 drachmae, ‘The figure was probably reached by multiplying by ten the Solonian income of 200 metra . . . , converted on the Solonian scale of values (Plut., Sol. 23) into 200 drachmae; this figure is slightly confirmed by the & value of the Lesbian klZroi [‘‘allotments’’] (Thuc. III 50.2), which were probably intended to raise thetic occupants to zeugite status.’ As I do not believe that any official qualification can ever have been expressed in terms of money income,232 I of course do not believe that the annual income of 200 dr. received by each of the Lesbian cleruchs of 427 can have turned him into a hoplite per se,233 although if (as I think) hoplites became such by the ability to perform hoplite service, the possession of an income of 200 dr. in addition to whatever else the cleruch had might very well make it possible for him to achieve hoplite status.

ENDNOTES i (from n. 6). The eupatrid monopoly of the archonship Later (in A 247–8 n. 49), Jacoby argued cogently that Plut., Thes. 25.1–2 (forming the earlier part of Ath. Pol. fr. 2 in Sandys’s edition and the OCT, ed. Kenyon, but omitted from H. Oppermann’s Teubner text, fr. 4), the one major source to assert explicitly an original ’ eupatrid monopoly of the archonship, in the words E ypatr
idaiB ’
’ de . . . par
exein arxontaB apodo yB [o‘ QZse
yB] ktl. [‘(Theseus) entrusted to the eupatrids . . . the provision of archons’], probably does not come from Aristotle himself but from ‘an author who, historically of no importance whatever, described the State of Theseus obviously under the influence of the Platonic State’: Jacoby suggests Theophrastus, but the source could be an Atthidographer or a political pamphleteer. Cf. F. R. Wu¨st, in Historia 6 (1957), at pp. 181–2. Ruschenbusch, PP, at p. 415 n. 64, thinks the passage in Plutarch goes back to Androtion, but without giving reasons. 231

232 AD 142 n. 50. See pp. 41 ff. above. As believed by Jones, loc. cit.; Glotz-Cohen, HG II 201; and others. I cannot go into the figures here, but I believe it can be proved beyond reasonable doubt that an income of 200 dr. per year was very far below what was needed to maintain hoplite status even in the fifth century, and still more in the fourth. 233

64

Athenian Democratic Origins

We certainly cannot claim with any confidence the authority of Aristotle for the view that in the aristocratic state only eupatrids could be archons. Nevertheless, it very probably represents what Aristotle believed (I see no reason why he should not have continued to hold it for the period before Draco even after ch. 4 was inserted into the Ath. Pol.), and it appears in later sources: Dion. Hal., Ant. Rom. II 8.2; Euseb., Chron. II (ed. A Schoene, 1866) 84, s.a.Abr. 1333 ¼ Olymp. XXIV 1 [Hieron.] or 2 [Arm.] (see esp. Sync., p. 399.21); cf. Plato, Critias 110CD. [The reality of a ‘Eupatrid order’ is doubted e.g. by Welwei, Athen, 108–9; cautious acceptance in J. K. Davies, OCD3, s.v. Eupatridai; Parker, AR, 63–5.] ii (from n. 12). Arist., Ath. Pol. 7.4 and 55.3 There is an obvious discrepancy between the forms of the question put to the candidate for office in Arist., Ath. Pol. 7.4 (line 16 OCT) and 55.3 (line 14 OCT): in the first passage it is in effect ‘Which of the tele¯ do you belong to?’ and in the second ‘Do you pay your taxes?’. Connected with the latter are the ei’ t a t
elZ tele&i [‘does he pay his taxes?’] in Deinarch. II (c. Arist.) 17 and the t a t
elZ telv& [‘I pay my taxes’] of Cratin. Jr. fr. 9 K/A ¼ 9 K, line 5, ap. Athen. XI 460F. The solution is that there were two different questions, asked on two different occasions, the first (‘What is your telos?’) at the preliminary ’ akrisiB, when the candidate was still m
ellonta klZro &ysuai [‘a an
candidate for allotment’] (Ath. Pol. 7.4), and the second (’Do you pay your taxes?’) at the dokimas
ia subsequent (Ath. Pol. 55.2–3) to the election. Poll. VIII 85–6 is confused: only the last item in his list, ’
’ iB [untranslatable: read kai t
i t ayto& o t
imZm
a kai ei’ t o t
imZma estin ’ estin, ‘what their census class is?’ PJT], belongs to the anakrisis to which he attributes it; all the other questions (see Ath. Pol. 55.3) were asked at the later dokimasia. [So too Rhodes, CAAP 145.] iii (from n. 45). Antiquity of Athenian Hippeis In Aristotle’s account of the census classes, in both Pol. II 1274a5–21 ‘ i‘ pp
and Ath. Pol. 7.4, the Hippeis are Z aB, a feminine adjectival form (presumably with t
ajiB understood) used as a noun, and found in classical Attic only in survivals, such as (a) the name of the census class of Hippeis; (b) the ‘Ipp
adeB p
ylai [‘Hippad gate’], one of the city gates (Heliodorus ap. Plut., Mor. 849C); (c) bo &yB i‘ pp
aB [‘hippad ox’], explained by Hesych., s.v. i‘ pp
aB, as used for a sacrifice in the uys
iai i‘ pp
adeB [‘hippad sacrifices’], in which the Hippeis were con‘ i‘ pp cerned; and (d) Z aB paidi
a [‘the hippad game’], a children’s game

Endnotes

65

mentioned by Poll. IX 122. I am grateful to Miss L. H. Jeffery for drawing my attention to this. iv (from n. 47). Early mounted infantry Miss Jeffery has kindly drawn my attention to the predella of the hoplite stele in the Metropolitan Museum, New York (G. M. A. Richter, The Archaic Gravestones of Attica, no. 45 and fig. 126, c. ‘ ioxoB [‘charioteer’] and 530 B.C.), showing a chariot with Zn
’ apob
atZB [‘dismounter’]: this would demonstrate the man’s wealth and high social rank among the horse-breeders, but evidently he served as a hoplite. As Miss Jeffery remarks, he may well have been one of those hoplites who rode to the battlefield with their squires and then dismounted; cf. J. K. Anderson, Ancient Greek Horsemanship, 146 and pl. 29; and see chs. xi–xii, pp. 128–54, esp. 129, for a good account of the part played by Athenian mounted warriors in the sixth century as illustrated by vase-paintings. This monument, the stelecapital of c. 550–30 from Lamptrae (Richter, op. cit. no.20 and fig. 68: the horseman must be the squire, who carries two shields and a spear—surely not a mere hunting-javelin), and the stele of c. 525 in the Museo Barracco at Rome (Richter, op. cit. no.64 and fig. 154: note the sword as well as the two spears), all seem to testify to specifically military activities, and several other monuments of the same sort which I need not mention here may well have a military significance. v (from n. 54). Katalogoi Jones, AD 163, gives a brief and clear explanation of the mechanism of the call-up. I might add that the expression kat
alogoB occurs in this sense (call-up for a particular expedition) in the singular as well as the plural (see Ar., Eq. 1369). Lys. XIV 6–7 may appear at first sight to be a reference to the compilation of the main hoplite katalogos, but the question asked is, ‘Who are those who must appear in the infantry ’ tfi Z& pezfi Z& stratia&fi )?’, and the answer, ‘Are they not those who ranks (en ’ to&iB epvn
’ are of the proper age [a reference to the call-up en ymoiB [‘by age-groups’]: see Jones, loc. cit.], those whom the generals enrol?’, must therefore refer to the enrolment by the generals for the particular expedition. (Registration in the main hoplite katalogos itself was the business of the tribal taxiarch: see Lys. XV 5; cf. XIII 79.) The speaker then seems to say (in a passage the text of which is not very ’ satisfactory) that Alcibiades is guilty of astrate
ia [‘avoidance of service’] because though enrolled he did not march out.

66

Athenian Democratic Origins

vi (from n. 61). Dem. XXIV 197 ’ ‘ Jones, AD 84, translates u
yraB awaire& in kai str
vmau ’ yposp &an kai ’ Zto, ’ & di
akonon, e’
i tiB exr ta
ytZn enexyr
azein by ‘removing doors, and seizing blankets, and distraining on a servant girl, if anyone employed one’, and adds, ‘If Demosthenes is speaking the truth, some of the 6,000 eisphora payers could not even afford to buy a single slave girl to help in the house.’ There are three arguments against this interpretation: 1. This is part of an orator’s case. The speech is full of tendentious statements, and even if Demosthenes were pretending that some of Timocrates’ victims might be so poor as to possess no slave, we should not be justified in assuming that perhaps some of them actually did not. 2. The e’
i tiB clause will not bear the interpretation Jones gives it. There is certainly no necessary implication that there might really be people among Timocrates’ victims who did not have a di
akonoB at all. For uses of e’
i tiB in which no corresponding doubt is present, see e.g. Thuc. VII 10; VIII 76.2; Dem. XIV 16, and some texts cited by R. Ku¨hner and B. Gerth, Ausf. Gramm. d. griech. Sprache, II3 ii 573–4 (in support of the statement, ‘so kommt es, dass e’
i tiB . . . u.s.w. geradezu statt tiB . . . u.s.w. stehen’), e.g. Thuc. I 14.3; 17; IV 26.5; VII 21.5; Xen., Hell. IV ii 21; V iii 3. ’ Zto ’ & 3. Here exr [‘employed’] stands for eplZs
iaze [‘had intercourse with’]: this was realised by Wm. Wayte, in his edition of Dem. XXII and XXIV, ad loc. (Note the holding back of the gender of the slave until the last two words.) The sense is, ‘Whenever he came across a favourite servant, why, he took her off too’. And see the Scholia (ed. Dindorf, IX 801–2). According to Wayte, ‘the MS. authority is de’ Zto & cisive in favour of fi Zffl tiB exr [printed by Wayte, following Benseler], which yields exactly the same sense as e’
i tiB.’ vi i (from n. 115). Relative prices of wine and barley The only comparable simultaneous figures we seem to have which are at all reliable are from the Eleusis accounts of 329/8 where 16 dr. were apparently spent on 2 metretai of wine (at 8 dr. per metr., therefore) for the public slaves (hardly the best wine, therefore) at the festival of the Choes (IG II2 1672. 204–5). In the same year the barley first-fruits ’ (aparxa
i) were sold at 3 (to 3 5/6) dr. per medimnos, and the wheat ’ aparxa
i at 5–6 dr. per med. (ibid. 281–3, 297–8). Phaenippus in the 320s was alleged (Ps.-Dem. XLII 20) to be producing more than 1,000 med. of corn (in fact barley, as the context shows) and 800 metr. of wine on his estate, and selling the barley at 18 dr. per med. and the wine at 12 dr. (apparently per metr.), but we cannot trust these figures at all—cf. my contribution to ASI

Endnotes

67

(Ehrenberg), 109–14. The whole speech gives every sign of being a thoroughly dishonest one, and from § 24 it appears that Phaenippus admitted to producing only a tenth as much as the speaker alleges. In § ’ 31 Phaenippus is said to have been obtaining a triplas
ia tim Z Z pr
oteron [‘price three times as great as before’]. Although the barley figure of 18 dr. may well be a gross exaggeration, the price Phaenippus had been getting may well have been treble that of a few years earlier, for a nearly contemporary speech (Ps.-Dem. XXXIV 39) mentions that in a year of high prices, probably 330, s&itoB (which in the context means wheat) rose from its usual price (kauestZky&ia tim
Z) of 5 dr. per med. to 16 dr.; and an inscription (IG II2 408. 12–14), also of about the same date, honours two Heracleotes for selling at Athens (evidently in a year of severe shortage) Sicilian wheat at 9 dr. per med. and barley at 5 dr. For other material on wheat, barley, wine and oil prices, see Jarde´, CAG I 164–86; W. K. Pritchett in Hesp. 25 (1956), at pp. 184–6, 196–203; F. Heichelheim, Wirtschaftliche Schwankungen (1930), 51–2, 57–72, 118–22, 128–35. [For more recent discussion, see P. Garnsey, ‘Grain for Athens’, in Crux, updated in his Cities, Peasants and Food in Classical Antiquity (1998) 183–200.] viii (from n. 130). Attic wheat It is possible that the proportion of barley to wheat was of the order of twelve to one, but it looks as if in the year 329 at any rate the quantity of wheat grown in Attica, excluding Oropus, was c. 27,500 med. or more, and one cannot call this ‘practically no wheat’. The figures are of course obtained from IG II2 1672: see the interpretation of Jarde´, CAG I 36–57, 95–98. I am taking into account only the returns of the ten Athenian tribes, which, if we make all the necessary assumptions, give us 27,487 1/2 med. of wheat and 340,575 of barley, or 371,062 1/2 med. in all. (I should perhaps mention that anyone who uses the table on p. 41 of Jarde´’s book should beware of a misprint: the total production of the tribe Akamantis should be 44,700, not 47,700.) The most important of several ‘necessary assumptions’ is that the ’ respective rates of aparx
Z [‘first-fruits’] were still 1/1200 for wheat and 1/600 for barley, as laid down by M/L 73 ¼ Fornara 140 and amended by IG II2 140. ix (from n. 168). Van Groningen’s version of Wilcken’s interpretation of Plut., Sol. 23.3 ’ vn Van Groningen, who accepts oysi & but believes that movables were not assessable in Solon’s time, has suggested that ‘medimnos, sheep and

68

Athenian Democratic Origins

drachma’ were intended to correspond with the three main ways in which land might be utilised: for agriculture, for pasturage, or for letting. This seems to me over-ingenious, and I would criticise it as follows: 1. The form of words used by Plutarch (ei’ B . . . t a tim
Zmata tvn & ’ vn ’ i med
imnoy) [‘for the oysi & log
izetai pr
obaton kai draxm Zn ant purpose of fixing the assessments of property, (the Solonian law) reckons a sheep (or goat) and a drachma instead of a medimnos’], although on van Groningen’s interpretation it puts the right emphasis on med
imnoy, would be an exceedingly strange way of saying what van Groningen wants the passage to mean, and no one else has ever understood it in that way. The responsibility for the cryptic wording would presumably not be Plutarch’s. But why should Solon express himself with such unnecessary obscurity? 2. If in Solon’s time a medimnos of barley was really worth a drachma, then a piece of land let for an annual rent of so many drachmae must have been very much more valuable than a piece of land merely producing as many medimnoi of barley, for the lessee would have to maintain himself and his family in addition to paying the rent. Land which was leased would therefore be seriously undervalued. 3. It seems unlikely that letting for ‘money’ rents (even in silver by weight) was much practised in Solon’s day. 4. It seems unlikely, too, that at this time more than a small fraction of the land of Attica was used for pasturing sheep, or even sheep and goats. 5. Assessing pasture land in the way suggested by van Groningen would surely be an absurd method of valuation. A man could have counted his sheep, for this purpose, in one of two ways. Either (a) he included all his sheep every year, or (b) he included only those newly acquired, by birth (and perhaps purchase?), in the year of assessment, or in an average year. But if (a), then capital units would have been equated, most inappropriately, with units of income, as even a sixth century Greek could hardly have failed to realise. And if (b), then the undervaluation of a sheep (see p. 40 and n. 149 above), as compared with a drachma and a medimnos of barley, would result in an unfairly low assessment of the land. By the method of assessment of pasture land which I have suggested above, on the other hand, no such unfairness would arise. x (from n. 191). A further objection to Kahrstedt’s theory of the qualifications of the tele¯ Against any theory that the qualifications of the census classes were always assessed in terms of landed produce reckoned in kind I see another objection which I do not think has been noticed by Kahrstedt and those who like him accept Aristotle’s account and deny a subse-

Endnotes

69

quent change in the method of assessment. This objection derives from the most intelligent discussion by Aristotle, in two places in the Politics (V 1306b6–16; 1308a35–b10), of the consequences of economic changes resulting in a larger proportion of the citizens (in 1306b13–14 Aristotle actually says ‘all’ citizens) rising above the fixed monetary time¯ma which is the qualification for important political rights. Aristotle is speaking not only of oligarchies but also of polite&iai in the special sense in which he sometimes uses that term, i.e. mixed consti’ tutions; and the phrase he uses at the beginning of the first passage, en ’ o tim
’
’ aB o‘
saiB ap ZmatoB boyle
yoysi kai dik
azoysi kai t aB allaB arx ’
arxoysin [‘(oligarchies) in which people serve as councillors and jurors and fill the other magistracies on the basis of their time¯ma’], would apply, but for the word dik
azoysi [‘serve as jurors’], to the constitution of Solon as Aristotle himself pictures it both in the Ath. Pol. and in the Politics (see esp. II 1273b35–74a21; III 1281b25–38). In the second passage explicitly (note nom
ismatoB [‘money’] in 1308a38), and in the first as surely, by implication, the time¯mata are conceived in terms of money capital only—note Pol. 1306b12–13: it is not just that many citizens have greater possessions, but that the same possessions, ’ aB kt
’ iaB t aB ayt ZseiB, become pollaplas
ioy . . . tim
ZmatoB aj
[‘worth many times the time¯ma’]. The only remedy Aristotle can suggest (Pol. 1307a8–b6) is to review the assessments frequently, and raise or lower the qualifications if necessary. But if in a very conspicuous case, Athens, the alleged Solonian system of assessments in kind, in terms of landed produce, had still existed, why did Aristotle not refer to it? It would have provided a secure method of preventing the very sort of metabol
Z [‘(constitutional) change’] which Aristotle is concerned about in these passages! The situation he is thinking of, namely a fall in the value of money, would not have affected the distribution of citizens on what Aristotle represents as the Solonian basis, even if there was a certain tendency for people to rise into higher classes simply because they became absolutely richer. I realise very well that in the two passages from the Politics which I have brought forward here we are dealing not with qualifications for military service, or gradations within a democratic system (such as the Solonian tele¯) giving access to offices, but with the basic qualification for exercising political rights at all—comparable, as far as Athens is concerned, only with the time¯ma of 2,000 dr. established by Antipater, to which I have already alluded. The principle, however, is exactly the same. If the situation at Athens had been as Kahrstedt pictures it, Aristotle could perfectly well have appealed to it as an alternative to the kind of time¯ma he is discussing, an alternative not subject to the defects of the usual method which he emphasises.

70

Athenian Democratic Origins

APPENDIX The Dedication of Anthemion (Arist., Ath. Pol. 7.4) Miss L. H. Jeffery, with characteristic kindness, has communicated to me a daring and highly original interpretation of the dedication of Anthemion, recorded by Arist., Ath. Pol. 7.4, as accompanying a ’ akeitai en ’ statue (ei’ k
vn) of a man with a horse beside him,1 which an
’akrop
olei [‘stands dedicated on the Acropolis’]: ’ euZke ueo&iB, Diw
iloy ’Anuem
ivn t
Znd’ an
’ i t
eloyB i‘ pp
’ QZtiko &y ant ad’ ameic
amenoB. [‘Anthemion son of Diphilos dedicated this (statue) to the gods, having exchanged membership of the hippad telos for the thetic’]

This couplet is also quoted, with a reference to the statue, by Pollux VIII 131. Editors and commentators since the Aldine editio princeps (1502) have usually printed Pollux’s first line in some such form as ’ euZken (thus Dindorf, 1824, Diw
iloy ’Anuem
ivn t
ond’ ‘
ippon ueo&iB an
and Bekker, 1846). E. Bethe. in his edition of 1931, has Diw
iloy ’Anuem
ivn t
’ euZken ueo&iB ( ’Anuem
ivn ‘
ippon t
ond’ ‘
ippon an
ond’ ACL ’ euZken Ed. pr.). Dindorf, in the commentary to his edition of ueo&iB an
Pollux (V i 815), indicated a preference for ’Anuem
ivn ‘
ippon t
ond’ ’ euZke ueo&iB. I believe that Pollux wrote Diw
iloy ’Anuem
ivn t
an
ond’ ’ euZke ueo&iB, which I understand appears in at least one MS (see an
Sandys, ACA2 29, apparatus to Ath. Pol. 7.3),2 and that ‘
ippon is an intrusive gloss. This dedication, although presumably after 480, must be a good deal earlier than Aristotle’s day, for it is given as an example of t a ’ ’ anau
Zmata tvn & arxa
ivn [‘dedications by people of old’].3 Among the difficulties which arise here are the following: 1. Although there are possible precedents for a single pentameter in Attic epigrams prior to the fourth century,4 there are none for two pentameters. 1

In the London papyrus of the Ath. Pol. the statue is said, no doubt by a slip, to be ‘of Diphilus’. 2 [Sandys’s statement has no independent authority. It comes from Dindorf’s note ‘Falckenburgii Codex habet etc.’, i.e. the MS in question is Bethe’s F. But Bethe does not report this variant, although he is usually diligent in recording the readings of this important family (his II). PJT] 3 Raubitschek (DAA p. 206, no. 174) has suggested that the Anthemion concerned is the father of Anytus, whom we know to have been a self-made man: see Xen., Apol. Socr. 29–31; Plato, Meno 90AB; Schol. Plat., Apol. 18B. 4 IG I3 1227 is almost certainly a single pentameter, and II2 3123 (¼ DAA 174), which, like Anthemion’s dedication, is clearly the offering of a Hippeus, may well be.

Appendix

71

2. No surviving epigram on the Acropolis has the plural ueo&iB [‘to the gods’]. We should expect tfi Z& uev fi & [‘to the goddess’], i.e. Athena,5 except in some very special circumstances, for instance a great campaign in which vows had been made to all the protecting gods. 3. It is very strange to find an Athenian gratuitously advertising his former thetic condition in commemorating his elevation to the status of Hippeus. 4. Diphilus is perhaps not a very likely name for a Thes in early times. The first objection alone is almost insuperable, and tempts us to look for some interpretation which does not involve accepting the double pentameter as an actual dedication. The alternative interpretation offered by Miss Jeffery, which I find very attractive indeed, was suggested to her by Plut., Cimon 5.2–3, an anecdote about Cimon which may very well come from his friend Ion of Chios, whom Plutarch quotes in the very next sentence.6 During the Persian invasion of 480, we are told, Cimon, followed by his compan‘ iroi), went in procession through the Cerameicus to the ions (eta& Acropolis, and there dedicated his horse’s bridle to Athena, ‘on the ’ Z [‘‘horseground that what the city now needed was not i‘ ppik Z alk
’
power’’] but naym
axoi andreB [‘‘men fighting at sea’’]’; he then took a hoplite shield from among the temple dedications, prayed to the goddess, and went down to the sea, an encouragement to all. He was ’ going to serve, then, as a marine, an epib
atZB—and, as we saw ’ atai of the Athenian fleet were drawn from the earlier,7 the epib
Thetes in the late fifth century, and are very likely to have been of the same class at the time of the Persian wars, so that Cimon was in effect exchanging his status of Hippeus for that of a Thes. Anthemion, Miss Jeffery suggests, may have been another such Athenian as ‘ iroi, who vowed a statue to the Cimon, perhaps one of his eta& gods—perhaps the united gods to whom, we are told, the city had prayed for victory8—and duly dedicated it in thanksgiving after the victory, with a simple pentameter. The second line, taken by later writers, from the source of Aristotle onwards, to be part of the dedication, may be simply an addition or comment by some wit, perhaps Ion himself, recording the fact that Anthemion, like Cimon, had given up ’ ibein his hippad status for that of a Thes. The middle voice of ame
could evidently be used to mean ‘give A (accusative) in exchange for B (genitive)’, as in Pind., Pae. IV 16; Ps.-Plut., Mor. 607E, as well as ‘get A (accusative) in exchange for B (genitive)’, as in Solon fr. 2.2 (West); ’ ibv B II 1, may mislead here. Soph., Trach. 737. LSJ9, s.v. ame
5 6 8

As in IG I3 603, 647, 817 ¼ DAA 2, 90, 31. 7 FGrH III B 392 F 12. p. 21 above. See lines 4–6 and 38–40 of the decree of Themistocles, M/L 23 ¼ Fornara 55.

72

Athenian Democratic Origins

AFTERWORD There has been, to our knowledge, no substantial treatment of the main topics discussed by Ste. Croix. Rhodes, to whom Ste. Croix had shown his paper, cites Ste. Croix’s views extensively and in the main without dissent in his notes in CAAP on the key text, Aristotle, Ath. Pol. 7.4; but on p. 143 he remarks that ‘it is hard to believe that the liability to cavalry and to hoplite service were never defined’. Whitehead reinforces the case (accepted by Ste. Croix) for a military, not an agricultural, interpretation of the term zeyg&itai. Rosivach, by contrast, seeks to dissociate zeyg&itai from hoplites altogether. Connor suggests that Solon was reusing for new purposes an existing set of divisions that related to participation, according to the scale of offerings that one provided, in civic processions: ‘the Pentakosiomedimnoi in the early Archaic period would drive their chariots in the festival procession; the Hippeis would ride horses; the Zeugitae would march.’ Foxhall discusses the implications of the classifications for estate sizes; on the issues discussed by Ste. Croix she is mostly agnostic. Bugh, 20–34, denies all connection between the second Solonian class and military service; he is not followed by Spence, 180–2. Both these scholars believe (against Ste. Croix, but in a way that does no harm to his argument) that there were already true cavalry in 6th c. Athens (Bugh, 1–19; Spence, 9–17; so too Worley, 63–6). Jeffery’s theory on the dedication of Anthemion (Appendix) is briefly discussed by Rhodes, CAAP, 143–5. He questions whether the verb ’ ibomai can bear the required meaning, and with justification: of ame
the texts adduced by Ste. Croix in support of Jeffery, the context of the Pindar fragment is too broken for readings to be secure, and the passage of Ps. Plutarch in fact goes the other way. Bugh, G. R., The Horsemen of Athens (1988) Connor, W. R., ‘Tribes, Festivals and Processions; Civic Ceremonial and Political Manipulation in Archaic Greece’, JHS 107 (1987), 40–51, at 47–9 Foxhall, L., ‘A View from the Top. Evaluating the Solonian Property Classes’, in Mitchell/Rhodes, DPAG, 113–36 Rosivach, V., ‘Zeugitai and Hoplites’, Ancient History Bulletin 16 (2002) 33–43 Spence, I. G., The Cavalry of Classical Greece (1993) Whitehead, D., ‘The Archaic Athenian ZEUGITAI’, CQ 31 (1981) 282–6 Worley, L. P., Hippeis (1994)

2 Five Notes on Solon’s Constitution

(I) THE INTRODUCTION OF MAJORITY VOTING AT ATHENS By far the most important single step in the development of democracy was the decision to settle major political questions by actually counting heads, or rather hands (or pebbles), and allowing the numerical majority to have its way—making political decisions by majority vote.1 When this began in Greece, it may have been the first time in human history that such a thing had regularly happened, at any rate in a civilised society. It is very strange indeed that all writers known to me, both ancient and modern, with the single exception of Larsen,2 have ignored this extraordinarily significant development. Other writers, while giving careful attention to the emergence of relatively minor features of democracy, such as the Council and its probouleusis, have strangely neglected the fundamental innovation, which is similarly overlooked by all ancient writers, even Aristotle. It is often taken for granted that in societies such as those of ‘Dark Age’ Greece, ‘sovereign power’ is likely to have resided in the hands of the ‘demos’ in some form, whether as the assembled people at home, or as the collective body of warriors, on campaign and perhaps at home also. This, I believe, is likely to be true only in the sense that the rulers, whether kings or aristocrats, could not afford to go too far in opposition to the will of the great majority of their subjects, and might—although if they felt strong enough they might not—consult the common people, or at any rate the army, before they took some step which might not be effective without 1 [Much of the substance of the following two pages reappears as Appendix xxiv, ‘By shouting and not by vote’, of OPW. We have preserved the text here for continuity, but for full footnotes refer the reader to OPW.] 2 J. A. O. Larsen, ‘The Origin and Significance of the Counting of Votes’, in CP 44 (1949) 164–81. This article is of outstanding importance and has received far less attention than it deserves.

74

Athenian Democratic Origins

their co-operation. Legalistic thinking in the modern manner, which demands an explicit recognition of ‘sovereignty’ in some organ of state, whether kings or nobles or assembly of the people or the warriors,3 should not be forced upon such societies, where power is likely to be exercised by whoever possesses it at a given moment, above all at periods when institutions are unstable and in process of change, as in the Greek ‘Dark Age’. In Homer the demos decides nothing: it is the king who decides; the army or the people in assembly merely express their approval or disapproval by shouts or silence or—at a pinch—mutiny. Similarly, the most probable etymology of the Latin word for ‘vote’, suffragium, suggests that the earliest assemblies at Rome made their will known by noisy demonstrations. Among the Israelites the people (not in any formal assembly) signified their approval by shouting: at the election of Saul as king, ‘all the people shouted and said, ‘‘God save the king’’ ’; and Nehemiah’s seisachtheia was ratified when ‘all the congregation said ‘‘Amen’’ and praised the Lord’— never do we hear of a vote being taken. In what has been described as the ‘primitive democracy’ of some of the ancient Mesopotamian cities there is equally no clear sign of any voting, in the form of counting ‘those in favour’ and ‘those against’, nor is any such procedure visible in our records of other ‘pre-Classical’ Near Eastern societies such as those of the Hittites or the Phoenicians. We do not know when or in which state or by what stages the demos as a whole first gained the constitutional power of decision in a Greek city. By the mid-seventh century the Spartan damos had it;4 but it exercised it then and thereafter ‘by shouting and not by vote’,5 the presiding magistrate—an ephor, in the historical period—no doubt deciding which shout was the louder or loudest. Athens, like other Greek states, was ruled first by a king.6 During the Dark Age, monarchy was succeeded by a form of aristocratic government. Where supreme power lay at this period, in theory or in practice, we cannot say: perhaps no firm consti3 Against excessive theorising of this sort applied to the Rome of the regal period, see the sensible treatment by G. W. Botsford, The Roman Assemblies (1909), 168 ff., esp. 173–7. 4 See Plut., Lyc. 6.2–8. I agree with Wade-Gery, EGH 37–80, that the ‘Great Rhetra’ is to be dated some time after the second Messenian war (i.e. the first Messenian revolt), and not before it, as by G. L. Huxley, Early Sparta (1962), 49. 5 Thuc. I 87.2: bofi Z& kai oy’ c
Zwfiv. 6 [This claim is now controversial: see the works cited in Welwei, Athen, 80 n. 9.]

The Date of Solon’s Nomothesia

75

tutional principles had crystallised at this stage. Eventually—not later, surely, than Solon—sovereign power was held to reside in the Assembly, and was exercised by majority vote, by counting hands. As I have indicated, this development is of enormous interest and importance, as the indispensable sine qua non of political democracy. How and when did it take place? There must have been a particular occasion on which, for the first time, a political decision was taken by ‘the demos of the Athenians’ after a counting of votes, in which each man’s vote counted equally—a principle calculated to make an aristocrat shudder. When did this happen? Were such decisions taken de facto before being accepted as constitutionally proper? Was the practice copied from another Greek state? At any rate, if any single step taken by any Greek state, Athens included, in the whole field of social institutions is to be regarded as the most fruitful, it must surely be this. I should like to leave open the possibility that formal recognition of the taking of political decisions by majority vote at Athens was first given, expressly or implicitly, by Solon’s constitution. I do not think that a later date is possible, but of course this fundamental step towards democracy may have been taken at some time during the seventh century.

(II) THE DATE OF SOLON’S NOMOTHESIA I shall concentrate here on proving that Solon’s Nomothesia must be dated in the same year as his archonship. The great majority of historians, I would say, now accept 594/3 as the year of the archonship. This date is not absolutely secure, but it is much the best on the evidence we have, and I shall take it for granted and ignore the variants proposed by Beloch (592/1) and Kirchner (591/0).7 1. There are three main alternative views which I would reject: (a) First, there is Hammond’s theory8 that the Nomothesia took place in 592/1, two years later than Solon’s archonship and 7

It will be sufficient to refer to Cadoux, AAKH 93–103 (with 104–6; 83 & n. 50), where full references are given to the earlier literature. [See now Rhodes, CAP 120–1.] Less satisfactory, to my mind, is G. V. Sumner, in CQ 55 ¼ n.s. XI (1961) at pp. 49–54 (with the first complete para. on p. 38), who thinks that Aristotle’s date was 592/1, but that 594/3 is rather more likely to be right. 8 N. G. L. Hammond, ‘The Seisachtheia and Nomothesia of Solon’, in JHS 60 (1940) 72–83.

76

Athenian Democratic Origins

Seisachtheia. I need spend no time on this view, as it has already been sufficiently refuted by Cadoux.9 (b) Secondly, I would express total disagreement with the position recently adopted by Raubitschek,10 who invents for Solon an ’ Z [‘rule’] lasting sixteen or eighteen years, beginning in ‘592/1 arx
(594/3)’, evidently conceived as the year of Solon’s actual archonship, and ending in 576/5: Solon’s legislation is apparently to be taken as distributed over the whole period. In fact there is no justification for Raubitschek’s assertion that Solon’s legislative activities are described by Aristotle, Ath. Pol. 10.1, ‘as if they were spread over some years’. And Raubitschek’s whole position involves a series of absurdities. (i) First, we should have to accept the extraordinary notion that in Arist., Ath. Pol. 13.1, dating the ’ Zn [‘after first ‘anarchia’ in the fifth year met a t Zn S
olvnoB arx
’ Z means ‘period of power’. RauSolon’s arche’], the word arx
bitschek vainly tries to justify this on the ground that the same word is also used to mean ‘period of power’ in Ath. Pol. 17.1 ’ ’ fi Z& [‘Peisistratus grew old (Peis
istratoB . . . egkateg
Zrase tfi Z& arx in his arche’]). In fact there is no real parallel at all. By 17.1 the reader of the Ath. Pol. knows quite well what the arche of Peisistratus is: his arche in 14.3; 15.1; 16.7 is not an annual office but a rule; it is explicitly a tyranny (tyrann
iB) in 16.1, 7. Equally, the reader knows from 5.2 that the arche of Solon is something quite different, namely an archonship: nothing has been said to connect Solon with any other kind of arche. (ii) Secondly, what would be the point of dating in the curious way Raubitschek proposes, without first defining the length of the supposed arche? Unless this is given (and it is not), the formula is a useless one. (iii) Thirdly, Aristotle uses two successive formulas for dating from ’ Zn [‘after Solon’s arche’] in 13.1, and Solon: met a t Zn S
olvnoB arx
shortly afterwards, in 14.1, met a t Zn tvn & n
omvn u
esin [‘after the enactment of the laws’]. Raubitschek, who believes that the n
omvn u
esiB [‘enactment of the laws’] covered a period of sixteen or eighteen years, nevertheless wants us to believe that Aristotle would use the expression ‘in such-and-such a year after the n
omvn u
esiB’ as a dating formula—and in the opposite way from 9

AAKH 93–9 (with 99–103). A. E. Raubitschek, in his review of A. Masaracchia, Solone (1958), in CP 58 (1963) 137–40, at p. 139. 10

The Date of Solon’s Nomothesia

77

’ Zn [‘after Solon’s arche’]: there, the formula met a t Zn S
olvnoB arx
the dating is from the end of the undefined arche, while here, for no discernible reason, it is from the beginning of the n
omvn u
esiB! In view of 5.2; 6.1; 7.1; 10.1; 11.1, of course, the reader naturally and rightly accepts these dating formulas as equivalents, both referring to a particular year, that of Solon’s archonship. It is perverse to insist, in the interests of a preconceived theory, upon forcing an interpretation on Aristotle’s text which no ancient reader could conceivably have given to it. I may add that the argument11 that Plato, Timaeus 21C, dates Solon’s return from Egypt just before the coming to power of Peisistratus is also groundless: cf. Ath. Pol. 13 (especially the first sentence of §3), from which it is evident that the disorders were prolonged. It is no use pretending that there are no errors in the dating intervals of our text of the Ath. Pol. Where one of the figures given cannot be reconciled with plain statements in the narrative, it is likely to be wrong.12 (c) The third theory I wish to oppose is that of Hignett,13 who would put all Solon’s reforms in the late 570s. I can see no force in any of his arguments, or any reason for abandoning the established chronological tradition. For example: (i) even if there is some basis for the tradition (which by Herodotus’ time had received fictitious accretions, bringing in Amasis and even Croesus) of a ten-year period of travel by Solon just after his legislation, I see no reason to suppose that the poem he addressed to Philocyprus (fr. 19 West) was written during that particular set of journeys. If Aristocyprus, killed in 497,14 was indeed the son of Solon’s Philocyprus, it would be quite possible for him to have been born in the 550s and for Solon to have visited his father in the 560s or even 570s. And (ii) as for the appearance in the ‘laws of Solon’ cited by Plutarch (Sol. 23.3) of prizes of 100 drachmae for Athenian victors in the 11

Raubitschek, op. cit. 138, 139. In this connection it is worth noticing that the London papyrus of the Ath. Pol., unlike some other ancient MSS, does occasionally use figures (in the alphabetic script, of course) for ordinal numbers: see Gomme in JHS 46 (1926), at p. 176 n. 12; Cadoux, AAKH 83 n. 50; Jacoby, A 379 n. 141. And as Cadoux says, ‘the fact that most numbers in our present text are written out in full does not mean that they were not cyphers in earlier copies.’ This makes it even easier to believe in the corruption of such a figure as that in 14.1, where I would read tet
artfiv [‘fourth’] instead of deyt
erfiv [‘second’] (cf. Cadoux, AAKH 98–9). 13 HAC 316–21; followed by R. Sealey, in Historia 9 (1960), at pp. 159, 171, 177. 14 Hdts V 113.2. 12

78

Athenian Democratic Origins

Isthmian Games, scarcely a Hellenic festival before the late 580s, surely Corinth was such a near neighbour that Athenians might show special interest in its Games even before the numbered series began. In any event, however, I am perfectly willing to admit15 that some of what were called in later times ‘the laws of Solon’ were in fact post-Solonian additions, or alterations of Solonian originals. If, for example, Solon had provided only for prizes to Olympic victors, and a similar rule was made not long afterwards in favour of Isthmian victors, the two provisions might well come to be quoted together eventually as ‘laws of Solon’. (I may add that the prizes to Isthmian and Olympic victors are not actually said to have figured in Solon’s ‘Sixteenth Axon’: that description applies only to the ‘values of choice victims’, in Sol. 23.4.) Needless to say, we cannot altogether exclude the possibility (suggested by Cadoux)16 that Solon was responsible for further laws, outside his main ‘Nomothesia’; but I know of no evidence for anything of this sort, and to venture further here would be to rely on pure conjecture. 2. I do not think it very likely that, as Jacoby believed,17 the Athenian archon list contained ‘very brief annotations referring to ’
the list itself’, for example, that Solon was diallakt ZB kai arxvn [‘mediator and archon’], although this is certainly possible. However, to my mind there are two incontrovertible arguments, one negative and one positive, in favour of dating the Nomothesia in the same year as the archonship of Solon: (a) The negative argument is that we never find so much as a hint in any source that any part of Solon’s legislative activity was carried out in someone else’s archonship—as we hear, for instance, that Draco published his laws in the archonship of Aristaichmos (Arist., Ath. Pol. 4.1). In Arist., Ath. Pol. 5.2, the Athenians ’
rxonta S
e‘
ilonto koinfi Z& diallakt Zn kai a olvna kai t Zn ’ etrecan ayt ’ fiv& [‘by common consent chose Solon as polite
ian ep
mediator and archon, and turned over the constitution to him’], and in spite of the vagueness of 10.1 it seems clear from 13.1 and 14.1 that Aristotle conceived the whole of Solon’s legislative activity as falling in the year of his archonship. Sosicrates, according to 15 16 17

See pp. 317–22 below. AAKH 98 & n. 140. A 185, cf. 96–7, 175–6, 177, 185 etc.

The Date of Solon’s Nomothesia

79

Diog. Laert. I 2, and a majority of the MSS of Jerome’s version of the Chronika of Eusebius,18 date the legislation of Solon to Olymp. 46.3 ¼ 594/3 B.C. Even in the discrepant tradition recorded in Plut., Sol. 16.5, the sacrifice and the appointment of Solon as nomothetes come soon (tax
y) after the cancellation of debts, and are surely not thought of as falling in a later archon year. (Contrast Plut., Sol. 14.3, where Solon is nomothetes as well as diallaktes from the first.) As we have seen, the expressions met a t Zn S
olvnoB ’ Zn [‘after Solon’s arche’] in Ath. Pol. 13.1 and met arx
a t Zn tvn & n
omvn u
esin [‘after the enactment of the laws’] in 14.1 are most naturally accepted as equivalents. (b) The positive argument, even more decisive, is provided by two different laws the wording of which proves they were passed during Solon’s archonship: (i) The first is Solon’s amnesty law, specifically quoted by Plutarch (Sol. 19.3) as the eighth law on Solon’s ‘Thirteenth
’ imvn o
‘ soi a ’
timoi Zsan ’ S
’
Axon’. It begins, at
prin Z olvna arjai ’ epit
imoyB ei
nai pl Zn o
‘ soi ktl. [‘all those who were deprived of civic rights before Solon’s arche are to enjoy civic rights, except those who etc.’]. Hignett19 doubts the authenticity of this text, and thinks that possibly Plutarch or his authority ‘got it from Demetrios of Phaleron, who may have translated the archaic language of the original into the terminology of his own age. The alternative is to suppose that the law was taken from a later revision of Solon’s code, and that the reference in it to Solon was an editorial interpolation’—though what possible reason an editor (or Demetrius) could have for wishing to interpolate a reference to Solon’s archonship at this point is hard to imagine. Hignett has a second string to ’ S
’
his bow: he also suggests that the words prin Z olvna arjai [‘before Solon’s arche’] may refer not to Solon’s archonship but to ‘the special appointment in virtue of which he was legislating; ’ Z earlier need not have caused the fact that he had held another arx
any practical difficulty in the application of the law’. This is another curious argument and shows the desperate straits to which the defenders of a late date for Solon’s Nomothesia are driven. Everyone would naturally take the words I have quoted to be a reference to Solon’s archonship. If the law had been passed

18

Euseb., Chron. (ed. Helm), p. 99b.

19

HAC 313, 318–19.

80

Athenian Democratic Origins

in a later year, the phrase in question would have been quite unnecessarily misleading. (ii) The second law, which is of quite a different character, seems to have escaped notice in this connection; and it surely settles the matter beyond reasonable doubt. Quoted in slightly different terms by Ps.-Dem. XLIV 68 and XLVI 14, this law allows the making of wills in certain circumstances. In both cases the quotation from the law (specifically called ‘the law of Solon’ in XLIV 67) ’ fi Zei begins with the words o
‘ soi m Z epepo
iZnto . . . o
‘ te S
olvn ei’ s
Z ’ Zn [‘Whoever had not been adopted . . . at the time ei’ B t Zn arx
when Solon entered into his arche’]. Here it is surely impossible fi Zei to believe that the form of words used, o
‘ te S
olvn ei’ s
Z ei’ B t Zn ’ Zn [‘at the time when Solon entered into his arche’], can refer to arx
anything but the beginning of Solon’s archonship. And there is no reason why the law should have referred to that moment unless it was passed during the year in question. The expressions I have cited from the two laws, differently phrased as they are, reinforce each other: to explain away both is impossible. It would be perverse to pretend that these laws could have been passed in any year other than that in which Solon was archon. The Nomothesia, then, took place in Solon’s archonship. *** As I have already indicated, and will demonstrate in detail in a later essay,20 it is wrong to imagine that Solon’s Nomothesia left the constitutional framework unaltered and merely passed laws intended to operate within a constitution that existed before Solon and continued with little change after him. Among Solon’s laws were important constitutional innovations, including the introduction of property qualifications for office, appeal to the Heliaia, and very probably the creation of a Council of Four Hundred and the formal recognition of the right of the demos to elect the magistrates.

(III) THE EUPATRID MONOPOLY OF THE ‘STATE MACHINE’ I have already expressed the view that Solon’s most important political reform was the destruction of the eupatrid monopoly (or 20

pp. 310–17 below.

The Eupatrid Monopoly of the ‘State Machine’

81

virtual monopoly) of state office,21 and the basing of qualification for such office upon the four property classes. Now we ourselves are used to states which have an elaborate machinery of government, an apparatus which in many cases will continue to function irrespective of what individuals or groups hold the reins of power, and can often be taken over more or less intact even by a revolutionary regime. We too easily forget, therefore, that at the moment when the Solonian constitution began to operate, virtually the whole of such rudimentary ‘state machinery’ as existed at Athens must have been entirely in the hands of the eupatrids, and that there was no other apparatus which could be used to coerce the eupatrids if—as would be only too natural—they refused to work the new constitution, or quietly sabotaged it. (There is an obvious analogy in the ‘Struggle of the Orders’ at Rome, where the mere passing of plebiscita or leges was sometimes evidently insufficient to procure the observance of constitutional innovations disliked by the patricians.) The scorn and bitterness with which Greek aristocrats could regard wealthy upstarts who deprived them of their monopoly of political privilege, or merely married into noble families, is nicely illustrated by the poetry of Theognis, a citizen of Athens’ nearest neighbour, Megara, not so long after Solon’s day.22 To my mind the appreciation of this fact is fundamental to an understanding of the course of events between Solon’s archonship and the tyranny of Peisistratus. It was the eupatrids alone who, as a class, lost by Solon’s constitutional reforms; and the most likely ’ explanation of the anarx
iai [‘years during which there was no archon’] of (probably) 590/89 and 586/5 and the Damasias episode of c. 582-023 is surely that the eupatrids refused, when they could, to allow the election of non-eupatrid archons. I do not see how we can decide whether the arrangement described by Aristotle, Ath. Pol. 13.2 (five of ten archons from the Eupatridai, three from the Agroikoi and two from the Demiourgoi), represented an outright reaction, in which the eupatrids succeeded in claiming a half share in the archonship, or was a relative setback for them, a realistic compromise whereby the eupatrids, who had been having a monopoly of elected archons, were forced to draw back and claim only a 21

22 See pp. 6–7 above. [See CSAGW 278–9.] For the chronology and other problems of the early Athenian archon list, see Cadoux, AAKH. 23

82

Athenian Democratic Origins

half share, but at least received that recognition of their special position in the state which Solon had denied them. (I think the latter explanation is probably true, if only because a Damasias had been archon in 639/8,24 and it is not unlikely that the later Damasias was a relative of his—he could easily have been his grandson— and therefore a eupatrid. But all this is uncertain.) We may sympathise with Solon’s detestation of tyranny and his refusal to assume the role of tyrant which the discontented demos would have thrust upon him. At the same time we must admit that Solon’s great work might have been largely nullified had not Peisistratus, using violence to overcome aristocratic resistance, enforced Solon’s constitution—for which there is a remarkable unanimity in the best sources that Peisistratus did preserve the laws of Solon, apart from keeping hold on the offices.25 At the moment Solon’s constitution came into effect the administration of the state seems to have been mainly in the hands of the nine archons,26 above all the archon eponymus.27 Doubtless the Areopagus played an important part, but unfortunately we are quite unable to define the manner in which the alleged nomowylak
ia ‘guardianship of the laws’ was exercised by the Areopagus, or to identify any single action on its part at any time which might illustrate that guardianship. Aristotle28 believed that the Areopagus always consisted of ex-archons; and although this statement has been doubted for the pre-Solonian period,29 we 24

Dion. Hal., AR III 36.1. See Thuc. VI 54.6; Hdts I 59.6; Arist., Ath. Pol. 14.3; 16.2,8; Plut., Sol. 31.3. Ath. Pol. 22.1 is presumably to be explained in the light of the end of 16.7 (syn
ebZ ’ Zn [‘for g ar y
‘ steron diadejam
envn t vn & yi‘
evn pollv fi & gen
esuai traxyt
eran t Zn arx
it came about that later on, when his sons had succeeded him, the regime became much harsher’]), as referring to the rule of Hippias. 26 ’ ea See Thuc. I 126.8: in the 7th century t a p
olla t vn & politik vn & oi‘ enn
’
arxonteB e
’ prasson [‘the nine archons administered most of the affairs of the state’]. 27 & See Arist., Ath. Pol. 13.2: fivffl kai dZlon o‘
ti meg
istZn ei
xen d
ynamin o‘ ’
’ ZB & arx & [‘by which it is arxvn
wa
inontai g ar ai’ ei stasi
azonteB peri ta
y tZB tZB clear that the archon held the greatest power; for it is clear that there was always strife arising over this office’]. 28 Ath. Pol. 3.6. 29 e.g. by Jacoby (ii 108 n. 33) and Ehrenberg (GS 61). Here much depends on whether we accept the statement of Arist., Ath. Pol. 8.2, that the archons before Solon’s time were elected by the Areopagus (cf. Busolt, GG II2 143 n. 2), ’ prokr
itvn [‘sortition from prokritoi’] which—whether we accept kl
ZrvsiB ek from Ath. Pol. 8.1 or believe that Solon provided for plain election (see pp. 89ff. 25

The Solonian Council of Four Hundred

83

must surely accept it as a fact from at least the time of Solon onwards. Once non-eupatrid archons began to be elected, the composition of the Areopagus would begin to change; but it must have remained predominantly eupatrid for many years. Apart from the archons and the Areopagus there were doubtless other officials, such as the tribe-kings (wylobasile&iB), who are said to have been eupatrids,30 and the kolakretai, whose very name attests their antiquity; but there is no evidence what powers any of these magistrates enjoyed.31 An appreciation of the eupatrid monopoly of the machinery of state in Solon’s day is important not only for the understanding of the period between Solon and Peisistratus but also in connection with the problem of the Solonian Council of Four Hundred, to which I now turn.

(IV) THE SOLONIAN COUNCIL OF FOUR HUNDRED Many modern writers have expressed their opinions on the question whether Solon created a Council of Four Hundred, as stated ’ iZse by Aristotle (Ath. Pol. 8.4) in the terse phrase, boyl Zn d’ epo
‘ ’ ek
‘ astZB wylZB & [‘and he set up a Council tetrakos
ioyB, ekat on ej of Four Hundred, a hundred from each tribe’]. The only other important source to mention such a Solonian Council, namely Plutarch (Sol. 19.1–2), cannot be shown to possess any earlier authority than Aristotle, although of course he may be using an older Atthidographer, directly or indirectly. Aristotle gives no details about the composition or method of election or functions of this council, and some scholars, including De Sanctis32 and Beloch,33 have treated it as a fiction. Recently Hignett has below)—conflicts with the statement in Pol. II 1273b41–74a2 that Solon made no ‘ t vn ’ vn change in Z & arx & a‘
iresiB [‘appointment to office’]. 30

Poll. VIII 111. Hdts V 71.2 speaks of oi‘ pryt
anieB t vn & naykr
arvn, o‘
i per e
’ nemon t
ote t aB ’Au
ZnaB [‘the prytanes of the naucraries, who governed Athens at that time’]; but this of course is contradicted by Thuc. I 126.8. The general opinion of modern scholars is well expressed in the words of Gomme (HCT I 426): ‘Hdts may be following a version favourable to the Alkmeonidai which said that the naukraroi and not the archons (of whom Megakles was perhaps the archon eponymos) were responsible for the sacrilege for which the Alkmeonidai were, however, blamed.’ 32 33 A2 251. GG I2 i 366 n. 2. 31

84

Athenian Democratic Origins

described it as a fabrication of oligarchic propaganda in 411 B.C.,34 and Jacoby, in the Addendum to his commentary on the Atthidographers, was ‘rather inclined to agree with him’.35 On the other hand, there are plenty of historians who have not seriously doubted the existence of a Solonian Council.36 The reader of Wade-Gery’s Essays37 would not even become aware, I fancy, that the existence of Solon’s Council had been called in question but for the statement on p. 199 that ‘the Chian Bouleˆ is surely probouleutic and is good evidence for the reality of Solon’s 400’. I have dealt elsewhere in this book38 with the general problem of the credibility of Aristotle’s account of early Athenian history and with some particular parts of that account. In the light of the conclusions which emerge, I would maintain that we cannot tell for certain whether or not Aristotle had good authority for his statement. Some have found it suspicious that he does not explain the functions of the Solonian Council or how it was elected or from whom; but this is quite unjustified, for—not to mention other inexplicable omissions in the Ath. Pol.—Aristotle is equally silent about the Cleisthenic Council. Whether he had reliable authority or not, Aristotle may be reporting a correct tradition. To my mind, some of the arguments which have been used in this controversy are worthless. For example, how can we decide whether the oligarchs in 411 first decided upon the number 400 because that was the size of the council they wanted, and then pretended there was ancestral precedent for it,39 or whether they chose the number 400 because there was already a tradition that this had been the number of Solon’s Council? To me the second alternative seems much more likely, but that is probably because I believe in a Solonian Council on other grounds. Again, how can we 34

HAC 92–6, cf. 273 etc. Jac. ii 530. Jacoby is one of the few scholars who have not declared strongly for or against the Solonian Council; cf. his ii 107 n. 25. 36 Including Busolt, Cloche´, Ehrenberg, Gilbert, Ledl, Meyer and Wilamowitz: see in particular Ledl, SAAV 275–85; Busolt-Swoboda, GS II 845–6 (esp. 845 n. 3), 1586. For a more recent defence of a Solonian Council, see Kahrstedt, BGRF, esp. 6–8. (I cannot see why he should want to have Peisistratus abolish the Council.) It is interesting that four years earlier (in UMA 49 n. 7) Kahrstedt had declared the Council to be ‘wohl erheblich ju¨nger als Solon’ (‘appreciably later than Solon’). 37 EGH, esp. 145–8. 38 Ch. 7. 39 See Arist., Ath. Pol. 31.1. This is the earliest evidence for a Council of Four Hundred. 35

The Solonian Council of Four Hundred

85

possibly be sure whether the ‘two anchors’ of Plutarch, Sol. 19.2, come from one of Solon’s poems or not?40 I shall merely state those considerations which seem to me to have some real force: 1. The principal argument which leads me to accept the existence of a Solonian Council of Four Hundred is as follows: (a) As we have just seen, all such ‘state machinery’ as existed at the moment when the Solonian constitution began to operate must have been entirely in the hands of eupatrids, the one class which had suffered politically by Solon’s reforms; and of course it was all too likely that for at least some time to come a high proportion of archons—and therefore new Areopagites—would similarly be eupatrid. This could not be prevented. Solon must have known perfectly well that (as indeed, I think, events proved)41 the eupatrids would not voluntarily surrender power; and in the absence of any state machinery to compel them, the only possible way of giving the new constitution a real chance to work was by giving a measure of real power to the demos as a whole: that is to say, ultimately to the Assembly. (b) Solon cannot have wished, however, to give such power directly to the Assembly, a mass meeting of all citizens, the great majority of whom as yet lacked political maturity.42 In the bitter class strife which had led to his appointment as diallakt ZB kai ’
rxvn [‘mediator and archon’] he conceived his own role to be a essentially that of an arbitrator, giving to each side exactly what it deserved: to the demos, he says, he gave such privilege as was 40 Jac., ii 107 n. 25, ‘cannot believe’ the two anchors appeared in Solon’s poems; but he does not say why. I think they may well have done, but there is no certainty of this, and I see no way of settling the matter. Miss Freeman, WLS 79 n. 1, sees poetical usage: she says that ‘parts of the passage seem to fall into trochaics’, and gives a specimen. Wade-Gery, EGH 146 n. 2, thinks Plutarch’s words, ‘as Miss Freeman has pointed out, . . . evidently come from Solon’s poems, probably the iambic poem of defence, quoted ’Au:pol. 12.4–5’; and proceeds to give most of two possible lines. J. M. Edmonds, in the Loeb Elegy and Iambus, I 154, ventures on a reconstruction in iambics. F. Sta¨helin, in Hermes 68 (1933) 343 ff., at p. 345, has slightly strengthened the case in favour of attributing the metaphor to Solon himself, by pointing out that the idea of the storm-tossed ‘ship of state’ occurs in other 7th/6th century poets: Alcaeus fr. 6, 326 (L/P); Theogn. 671 ff. (Sta¨helin would add Pindar, Isthm. I 36 ff.; but I cannot see that this is relevant). 41 See pp. 81–2 above. 42 ‘ evn d’ eiffl B men e
‘ kastoB al
’ vpekoB Solon fr. 11.5–6 (West) is relevant here: ym
’
ixnesi ba
inei, s
ympasin d’ ym& ‘ in xa &ynoB e
’ nesti n
ooB [‘your trouble is, each of you treads the fox’s way, but your collective wits are thin as air’ (tr. West)].

86

Athenian Democratic Origins

fitting, neither more nor less;43 and at the same time he provided that the rich and powerful should suffer nothing unseemly.44 He describes himself as standing with a strong shield before both sides, allowing neither to win an unjust victory.45 What he wanted was that the demos should ‘follow its leaders best, neither under too little restraint nor under compulsion’.46 Yet of the ‘leaders’ of the day he had spoken in stern rebuke in an elegiac poem apparently written before his archonship: they had an ‘unjust mind’ and were arrogant and greedy, and they took no heed of ‘the dread foundations of Justice’.47 What was more natural than that he should give to the demos a body which would guide and inform its deliberations, quite independently of the older organs of state, and at the ’ syn
same time act as a check upon spontaneous disorder en odoiB ’ to&iB adik
eoysi w
ilaiB [‘in conspiracies beloved of wrong-doers’],48 such as had occurred during the crisis preceding his archonship?49 (c) Unless a special Council were created to provide it, probouleusis would inevitably remain a prerogative of the Areopagus; and this would be unfortunate, both because of the eupatrid domination of the Areopagus and because of the prestige of that body— any proposals it might bring forward would have been invested with great authority, and the Assembly might well have tended to become a mere rubber-stamp. On the other hand, a new probouleutic Council elected from all three of the upper property classes, and consisting of as many as 400 men (large enough, especially if 43 Solon fr. 5.1–2 (West): d
Zmvi men g ar e
’ dvka t
oson g
eraB o‘
sson ’ ’ elvn oy
’ t’ eporej
’ & oy
’ t’ aw
eparke& in, timZB amenoB [‘The commons I have granted privilege enough, not lessening their estate nor giving more’ (tr. West)]. He cannot mean that he had neither added to nor taken away from what the demos already had: his standard is what he thinks they ought to have. In a later poem in iambics (fr. 37.1–3) he can say that the demos now (after his reforms) have what they could never have dreamt of. 44
’ Ibid. 5.3–4: o‘ i d ’ ei
xon d
ynamin kai xr
Zmasin Zsan agZto
i, kai to&iB ’ ’ ewras
amZn mZden aeik eB e
’ xein [‘the influential, I have saved them from all mistreatment too’ (transl. West)]. 45 ’ ’ ’ Ibid. 5–6: e
’ stZn d’ amwibal vn krater on s
akoB amwot
eroisi, nik &an d’ oyk ’ eroyB ad
’ ikvB [‘I took my stand with strong shield covering both sides, e’
ias’ oydet
allowing neither unjust dominance’ (tr. West)]. 46 ’
‘ ffl ’ arista & an syn Zgem
onessin e
‘ poito, m
Zte l
iZn Ibid. 6.1–2: dZmoB d’ vd’ ’aneueiB m
Zte biaz
omenoB [‘Thus would the commons and its leaders best accord, not given too free a rein, nor pushed too hard’ (tr. West)]. 47 Ibid. 4.7–16. 48 Ibid. 4.22. [w
ilaiB Bergk, accepted in Diehl’s edition used by Ste. Croix; w
iloyB West]. 49 For the latter motive, cf. Plut., Sol. 19.1.

The Solonian Council of Four Hundred

87

iteration were prohibited or restricted,50 to ensure the presence of a good number, surely an actual majority, of non-eupatrids), would have nothing like the prestige and authority of the Areopagus, and the Assembly might play its proper part. (d) And it was not only a probouleutic body which was required: an organ of state which could act in an administrative capacity, if only in routine matters, was also necessary, unless the still entirely or mainly eupatrid archons and Areopagus were to continue as the sole executive. (e) There is good reason to suppose that the Council was also given considerable judicial powers. According to Aristotle, Ath.
kyr
ia kai xr
‘ de boyl Pol. 45.1, Z Z pr
oteron men Zn Zmasin ’ & zZmivsai & kai dZsai kai apokte& inai [‘In former times the Council had the power to impose fines, imprisonment, and the death & boylZB & kr
iseiB ei’ B penalty’] (cf. 45.2–3; also 41.2: kai g ar ai‘ tZB ’ & elZl
t on dZmon yuasin [‘for the Council’s jurisdiction has passed into the hands of the demos’]). The date at which the powers of the Council were restricted is disputed;51 but even if Aristotle is thinking here (as he probably is) of the Cleisthenic Council only, we may surely infer that Solon’s Council is also likely to have had much greater authority in the judicial sphere than the Council which we know in the late fifth and fourth centuries. The ‘ dZmos
iZ [‘Council of the People’] Council described as the bol ZZ at Chios in the second quarter of the sixth century, to which I shall refer presently,52 had important judicial functions. In the circumstances with which Solon was confronted, therefore, a new Council, probouleutic and executive and probably possessing judicial powers, was certainly a wise innovation, and Solon might have reason to hope that it would prevent the new constitution from being wrecked by eupatrid hostility. 2. The second argument which I find attractive is that the Council which Cleomenes is said by Herodotus53 and Aristotle54 to have tried to dissolve, and to have led the resistance to him, must 50 We cannot say, of course, whether Solon’s Council was governed by the rule known in the fully developed democracy, forbidding anyone to be a councillor more than twice (Arist., Ath. Pol. 62.3). 51 See P. Cloche´, ‘Le Conseil Athe´nien des Cinq Cents et la peine de mort’, REG 33 (1920) 1–50; Busolt-Swoboda, GS II 1046–7; Lipsius, ARR 45–7 (esp. n. 142); Hignett, HAC 240–1, with 152–5; Bonner and Smith, AJHA I 336–45, with 195–209. 52 53 54 P. 89 with n. 59 below. V 72.2,3. Ath. Pol. 20.3.

88

Athenian Democratic Origins

surely have been the Solonian Four Hundred. In the first place, it certainly cannot have been the Areopagus, as some scholars have supposed.55 Both Herodotus and (following him) Aristotle speak ‘ boyl
‘ boyl ‘ en ’ ’Are
ifiv p
of Z Z ‘the Council’, not of Z ZZ agfiv [‘the Council on the Areopagus’] or any similar phrase; and I do not see ‘ boyl
how Z Z tout court, in such a context, can possibly mean the Areopagus. (Is there a parallel in the whole of the surviving sources?) Nor, one may think, is the Areopagus very likely to have been the man source of resistance to the ally of Isagoras. ‘ boyl
That Z Z in question is the Five Hundred of Cleisthenes has been maintained by some historians.56 However, it seems to me (as, I think, to most people) very unlikely that the intervention of Cleomenes can have come as late as this, when the new constitution was actually working—the Council could not have come into being until the membership of demes and trittyes and tribes had been settled (a lengthy process), and elections had taken place.57 That Cleisthenes ‘may have secured the appointment of a provisional Council of Five Hundred to hold office until the necessary preparations for his new constitution were completed’58 seems to me an unnecessary and exceedingly implausible conjecture, inspired by nothing but a determination to reject a Solonian Council. Herodotus, I conclude, was certainly thinking of the Four Hundred. There was, then, in his day a tradition of a pre-Cleisthenic Council, known to us in a context which has nothing to do with the constitutional activity of the Council, and therefore the less likely to have been invented for a propagandist purpose, such as attributing as many as possible of the institutions of the developed democracy to the great lawgiver. 3. A third argument which seems to have some weight is that the Cleisthenic constitution evidently worked quite remarkably well from the moment it was instituted; and it would be much easier to understand this if the Council, a vital part of the whole fabric, were 55 For instance, Hignett, HAC 94. Such a passage as Arist., Pol. II 1274a1–2, where t Zn boyl
Zn [‘the Council’] is certainly the Areopagus, provides no fair ’ Zn en comparison, for only three lines earlier (1273b39–40) Aristotle has written t ’Are
ifiv p
agfiv boyl
Zn [‘the Council on the Areopagus’]. 56 Most recently Kahrstedt, BGRF, esp. 2–6; F. Schachermeyr, in Klio 25 (1932) 334–47. Contrast P. Cloche´, in REG 37 (1924) 1–26; Wade-Gery, EGH 147; Cadoux, AAKH 115 n. 249. 57 See esp. Ledl, SAAV 275–85; Eliot, CDA 145–6. 58 Hignett, HAC 94. Contrast Cloche´, op. cit. (n. 56 above) 21–3.

‘Sortition from Pre-selected Candidates’

89

not an entirely new and untried invention. No one nowadays will suppose that it was the tyrants who created the Council, and if we feel that it is likely to be older than Cleisthenes we have no reason to doubt a Solonian origin. ‘ dZmos
iZ [‘Council of 4. The Council described as the bol ZZ the People’] which existed at Chios in the second quarter of the sixth century59 affords evidence that within a generation after Solon a Greek city could have an elected body of this nature. Perhaps, as Tod believed, ‘the epithet dZmos
iZ twice added to bol
Z suggests that another council continued to exist at Chios, just as at Athens the Areopagus lived on side by side with the Solonian, and later the Cleisthenic, boyl
Z.’ ‘ boyl 5. Diog. Laert. I 49 uses the phrase Z Z, Peisistrat
idai o
’ nteB [‘the Council, who were Peisistratids’], in relation to the Council existing just before the time at which Peisistratus seized power. I am not sure this passage is worth anything at all; but, such as it is, it can only refer to the Council of Four Hundred and not to the Areopagus. Consequently, although I think it may perhaps be wrong to assert the existence of a Solonian Council as a positive historical fact, the arguments for it seem to me very much stronger than any which can be brought against it: these amount to little more than that there could have been no need for such a Council in Solon’s day, when the importance of the Assembly had scarcely begun to develop, and the fact that Aristotle may have had no details of its composition and so forth—arguments which I have dealt with already.

’ prokr
itvn [ ‘ S O R T I T I O N ( V ) kl
ZrvsiB ek FROM PRE-SELECTED CANDIDATES’] (ATH. POL. 8.1) ’ prokr
itvn The only explicit authority for the use of kl
ZrvsiB ek [‘sortition from pre-selected candidates’] in Solon’s constitution is Tod 1 ¼ M/L 8 ¼ Fornara 19. In the preserved portion of the Chian decree the ‘ dZmos
iZ [‘Council of the People’] are mainly judicial: see lines duties of the bol ZZ C1-15. Sealey, in Historia 9 (1960), at p. 161 n. 36, makes the astonishing suggestion that bol Z dZmos
iZ ‘might mean something quite different, namely ‘‘village council’’ ’: he has evidently failed to notice that the council is to consist of as many as 50 from each tribe; and dZm
osioB is never, as far as I know, used in the sense of ‘belonging to a deme’. [But for other doubts about Tod’s interpretation of 59

90

Athenian Democratic Origins

Arist., Ath. Pol. 8.1 which, as is well known, is irreconcilable with the statement in Pol. II 1273b41–74a2, 1274a15–17, that Solon ’ vn seems not to have changed t Zn tvn & arx & a‘
iresin [‘the election to magistracies’].60 There are three or four other texts which may reflect the same view as that of Ath. Pol. 8.1,61 but they are vague and unsatisfactory and have no weight: it is Aristotle’s statement which alone deserves to be taken seriously. Some passages in Isocrates which are commonly brought into this context are entirely irrelevant, as I shall demonstrate presently.62 Ath. Pol. 8.1 can be divided for convenience into four sections: ’ (1) the magistrates (arxa
i) were chosen by lot from prokritoi (‘pre-selected candidates’), according to tribes; (2) in the prokrisis (‘pre-selection’) for the nine archons, each tribe chose ten men, and among these the lot was cast; (3) from this the custom still ’
diam
enei) that each tribe chooses ten men by lot survives (o
‘ uen eti and then makes a further choice by lot among these; and (4) evidence (sZme&ion) that Solon had the magistrates chosen by lot according to their property qualification is furnished by the law concerning the treasurers (of Athena), which is still in use: for it orders the selection of the treasurers by lot from the Pentakosiomedimnoi. The question now arises: are the statements in my sections 3 and 463 arguments, proofs, are they the basis of Aristotle’s belief that ’ prokr
itvn went back to Solon; or are they, as it were, kl
ZrvsiB ek dZmos
iZ see H. van Effenterre and F. Ruze´, Nomima, Recueil d’inscriptions politiques et juridiques de l’archai€sme grec, I (1994), no. 62.] 60 See Hignett, HAC 322–3, sufficiently refuting the attempt of Wilamowitz (AuA I 64–73) and Ehrenberg (in RE XIII ii 1469, 1472–3) to get rid of the contradiction between Ath. Pol. and Pol. 61 ’ palaio &y [‘allotted Plut., Per. 9.3 (the archonships at Athens were klZrvtai ek
’ Zsan since ancient times’] ); Paus. IV v 10 (in the 740s B.C. at Athens oyk pv t
ote ’ ’
oi‘ tv fi & kl
Zrfiv kat’ eniayt on arxonteB [‘there did not yet exist the yearly archons appointed by lot’]); Dem. XX 90 (speaking of Solon in connection with toyB ’ i toyB n
uesmou
etaB toyB ep omoyB klZroym
enoyB [‘the Thesmothetai, appointed by lot to administer the laws’]); and Ps.-Dem. LIX 75 (referring to Theseus as effecting a synoecism and setting up a democracy, and continuing, t on men ffl ’ en Ztton ‘ ito ek ’ prokr
itvn kat’ andragau
’ & basil
ea oyd o‘ dZmoB fi Zre& ian xeiroton vn & [‘just as before, the people continued to chose the Basileus from prokritoi, casting their votes according to the excellence of the candidates’] ). 62 At pp. 101–3 below. 63 Like the similar ones, based on circumstantial evidence on survivals, in Ath. Pol. 3.3 (lines 4–7 and 9–11 OCT); 3.5 (lines 19–21); 7.1 (line 20); 7.4 (lines 5 ff.); 13.2 (lines 25–7); 13.5 (lines 11–13).

‘Sortition from Pre-selected Candidates’

91

mere footnotes, illustrations? The former is often assumed, as by Hignett;64 but a good case can be made for supposing that Aristotle at least believed he knew independently what Solon had enacted, and in sections 3 and 4 was simply adding some information for the benefit of those who did not share his own special knowledge of the Solonian laws.65 On this I will merely remark that even if Aristotle believed he had good information about the Solonian law in question, he may have been mistaken in that belief, as he was in some of his other statements in the Ath. Pol. (for instance, those about the ‘constitution of Draco’ and the qualifications of the Solonian census classes);66 and that those who are sceptical about Aristotle’s assertions in sections 1 and 2 will certainly not be impressed by the additions in 3 and 4, whether they are to be taken as proofs or footnotes. In view of Ath. Pol. 22.5, the custom mentioned in section 3 is obviously no evidence at all for any period before 487. About my section 4 two things need to be said. (a) It is not as illogical as it appears. At first sight one is inclined to object that there is no reference to prokrisis in the election of treasurers. But this would be to take too much notice of the modern division into chapters, which has put my section 4 (the last part of 8.1) in a different chapter from 7.3 (describing the election of magistrates according to property qualifications). In fact chs. 7 and 8.1 form a ‘ astoiB an
’ alogon single whole, and it is 7.3 (especially the words ek
’ ’ Zn [‘assigning offices tfiv& meg
euei to &y tim
ZmatoB apodido yB t Zn arx
to each [sc. of the property classes] in accordance with the size of their property rating’]) to which the last part of 8.1 (my section 4) harks back, not the earlier part of 8.1 (my sections 1 and 2). (b) On the other hand, even if the law ordering the treasurers to be chosen by lot did go back to Solon, no inference would be justified about the method of election of archons, whose office at this time was very much more responsible and onerous. (As Aristotle remarks in Pol. V 1309b4–8, the post of treasurer demands more than average ’ virtue (aret
Z), but everyone possesses the required knowledge ’ (epist
ZmZ).) The statement of Aristotle which we are considering has been rejected by the great majority of recent historians, including 64

HAC 323–4. Prof. Andrewes has been kind enough to show me some persuasively written unpublished material defending this position. 66 See pp. 28–32 with 7 n. 10 above. 65

92

Athenian Democratic Origins

Beloch, Bengston, Busolt, De Sanctis, Hignett, Kahrstedt (‘diese groteske Vorstellung’), Ledl, Meyer (‘undenkbar’) and apparently Jacoby.67 That there is nothing inherently absurd in the idea of election of archons by lot, at any rate if preceded by a prokrisis, has been maintained by a minority of scholars, including Ehrenberg, Heisterbergk, Le´crivain, and most recently Wade-Gery.68 In my opinion, the most likely explanation is that in response to the practice of choosing magistrates by sortition some fourth-century democratic author invented a Solonian precedent, which Aristotle mistakenly accepted, fortifying the theory in his characteristic manner by arguing back from survivals. Hignett believes that ‘nothing can be decided either way by a priori arguments; thus it is futile to argue . . . that partial sortition could not have been employed in the appointment of the chief magistrates of the state.’69 He asserts that ‘the problem can only be resolved by an examination of the ancient evidence, and the result of that examination is fatal to the alleged reform.’ He presents quite a strong case, but I do not feel that the results of his analysis are nearly as conclusive as he imagines. For me it is the more general arguments which are decisive. Because of what seems to many of us the extreme implausibility of Aristotle’s statement, most of those who have rejected it have not thought it worth while to set out these more general arguments; and this is mainly what I now propose to do, although I intend always to keep as close as possible to what evidence there is. *** 1. To say it is ‘impossible that Solon permitted the lot to choose the nine archons, one of whom was to be Polemarch, and command in war’, Wade-Gery claims,70 is ‘demonstrably wrong’—but he does not demonstrate this; he merely asserts that ‘any good man could command in war, just as any good man could govern, or 67 Beloch, GG I2 ii 319; Bengtson, GG2 122; Busolt, GG II2 274–7; GS II 842–4; De Sanctis, A2 243; Hignett, HAC 228–9, 321–6; Kahrstedt, SSA 48–9; Ledl, SAAV 344–78; Meyer, GdA III2 608 & n. 2; Jac. ii 116 n. 17. 68 Ehrenberg, in RE XIII ii 1451 ff., at cols. 1468–74, s.v. Losung; C. Le´crivain, in DS IV ii 1401 ff., at pp. 1403–6, s.v. Sortitio; Wade-Gery, EGH 110 ff., esp. 113–14. Going one better than Aristotle, Glotz and Cohen (HG I 437–8, with 398) assert that Solon merely added prokrisis by the four tribes to ‘le tirage au sort des magistrats, vieux mode de nomination qu’il conserva scrupuleusement’. 69 70 HAC 324–5. EGH 113–14.

‘Sortition from Pre-selected Candidates’

93

sacrifice, or give judgement; the notion that the conduct of war is more likely to go wrong, or demands more special skill, is a later notion: the mark of its recognition is when the iterable strategos ousts the yearly-changed Polemarch, about 490 B.C. Till then the yearly change of Polemarch is the proof that, as between a large handful of competent men (a fool or coward would not be among the forty prokritoi), it is indifferent who commands. The great professional strategoi of the fourth century have a military prime of about twenty-five years: at this rate, during a great soldier’s military prime, sixth-century Athens acquiesced in twenty-five different commanders. This is sufficient to show how little they were concerned to have the one proved man, the Napoleon or Chabrias, in command.’ Now Aristotle, at any rate, did not agree at all with Wade-Gery that in the archaic period ‘any good man could command in war, just as any good man could govern’. Not only does he observe in Pol. V 1309a33–b6, when discussing the qualities needed by candi’ dates for high office (ai‘ k
yriai arxa
i), that for a general the first ’ ’ requisite is experience (empeir
ia) rather than virtue (aret
Z), since ’ military ability is a rarer quality than moral goodness (epie
ikeia); in Ath. Pol. 3.2, rightly or wrongly, he attributes the institution of the office of polemarch in archaic Athens to the fact that some of the Basileis were feeble in the military sphere (t a pol
emia malako
yB)! Even if military skill was relatively less important in the sixth century than in the fourth,71 exceptional bravery was a quality just as desirable in a commander in the early period as it was later, for hoplite generals commonly led their men into battle in the most literal sense. 2. But the argument can be put much more strongly. Those who accept Ath. Pol. 8.1 should be asked to produce evidence of the choice by lot at some time of what I shall call ‘important magistrates’, that is to say, those who exercised major political powers or held high military commands, in non-democratic states, or indeed any states whatever, in ancient Greece72—for even democracies are 71 Even in the 5th century being a Strategos was likely to involve being admiral as well as general; and commanding a fleet was a more highly skilled undertaking. 72 I do not think there is much point in discussing the practice of non-Greek states, because so many factors are different that fruitful comparison is scarcely possible. I have, however, said something on this subject in the Appendix to this chapter.

94

Athenian Democratic Origins

not known to have chosen by lot those of their magistrates who were ‘important’ in my sense, in the way that fifth/fourth-century Strategoi and sixth-century archons were and fifth/fourthcentury archons were not. As far as I am aware, no such evidence exists, except perhaps for one city: Heraea, mentioned by Aristotle in Pol. V 1303a13-16, where the implication may be that it was the more important magistracies to which sortition was applied—and here, I may point out, the consequence of introducing sortition was a change of constitution, evidently not at all the result desired! The Rhetorica ad Alexandrum 273 has actually been quoted by Scho¨mann and Headlam74 as evidence that the lot was used even in oligarchies. In fact the passage in question does not describe an existing situation but gives advice about the practice that ought to be followed by oligarchs,75 and it distinguishes carefully between two kinds of magistracies: t aB men ple
istaB klZrvt
aB, t aB de ’ Zwfiv meu’ o
‘ rkvn kai ple
istZB akribe
iaB meg
istaB kryptfi Z& c
diacZwist
aB [‘the majority of them appointed by lot, but the most important ones voted on by secret ballot, under oath and with the greatest precautions’]. A little earlier in the same work76 ’ aB we find the statement that in democracies de&i . . . t aB mikr aB arx ’ kai t aB poll aB klZrvt aB poie&in
astas
iaston g ar to &yto
t aB ’ o to &y pl
de meg
istaB xeirotonZt aB ap ZuoyB [‘the majority of the magistracies, being minor, ought to be appointed by lot, for in this way faction is avoided; but the most important ones ought to be voted on by the masses’]. 3. Those who believe in the sortition of Athenian magistrates in early times usually refer with respect to the writings of Fustel de Coulanges, who of course wrote before the rediscovery of the Ath. 73 P. 22, ed. L. Spengel and C. Hammer ¼ Anaximenes, Ars. Rhet. II 18, ed. M. Fuhrmann (1966). 74 ELA2 80 & n. 1, with Macgregor’s note, pp. 201–2, where the passage is given in full. 75 As Macgregor pointed out in his second edition of Headlam’s book, pp. 201–2. The other texts Macgregor cites in support of his statement, ‘Still, there can be no doubt that the lot was in fact used in oligarchies’, are all valueless for his purpose, except perhaps Arist., Pol. V 1303a13–16 (Heraea), on which see my remarks above. In Thuc. VIII 70.1 and Arist., Ath. Pol. 30.3, the lot is merely used to decide the order in which the men concerned are to perform their duties, and the same would be true of the Boeotian Federal Council. In Ath. Pol. 30.2 it is the minor magistrates who are to be appointed by lot, and in 30.5 (OCT) the lot merely picks out the chairman of the Council. Plut., Mor. 597A need not refer to an important official. 76 P. 21, ed. Spengel-Hammer ¼ Anaximenes, Ars. Rhet. II 14, ed. Fuhrmann.

‘Sortition from Pre-selected Candidates’

95

Pol. Fustel believed deeply in the ‘religious’ nature of sortition and saw no reason for rejecting the possibility of its use in early times. Unfortunately, he failed to take proper account of the essential distinction I have drawn above, between ‘important’ magistracies and minor ones; and his handling of the evidence was unsatisfactory. ‘The Athenians’, he says,77 ‘like many Greek peoples, saw no better way [of electing magistrates] than to draw lots . . . For the lot was not chance; it was the revelation of the divine will . . . It was believed that the gods designated the most worthy by making his name leap out of the urn. This was the opinion of Plato himself, who says,78 ‘‘He on whom the lot falls is the ruler, and is dear to the gods; and this we approve to be quite just. The officers of the temple shall be appointed by lot; in this way their election will be committed to God, who will do what is agreeable to him’’. The city believed that in this manner it received its magistrates from the gods.’ (There follows an account of the consular elections in the early Roman Republic which few historians would now accept.) The only source Fustel cites gives his case away completely. It is just as clear from the imaginary institutions set out in Plato’s Laws as it is from the actual institutions of the Greek cities that all the important magistrates, as well as many of the minor ones, are to be elected by vote. The only magistrates I can find at a quick glance through the Laws who are to be appointed by lot are indeed those in charge of the temples.79 And the sentence which follows shows the reason for this: the main desideratum for these officials is ritual purity rather than any special competence! I would not say that Fustel de Coulanges was being dishonest; but he certainly conceals from the innocent reader who does not know the Laws at first hand that even Plato, for all his religious outlook, was not prepared to entrust any important magistracies to that expression of the divine will, sortition. Nowhere does Fustel produce evidence to support his generalisations about the use of the lot in early times. 77 The Ancient City (Eng. tr. by W. Small, 1916), at pp. 182–3, from book III, ch.10 of La Cite´ Antique (1864). And see Fustel’s article, ‘Sur le tirage au sort applique´ a` la nomination des archontes athe´niens’, in Nouv. rev. hist. de droit franc¸ais et e´tr. 12 (1878) 613–43. 78 Laws VI 759; cf. III 690. 79 Laws VI 759BC. The lot is also to be used in the choice of boyleyta
i [‘councillors’], but only to reduce 720 elected candidates to the final number of 360: see VI 756E.

96

Athenian Democratic Origins

4. Wade-Gery80 can do no better. He refers to Laws 757B, E81 as ‘a religious man’s statement’, which ‘shows how completely sortition had lost its religious authority in the fourth century and become a weapon of democracy’. (It is a pity he did not quote the sentence in Laws 757E which immediately follows the one he cites on p. 113 n. 1: both kinds of equality must necessarily be used, but we should use as little as possible that which depends on chance82— i.e. the lot.) The question-begging phrase ‘had lost its religious authority’ conceals the fact that no evidence has been produced to prove that sortition was ever believed to have such authority in really important political or military matters. Nor does WadeGery bring forward any evidence which might explain this curious ‘loss of religious authority’. A recognition that the lot could be used in Greece in religious contexts in early times does not at all justify the conclusion that whenever it was used it was believed to have religious authority. Indeed, Wade-Gery can find only one single phrase to quote in support of his argument for the alleged ‘religious authority’ of sortition in Solon’s time, namely Iliad VII 170–80, where the lot had to be used to choose between a number of arrogant warriors none of whom would have been prepared to concede that any of the others was a better man than himself. Wade-Gery has to concede that the employment of the lot in this case was ‘partly perhaps to ‘‘avoid ill-feeling’’ ’; but he goes on to claim that it was ‘most in order that Zeus may choose his own man’—an inference for which there is no justification at all in Homer. Similarly, another distinguished believer in the use of sortition in early times has said (in an unpublished paper which I have been privileged to see), ‘The onlookers knew which champion they wanted, and prayed to Zeus to direct the lot aright; and this illustrates what is often and rightly emphasised, that there is a religious element in sortition.’83 It illustrates nothing of the sort. To conclude, from the fact that the onlookers prayed to Zeus, that the choice had in any way been committed to Zeus or was regarded as his decision is a non sequitur. Zeus could do almost anything, and prayers were addressed to him on numerous occasions: it was quite natural that the spectators 80

81 EGH 112–14. Cf. III 690C; V 741B; VI 759BC. Cf. Xen., Cyrop. I vi 46. 83 [Ste. Croix does not identify the ‘distinguished believer’. RP believes that in an Oxford lecture he heard this view ascribed to A. R. W. Harrison.] 82

‘Sortition from Pre-selected Candidates’

97

should pray to him to influence the way the lot fell—just as they prayed to him immediately afterwards, at Ajax’s request, that their champion should be victorious.84 But there is not the least suggestion either that entrusting the choice to the lot was a deliberate commission of it to the divine will, or that the decision of the lot was accepted as divine. Exactly the same situation is found in Iliad III 314–25, where lots are cast at the single combat of Paris and Menelaus, to decide which is to cast his spear first: it is again the watching hosts who pray to Zeus, and their prayer is that the guilty man should perish and peace prevail. And in any event, as WadeGery admits, ‘Things are done in poetry which are not in real life.’ Moreover, the Greek host at Troy, a loose alliance under the precarious suzerainty of a chief admittedly only primus inter pares, does not provide the best of comparisons with a Greek city administering its internal affairs. The vain appeal to Il. VII 170–80 only shows how hard put to it are the defenders of Ath. Pol. 8.1 to find any scrap of evidence to support their belief in the ‘religious authority’ of sortition in early times. Headlam,85 who did not doubt the religious origin of the drawing of lots, had to admit that ‘at Athens where the use of the lot was most common the evidence for its religious significance is smallest . . . In the Tragedians mention is often made of the lot; and occasionally in such words that we are reminded it had a religious origin. But this is the only case where it is mentioned in direct connection with the services of some temple.86 In no other cases is it spoken of as religious, and never do we find attaching to its use the awe and mystery which belongs to other more impressive means of divine utterance.’ 5. It may be true, as Wade-Gery says, that ‘any good man could command in war, . . . govern, or sacrifice, or give judgement’; but some could clearly do this very much better than others, and the community would ordinarily want to choose the best of them, subject only, in the case of aristocracies, to the general principle (visible also in the Roman Republic) that iteration must be forbidden or drastically restricted, so as to permit rotation of office around the circle of the nobility—a rule often found in aristocratic 84 85 86

II. VII 190–6, 200–6. ELA2 8–9. See the whole passage, pp. 4–12. Headlam cites Eur., Ion 416; Aesch., Eum. 32.

98

Athenian Democratic Origins

societies, where it may help to prevent the disruption of the constitution and social order by men greedy for personal power. Such an arrangement may lead to a loss of military efficiency and yet be felt to be an overriding necessity. It certainly cannot be claimed that non-iteration of archonships in Solonian Athens87 proves the Athenians were ‘little concerned’ about the capability of their potential magistrates and may therefore have been willing to choose them partly by lot: non-iteration was a common overriding aristocratic principle, and no proof at all of indifference as to who should hold office. This invalidates Wade-Gery’s argument quoted above (p. 92). The Roman consuls also changed yearly, and the Romans, because of their restrictions on the iteration of office, might ‘acquiesce in quite a number of different commanders during a great soldier’s military prime’; but that is very far from implying that ‘it is indifferent who commands’: the Romans elected those who appeared to them (sometimes on what may seem to us very inadequate grounds) the best men available in each year, outside the number of those disqualified by the leges annales, and allowed the lot to decide only such matters as the sharing out of a fixed number of provincial governorships among an equal number of qualified men. In many aristocratic societies it has been a common belief that any member of a noble family has an inherited capacity for ruling and even for commanding in war. At lower levels this is particularly true, and sometimes it has been applicable to the highest posts. Lord Howard of Effingham was chosen by Queen Elizabeth I to be Lord Admiral of the fleet that was to oppose the Spanish Armada ‘less because of any demonstrated fitness for command at sea than because he came of an illustrious line—three of his house having served the Tudors as Lord Admirals—and because he was an ardent Protestant of unquestionable loyalty’.88 Similarly, in the appointment of the Duke of Medina Sidonia to command the Armada ‘what certainly counted most was that he was head of the house of Guzman el Bueno, one of the most ancient and illustrious 87 I suppose it is not necessarily true a priori, just as it is certainly not demonstrable, that even in Solon’s time election to any of the nine archonships permanently disqualified a man from holding the same or any other archonship; but this seems to me very probable indeed, and I think it is what most historians believe. Cf. pp. 225–7 below. 88 Garrett Mattingly, The Defeat of the Spanish Armada (Pelican, 1962), 215.

‘Sortition from Pre-selected Candidates’

99

in Castile, a grandee of such dazzling eminence that no officer in the fleet could feel insulted by his promotion, or find it beneath his dignity to obey him’.89 Certainly in aristocratic societies technical competence for the job is often far from the decisive qualification for an office, even a military one. Nevertheless, it is a far cry from sixth-century Greece to sixteenth-century Europe and similar societies for which examples of the same sort could be given.90 6. And there is another weighty argument. I doubt if those who uphold the possibility of the election of important magistrates by lot have sufficiently realised that the choice of an Athenian polemarch, for instance, by lot before 487 would not merely represent putting to the hazard of the lot the choice of one man from the forty best possible commanders, even if we may assume that the best possible candidates came forward year after year until they were elected to an archonship. At any given moment it was likely that most of the best qualified men would have been elected already as polemarch or some other archon and would therefore be ineligible—and this makes all the difference. Choosing by lot among the forty best men would be foolish enough, but perhaps not utterly inconceivable. Choosing by lot among forty men when probably a high proportion of the best forty men are disqualified by having held an archonship before would be insane. 7. Again, if those who are superintending an election are not disinterested, the election can be ‘fixed’, with greater or less ease according to the mode of an election concerned. Now most modern elections are by secret ballot; but in ancient Athens elections not made by lot seem always to have been open, conducted in the way indicated by their name, xeiroton
ia [literally, ‘stretching out of hands’]. And open elections of this kind are obviously very much harder for dishonest magistrates to ‘fix’ than elections by ballot, because deliberate miscounting becomes much harder. Choice by lot, on the other hand, is absolutely dependent upon the honesty of those who draw the lots. No type of election is easier to ‘fix’ than sortition.91 Now Solon must have known very well that 89

90 Ibid. 221–2. Cf. the Appendix to this chapter. & Plato, Rep. V 460A, with the euphemistic phrase klZroi d
e tineB oi
mai poiZt
eoi komco
i [‘I suppose we will have to devise some subtle method of allotment’], contemplates a wholesale ‘fixing’ of the lot. And see Arist., Ath. Pol. 62.1, for the ‘sale’ of election by lot in the Athenian demes. For the draw ‘a mano’ in Florence under Cosimo de’ Medici, see the Appendix to this chapter. 91

100

Athenian Democratic Origins

administration of the elections would be for many years entirely or mainly in the hands of eupatrids, who were all too likely to be reluctant to admit ‘novi homines’ to the archonship. Can we really believe he would have been foolish enough to create an institution which gave the opponents of his new timocracy the best possible opportunity of sabotage? 8. A further argument against sortition of archons in the manner described by Ath. Pol. 8.1 is that the offices to be filled so indiscriminately were very different in character. The eponymous archon seems to have been much more important politically than the others. There was one, and only one, military post, but that was the highest of all. The Thesmothetai, who had judicial functions, must have been smaller fry. The Basileus by now acted mainly in the religious and judicial spheres; but—an important point—he alone was subject to a very ancient rule92 that his wife must have been a virgin at the time of her marriage. When they were electing their prokritoi, the men of each tribe could not have known which office each man would fill. Is it seriously contended that this is an arrangement worthy of the good sense of a Solon? If he had really wanted to make good use of sortition, was not the obvious course to have separate elections for archon, polemarch, Basileus and Thesmothetai? (In the developed democracy, with its tribal rotation, this was what happened in practice: the office each candidate would fill if the lot fell upon him was decided beforehand by the fact that he was a member of a particular tribe.) When those who are determined to justify Solonian kl
ZrvsiB ’ prokr
itvn are looking for parallels, they had better try to find ek something analogous to the peculiar situation I have just described, for we shall hardly be content with an example of choice by lot to a single post, or a college all members of which are on much the same footing. 9. As far as I can see, there are two arguments, and two only, in ’ favour of accepting the institution by Solon of kl
ZrvsiB ek prokr
itvn of archons, apart from the fact that Aristotle asserts it in Ath. Pol. 8.1. (a) The ‘ten archons’ (five Eupatridai, three Agroikoi and two Demiourgoi) who, according to Ath. Pol. 13.2 functioned in the year after Damasias, probably 580/79 (or the ten months which were left of it), have always been a puzzle. One 92

See Ps.-Dem. LIX 74–6.

‘Sortition from Pre-selected Candidates’

101

’ prokr
itvn, Cavaignac,93 has believer in Solonian kl
ZrvsiB ek suggested that Aristotle’s ten ‘archons’ are really the ten prokritoi of each tribe, who had to be chosen in the proportions indicated. This certainly makes better sense than any alternative explanation so far offered—although, as Wade-Gery has observed,94 ‘This is not indeed what Aristotle says, nor do I see any way of emending him (should one want to) into such a meaning.’ And (b) the use of the lot was undoubtedly recognised in antiquity, at any rate in the fourth century, as a useful expedient for allaying discontent or avoiding election intrigues and disturbances.95 Neither of these arguments, it seems to me, has nearly sufficient weight to counterbalance those I have developed above. 10. I shall conclude by discussing the use by Isocrates of the word prokr
inein, since it has been generally—but wrongly—believed that in certain passages he is referring to the practice of ’ prokr
itvn described in Ath. Pol. 8.1. Even Hignett kl
ZrvsiB ek says,96, ‘It is true that Isokrates in one passage (vii.22) regards ’ prokr
itvn as a characteristic feature of ‘‘the demockl
ZrvsiB ek racy established by Solon and restored by Kleisthenes’’ ’, although he then continues ‘but elsewhere (xii.145, cf. 130 and 148) he ascribes it to a mythical period of good government which is & S
presumed to have endured from Theseus m
exri tZB olvnoB men ‘ Zlik
iaB Peisistr
atoy de dynaste
iaB [‘until the time of Solon and the rule of Peisistratus’’].’ Hignett adds, with some justice, ‘The existence of the second passage . . . obviously discounts the evidence of the first. Probably Isokrates had no warrant for either statement.’ I myself have no doubt at all that Isocrates is employing prokr
inein, even in VII 22–3 and XII 145, not in the rare sense 93 ‘La de´signation des archontes athe´niens jusqu’en 487’, in Rev. de philol., n.s. 48 (1924) 144–8, at pp. 145–6. 94 EGH 103. Wade-Gery thinks (with Cavaignac) that ‘the compromise may have existed a good number of years’. Aristotle speaks of it as lasting for one year only. 95 See Plato, Laws VI 757E (dyskol
iaB t vn & poll vn & e
‘ neka [‘on account of the ’ ‘ &ynto masses’ discontent’]); Arist., Pol. V 1303a13–16 (di
a te t aB eriue
iaB . . . o‘
tifi Zro ’ toyB eriueyom
enoyB [‘due to electoral intrigue . . . for they used to elect candidates who canvassed for office’]—but see p. 99 above); Rhet. ad Alex., as quoted on p. 94 ’ & n. 76 above (astas
iaston g ar to &yto [‘for in this way faction is avoided’]— ’ but here it is only the mikrai arxa
i [‘minor magistracies’] for which the lot is recommended). 96 HAC 323.

102

Athenian Democratic Origins

of ‘make a preliminary selection’ (see LSJ9, s.v. prokr
inv, I b) but in the very much more usual sense of ‘choose before others, prefer, select’. The word is used by many other writers from Thucydides (IV 80.4) onwards in this way—see LSJ9, ibid., I a; and we may add two examples from the fourth century in which it certainly refers to outright election: Isae. VIII 19, where the wives of the members of a particular deme proy
’ krinan the speaker’s mother to preside at the Thesmophoria, and Dem. XXIII 197, where the Athenians proy
’ krinon as leaders Themistocles and Miltiades, as Strategoi of course. Isocrates employs the word prokr
inein some 17 times altogether. In 14 cases (III 26; IV 4; V 113; IX 13, 35, 74; X 46; XI 11; XII 114, 204; XIV 12; XV 120, 250, 302) no one will wish to pretend that it bears anything but the normal meaning, and in the remaining three cases (VII 22 and 23 and XII 145) there is not the least reason to suppose that the word is used in any other way. In VII 22–3, part of a disquisition on the ancestral democracy of Solon and Cleisthenes (§§ 16, 20), Isocrates is drawing a plain ’ contrast between election by vote and choice by lot: kl
ZrvsiB ek prokr
itvn is simply not there at all. In § 22 the distinction is ’ ap
‘ antvn t between (a) choosing by lot from all the citizens (ej aB ’ aB klZro &ynteB), and (b) electing (prokr
inonteB) the best and arx most suitable men for each job. It is true that in this section, considered in isolation, one could make a case for taking prokr
inonteB in the technical sense (‘making a preliminary selec’ tion by vote’), by emphasising in the first part of the sentence not ej ‘ antvn . . . klZro &ynteB but only ej ’ ap
‘ antvn, so that (a) becomes ap
‘choosing by lot from all the citizens’, and (b) becomes ‘making a preliminary selection of the best and most suitable men for each job’. Section 23, however, clinches the matter in favour of taking both prokr
inonteB and prokr
inein in the more usual non-technical sense. Isocrates here draws two successive sets of distinctions: first between (a) this arrangement (ta
ytZn t Zn kat
astasin) i.e. prokrisis of the best men, and (b) choice by lot (t Zn di a to &y lagx
anein gignom
enZn [sc. kat
astasin]), not explicitly by lot ’ tfi Z& ‘from among all’; and secondly between (a) choice by lot (en klZr
vsei), under which the Athenians realised that chance would decide, and (b) prokrisis of the best men, under which they realised that the people would have power to elect those who were most devoted to the existing constitution. At no point is there the least

‘Sortition from Pre-selected Candidates’

103

suggestion that the prokrisis would be followed by a sortition. All this shows beyond question that prokr
inonteB (§ 22) and prokr
inein (§ 23) mean not ‘make a preliminary selection by vote’, but ‘choose by vote in preference to others’, i.e. ‘elect by vote’. This is what we should anyway have expected from Isocrates, who rarely uses technical terms. In the remaining passage, XII 145, Isocrates is professedly speaking about the early constitution of Athens over a long period, from about the time of Theseus to that of Peisistratus (see §§ 119– 20, 126–9, 130–1, 134, 144, 148). Here we read that in the good old ’ i t ’ aB toyB prokriu
entaB times the Athenians kau
istasan ep aB arx ‘ o tvn yp & wyletvn & kai dZmotvn & [‘they appointed to offices those who were selected by their fellow tribesmen and demesmen’]. The imprecise word kau
istasan is unfortunate but again there is no hint of a subsequent sortition (kl
ZrvsiB) from the prokriu
enteB. I suggest that although one would have expected prokriu
entaB in this particular case to bear the technical sense of a preliminary selection (because the choices are made by tribes and demes), Isocrates is still thinking mainly in terms of the election in preference to others of the best man. He will have had in mind the offices ’ (arxa
i) in the widest sense, including the Council, in the elections to which he would no doubt have preferred xeiroton
ia [‘voting by show of hands’] to the sortition of his own day, still made by demes (Arist., Ath. Pol. 62.1); and he was probably thinking above all of the contrast between the contemporary choice of councillors by lot in the demes and the good old system under which ‘the best men’ could be chosen by vote. *** ’ prokr
itvn is very strong The case against Solonian kl
ZrvsiB ek indeed. It may seem virtually conclusive when we recall that against the one explicit testimony in favour of this institution, that of Aristotle in the Ath. Pol., we can bring up another voice of at least equal power: the Aristotle of the Politics.97 As I shall maintain in a later essay in the book,98 this is one of three major examples of conflict between Ath. Pol. and Politics; and in deciding between Aristotle’s two views in this case it is a material consideration that in both the other two examples it is undoubtedly the 97

See pp. 273–7 below.

98

Pp. 273–7 below.

104

Athenian Democratic Origins

Ath. Pol. which is to be rejected. The Ath. Pol. may well represent in each case Aristotle’s later view; but if so, his opinion had been changed for the worse, probably either by some bogus document which deceived him, or by a false statement made by a literary source he respected. In case it is argued that Ath. Pol. 8.1 may come ultimately from Solon’s Axones, I would retort that if Aristotle could be wrong, as I have shown he must be, about the qualifications of the census classes, he may equally have been wrong about the method of election of magistrates: both these subjects would form part of what in the later fourth century was called Solon’s politeia, as distinct from his nomoi, and I shall argue in a subsequent essay99 that much of Solon’s politeia in this fourth-century sense might not appear among the nomoi recorded on the Axones. If I seem to have devoted a disproportionate amount of space to refuting Ath. Pol. 8.1, it is because one of the aims of these essays is to demonstrate that the Athenians were not the quaint and rather silly people so often pictured in modern books, who devised eccentric political institutions: census classes fixed on a highly unsuitable basis, ostracism as a device for getting rid of potential tyrants (a purpose for which it was quite unsuited), and election partly by lot of their principal magistrates, first under Solon and ’ prokr
itvn was introduced for the selecagain in 487. kl
ZrvsiB ek tion of archons in 487; but by then, as I shall show, the archons had already ceased to have political and military importance.

APPENDIX Election to High Office by Lot Elsewhere than in Greece Election by lot in all kinds of different forms is found in the Italian cities of the later Middle Ages and Renaissance and in the early Swiss democracies. I shall not discuss the Swiss republics, in which I believe the lot was rarely if ever used for the election of magistrates of comparable importance with the Athenian eponymous archon and polemarch. In the Italian states the lot was used in many different ways. Some of them 99

Pp. 310–17 below.

Appendix

105

bear no comparison at all with Athenian kl
ZrvsiB: at Venice, for example, sortition entered into the exceedingly elaborate procedure used for electing the Doge from 1268 onwards; but it was only the electors, not the Doge himself, who were chosen partly by sortition. Although the magistrates themselves were actually chosen by lot to some extent in more than one city, there is little point in multiplying examples, and we may content ourselves with the one instance which certainly looks the most promising at first sight: the choice at Florence100 from 1328 onwards of the priori delle arti (six in number until 1343 and thereafter eight), who, with the gonfaloniere di giustizia, made up the signoria which was at once ‘the town council and the government of the State’.101 By a process known as imborsamento the names of the candidates, who were drawn in fixed proportions from the arti maggiori, medie and minori, were put into leather bags, borse, and lots were then drawn, the whole procedure being hedged about with extraordinary safeguards. After various changes, Cosimo de’ Medici and his faction in 1434 procured the appointment of a committee of ten accoppiatori, who were entrusted with the task of choosing the new priors from the names in the borse, still in theory by lot: the resultant drawing a mano, as it was derisively called, made a travesty of the whole procedure of sortition. However, ‘while the accoppiatori represent the finishing touch applied to an older system of control, it is absurd to maintain that only with their appointment was the Florentine constitution denatured.’102 In fact the mercantile oligarchy, or rather series of oligarchies, had always maintained a large measure of control over the elections, preceded as they were by a complicated scrutinio designed partly, through the divieto, to prevent any individual or family obtaining frequent re-election to high office, and partly (to a much greater extent than constitutional theory divulged) to exclude opponents of the regime. This qualifying procedure was tightened up in 1387, when the borsellino was also introduced, a special small borsa 100 Here I have found most helpful Ferdinand Schevill, Hist. of Florence (1936), esp. 209–10, 338–40, 355–6, 372; G. A. Brucker, Florentine Politics and Society 1343–1378 (Princeton, 1962), esp. 64–71, 76; C. S. Gutkind, Cosimo de’ Medici (1938), esp. 108–10, 124 ff., and 292–301 (Appendix 8: ‘The Florentine Constitution’); R. Davidsohn, Gesch. von Florenz III (1912) 862–5; IV (Die Fru¨hzeit der Florentiner Kultur) i (1922) 54–204 (‘Verfassung u. Verwaltung’), esp. 95–100, and ¨ mterauslosung’. see Index, s.v. ‘A 101 Gutkind, op. cit. 294. [Professor G. A. Holmes comments that Gutkind here exaggerates the power of the priors. ‘Their real powers were that they largely controlled diplomatic relations with other states and, internally, were able to propose legislation, but not to pass it. See J. M. Najemy, Corporation and Consensens in Florentine Electoral Politics 1280–1400 (1982) and N. Rubinstein, The Government of Florence under the Medici (1434 to 1494), ed. 2 (1997).’] 102 Schevill, op. cit. 355–6.

106

Athenian Democratic Origins

containing the names of faithful adherents of the oligarchy, out of whom two of the eight priors were now selected, and again in 1393, when the reintroduction in a new form of the ammonizione (formerly exercised by the Parte Guelfa) made it easier for the oligarchs to disqualify for office any suspected opponents on the pretence that they were Ghibellines—anathema at nearly all times in Florence. The Florentine system of electing priors proved itself a bad one and has been generally condemned by modern historians. It ‘placed little premium upon political ability or statesmanship’, says Brucker,103 ‘but tended instead to exalt mediocrity.’ Schevill’s verdict is harsher still. The constitution of Florence, he says,104 was ‘more or less flagrantly manhandled from its birth and indisputably from the time of the adoption of the practice of choosing the signory by lot. From the very start of this system an evil inseparable from a resort to chance as a method of appointing magistrates had made itself felt. This was the election to office of a high percentage of the flagrantly incompetent. If to this we add the perilous flux imposed on the executive department by the brief two months’ term of office, we shall be disposed to agree that anything approaching a considered and consistent general policy had become impossible . . . A political system, so offensive to the most rudimentary demands of common sense as this of Florence, cannot be maintained and . . . , if it cannot by legal means be adjusted to the most immediate needs of society, a way will unfailingly be found to circumvent it.’ Those who wish to attribute to Solon, as the inventor of kl
ZrvsiB ’ prokr
itvn, a merely equal degree of political folly must be told, ek however, that the complicated and intricate Florentine system of mixed nomination, election and sortition of priors bears in reality only a superficial similarity to the Athenian institution which appears to be analogous to it. Even apart from the features mentioned above, a whole series of factors made the position of the Florentine priors relatively much less important and responsible than that of the Athenian archons. To begin with, two major spheres of activity were entirely withdrawn from them: they were not concerned at all with jurisdiction, which was the business of the podesta`, nor did they hold any military command. They served for only two months at a time. Unlike the sixth-century Athenian archons, who differed very much among themselves in powers and functions, they were all theoretically equal and always acted together as a college, so that individual initiative was not required. ‘The priorate was a commission form of government and every conceivable difficulty was put in the way of any one prior exercising a greater authority than his fellows.’105 By the fourteenth 103

Brucker, op. cit. 70.

104

Schevill, loc. cit.

105

Schevill, op. cit. 210–11.

Afterword

107

century in Florence, again unlike ancient Athens, ‘the daily routine of government was to a considerable degree in the hands of professional civil servants. The activities of these men gave to the administration some degree of continuity and stability which compensated in part for the weakness inherent in a system of amateur, part-time government.’106 And then there were the guilds, which regulated so many features of the daily life of the citizen and exercised a great influence on state policy: this was a feature of the Italian cities which had no parallel in Greece. Finally, although there was nothing in Florence corresponding closely to the Athenian Assembly (for although a parlamento had full theoretical sovereignty, the government was under no obligation to summon one, and only did so in order to lend an air of constitutional respectability to a coup d’ e´tat), the liberty of decision of the priors was nevertheless far smaller in practice than in pure constitutional theory: important debates took place in the consulte e pratiche (records of which survive from the late thirteenth century, and in considerable numbers from the fourteenth and fifteenth), to the results of which the priors had to pay great attention. ‘These debates were advisory only, and technically the Signoria could ignore the views of these councils. But rarely, if ever, did the priors flout the consensus of opinion expressed in these consultative assemblies. This procedure became part of the unwritten code of the city’s political life.’107 It is very doubtful if even Areopagus and Council together provided the same kind of guidance to the Athenian archons. Although sortition entered into the selection of priors in late mediaeval Florence, therefore, it was very far from assuming the decisive ’ position it would have held if applied to the choice of archons (even ek prokr
itvn) in the sixth-century Athenian constitution.

AFTERWORD 1. Ste. Croix does not discuss the mechanics of voting procedures. Hansen has now made it very plausible that, at an ordinary vote by ’ show of hands (epixeiroton
ia) in the Assembly, the presiding magistrates assessed which side had the majority, and votes were not in fact counted. In the courts, and for votes in the Assembly on issues requiring a quorum of 6,000, voting was by pebbles which were counted. Such was the position under the developed democracy, but Ste. Croix could 106 107

Brucker, op. cit. 60. Brucker, op. cit. 76.

108

Athenian Democratic Origins

have rescued his case by the argument that ‘the decisions of the Assembly were called psephismata and the procedure itself often psephizesthai—both words are from psephos (‘pebble’), which shows that at some earlier time the Assembly had voted in the same way as the courts’ (Hansen, AD, 147 with reference to Staveley, 84–6). Note, however, the argument of Spivey that (mythological) scenes showing voting with pebbles first appear on Attic vases, in a cluster, early in the 5th c. 2. Rhodes, CAAP 121–2, and Wallace argue for the same conclusion as Ste. Croix. But Develin (in LCM, and Athenian Officials, 37) seeks to revive the position of Hammond briefly dismissed by Ste. Croix (note 8). 3. The existence of a Eupatrid order which dominated archaic Athens has become controversial in recent years: see e.g. Welwei, Athen, 108–9; J. K. Davies, OCD3, s.v. Eupatridai; Parker, AR 63–5. 4. Rhodes, CAAP 153, briefly defends the Solonian boule (see too AB, index s.v. Solon); it is also accepted by Welwei, Athen, 190–2, and others. 5. Hansen, AD 49–52 (‘Selection of magistrates by lot’) is very close to Ste. Croix in spirit and conclusions. The opposite view is defended by Rhodes, CAAP 146–8. Develin, R., ‘Solon’s First Laws’, LCM 9 (1984) 155–6 —— Athenian Officials 684–321 B.C. (1989) Hansen, M. H., The Athenian Ecclesia (1983), 103–21 (from GRBS 18 (1977) 123–37) Spivey, N., ‘Psephological heroes’, in Osborne/Hornblower, RFP 39–52 Wallace, R. W., ‘The Date of Solon’s Reforms’, AJAH 8 (1983) 81–95

3 Solon, the Horoi and the Hektemoroi

The projected chapter on this subject was never written. What follows derives from two long letters sent by Ste. Croix to his friend Antony Andrewes. By 1962 Andrewes had developed a new theory of the famous Seisachtheia or ‘shaking off of burdens’ of Solon: according to Andrewes, what Aristotle saw as a cancellation of debts was not that, but the abolition of a traditional and long-standing state of dependency, that of the Hektemoroi or ‘sixth-part payers’. The theory is briefly presented by Forrest, EGD, 147–50 and by Andrewes himself in Greek Society (1971 [first published as The Greeks, 1967]), 115–19; a modified version appears in Andrewes’s chapter in The Cambridge Ancient History, ed. 2, III.3 (1982), 377–82. In 1962 Ste. Croix was invited by Andrewes to comment on a preliminary draft and sent a long letter in reply; in 1968 Andrewes again consulted Ste. Croix and received a further long response. What follows is an edited extract from the 1962 letter, followed by that of 1968. Andrewes modified his position substantially in response to the criticisms of Ste. Croix and others; portions of the 1962 letter relating to claims which Andrewes later abandoned have been omitted where this could be done without losing continuity.

THE 1962 LETTER A. Deciding that the Hektemoroi were share-croppers who paid one-sixth of their produce to their landlords (or lords), you regard the very system itself as the main evil: you call it ‘a very deadening form of servitude’, and you say that Solon ‘eventually abolished the system under which they [the Hektemoroi] suffered’ by ‘a once-for-all administrative act’: this was the Seisachtheia, later misinterpreted by Aristotle and everyone else (except Androtion, who was even farther from the truth) as a cancellation of debts. The abolition of enslavement for debt is for you mere ‘consequential legislation’.

110

Athenian Democratic Origins

B. Before I begin in earnest, I do want to emphasise what an utterly disastrous effect your theory would have, if it were true, upon our confidence in both our main authorities, Aristotle and Plutarch—we need not bother about Androtion, whom no one now takes seriously. The part of the theory with which I am now concerned is that the Seisachtheia was mainly (perhaps entirely) concerned with the Hektemoroi, and that the Hektemoroi were not debtors. As far as Aristotle is concerned, you ask us to believe that Aristotle was obsessed with debt, and ‘forgot the Hektemoroi once he had mentioned them in his introductory sketch’. So Aristotle ‘forgot’ the really essential factor in the situation! And this curious lapse of memory led (on your theory) to Aristotle’s giving a totally false picture of the Seisachtheia. This, if true, would be exceedingly discreditable to Aristotle, who on numerous occasions presents the cancellation of debts as the essential feature of Solon’s reform. On your interpretation, one is forced to conceive Aristotle as an inexcusably bad and careless historian, who did not (as many of us believe) just have incomplete sources, but could not even make use of the good ones he had, and by ‘forgetting’ the most important element in the situation, produced an entirely false picture of Solon’s first (Ath. Pol. 10.1) and most important reform. I believe that in general I have a good deal less trust than you have in the Ath. Pol.; but I should find it very hard to believe that Aristotle could ‘forget’ the very people whom he knew to be the very centre of the whole affair—and if I could, I think I should then regard the Ath. Pol., apart from the quotations of Solon, as an almost worthless source for the period, because how could one tell what else Aristotle had forgotten and misinterpreted? Plutarch, in Solon 13–16, seems (on your interpretation) to make much the same sort of mistake as Aristotle. First he ‘distinguishes between Hektemoroi and debtors’, and he does this ‘firmly’ and ‘sharply’. Then, like Aristotle, he entirely forgets about the distinction, presents the Seisachtheia as a cancellation of debts, and so on. Is it not all highly implausible? Surely it would be better to look for an explanation which does not make both Aristotle and Plutarch contradict themselves and ‘forget’ the essential feature of the situation they are dealing with? C. 1. The misconception at the root of your theory, as I see it, is the drawing of a rigid, artificial and unnecessary distinction—

Horoi and Hektemoroi

111

which exists neither in the sources for Solon nor in real life— between (a) owing rent (i.e. what the share-croppers paid) and (b) owing a debt. In fact, of course, owing rent is one of the commonest forms of debt, and most legal systems will treat unpaid rent much the same (sometimes exactly the same) as any other kind of debt which is due and unpaid. They may give a man better remedies for recovering unpaid rent than other debts, as English law used to, and may still; in my days as a lawyer, you could ‘distrain’, levy a ‘distress’, for unpaid rent without the leave of the court, whereas you could not levy ‘execution’ for an ordinary debt without getting judgement in court first. On the other hand, it is conceivable that another legal system might put a landlord in a worse position than an ordinary creditor, especially, for instance, if he had unrestricted rights of ejection—although if that were the situation, the law would be likely to be evaded: a desperate tenant would often be willing to accept a loan from his landlord rather than be evicted, and then of course when the debt became due the landlord could put the harsher procedure into force against him. 2. In agricultural tenancies, rent and debt often become inextricably mingled. I will quote just one statement, by I. Mendelsohn, Slavery in the Ancient Near East (1949), 109. The great landowners, including the temples and kings, he says, instead of cultivating directly with slaves, ‘preferred to lease parcels of their land to tenant-farmers’ (including in the Neo-Babylonian period even slaves), who ‘received seed, animals and implements for the cultivation of the land, mostly in the form of loans from the landlords, who in turn received a definite ratio of the produce at the end of the harvest’. (These loans were usually ‘non-interestbearing’, but if they were not punctually repaid they often then became subject to tremendously high rates of interest—sometimes well over 100% p.a.) So, for instance, it is very common in the Old Babylonian period to find ‘an advance of seed-corn or of money bearing interest to buy it and payable out of the proceeds of the crop at harvest-time’ (G. R. Driver and J. C. Miles, The Babylonian Laws I (1955), 144 and n. 2, on the Code of Hammurabi sec. 48). Such an advance could well be called in Greek a daneism
oB. ’ i (6.1; Aristotle in the Ath. Pol. uses the expressions dane
izein ep
’ i (2.2; 4.5) in the sense of ‘(to create) debts at 9.1) and daneismoi ep
interest on the security of’ rather than, more narrowly ‘(to make) loans at interest on the security of’: I do not find any difficulty in

112

Athenian Democratic Origins

this, but if you do, you can apply the Babylonian analogy I have just given—such practices are widespread and very probably occurred in seventh-century Attica, either for overdue rent or because of non-payment of loans, and often the two will be indistinguishable. Mendelsohn (p. 111) goes on to speak of the situation in Syria and Palestine ‘in the El-Amarna and in the Israelite periods’. ‘We have evidence of the existence of a large class of landless people’, he says, of whom ‘the majority remained in their agricultural communities and became tenants on the land that had formerly been theirs . . . the dispossessed peasants remained on the land as tenants; hardworking and povertystricken share-croppers they were, but not slaves.’ Indeed, ‘the basis of Near Eastern society was the free tenant-farmer and sharecropper’ (pp. 121–2). 3. This brings me to my next point. Although these sharecroppers are not at first slaves, they may become slaves: right down to modern times, most legal systems have given creditors power to take very strong action against defaulting debtors, amounting sometimes to outright enslavement, sometimes to a condition of temporary ‘bondage’, approaching enslavement while it lasts; and any such powers will of course always extend to unpaid rent, as one of the most characteristic forms of debt. In the course of an informative discussion of the origins of slavery, the Finnish anthropologist G. Landtman (The Origin of the Inequality of the Social Classes (1938), 239) rightly describes it as ‘a very general rule that pawned debtors [as he calls them] find it impossible ever to be redeemed’. In the Near East, unlike the Greek and Roman worlds, by far the most important source of slavery in PreClassical times was the enslavement of defaulting debtors, or socalled ‘voluntary’ self-sale or the sale of children by debtors; and probably many of these had been tenants who had got into arrear with their rents. In Roman law, the position of the defaulting debtor (including lessee) was always much worse than most people realise: whatever it did to nexum, the Lex Poetelia1 certainly did not abolish personal execution for debt entirely, and the man against whom his creditor had successfully exercised manus iniectio [‘laying on of the hand’] might well be described, like the debtors 1 [On the tendency of modern historians to exaggerate the effects of this law, probably of 326 B.C., see CSAGW 165–6.]

Horoi and Hektemoroi

113

’ vgimoB, ‘liable in Aristotle’s picture of pre-Solonian Athens, as ag
to seizure’. D. Surely if we take what is the natural interpretation of Aristotle and Plutarch in the light of the foregoing, we get a perfectly satisfactory picture, even if we do not have all the detail we should like. 1. Different people will no doubt always believe (as I do not) that they can divine—differently, of course!—how the population of Solon’s Attica was divided between what we should call peasant proprietors (i.e. freeholders) and share-croppers (i.e. something in the nature of leaseholders, or I suppose you could even call them ‘landless labourers’). All I will say is that it would be very rash to exclude a substantial number of either category. If Greek conceptions of ‘ownership’ were always relatively vague, as Vinogradoff2 and others have suggested, and as A. R. W. H[arrison] (I understand) agrees, they may well have been very fluid indeed in this early period. But even quite primitive legal ideas can easily distinguish the ‘ownership’ of A, the man we call a freeholder, who has for instance borrowed on mortgage and is paying interest, whether in money or produce or a share of his crop, from that of B, the man we call a leaseholder, who pays a rent; because as soon as A pays off his debt he is entitled to possession without further payment to the mortgagee, whereas B’s payments to his landlord go on indefinitely, quite apart from the fact that his landlord will probably have the right to eject him when he wants to. 2. The Hektemoroi, I agree, were probably share-croppers, paying one-sixth. (I do not think this is certain, but this letter is already so enormously long that I dare not embark on that!)3 Many of them, like so many of their Near Eastern, and later Roman and other counterparts, may have been peasant proprietors originally, and had their land filched from them or had been induced to surrender it. Of these Hektemoroi, a large proportion, perhaps virtually all, had got into arrear with their rents, or with the repayment of loans made to them by their landlords, and had ‘ oxrevB as Plutarch (Sol. 13.4) has it, thus become indebted, yp
’ vgimoi, ‘liable to seizure’ (Ath. Pol. 2.2). and were therefore ag
2

P. Vinogradoff, Outlines of Historical Jurisprudence II, The Jurisprudence of the Greek City (1922), 198 and elsewhere. 3 [The reference is to the theory expounded in the letter of 1968.]

114

Athenian Democratic Origins

3. There must also have been many ‘peasant proprietors’, as there were at all times in the Greek and Roman worlds and the Pre-Classical Near East, except of course in Egypt, where land in some sense belonged to the king. A. H. M. Jones, in a splendid article (which few people seem to know about) in Past and Present 13 (1958) 1–13 on ‘The Roman Colonate’,4 has shown that even in the sixth century A.D. there must have been very many more peasant proprietors than most historians and lawyers have realised. Our nearest ‘analogy’ for seventh-century Attica is Hesiod, and in view of his evidence we cannot discount the probability of quite a large number of peasant proprietors. Many of these men had also got into debt and became liable to seizure on default. 4. Some of each of our two classes had been sold abroad, some illegally but some quite legally (Ath. Pol. 12.4 ¼ Solon fr. 36 (West), lines 9–10)—this seems to me very significant, and worth mentioning, because it shows both that Athenian law allowed enslavement and sale abroad of debtors and also that powerful men at Athens could take the law into their own hands. Many of these Solon freed and brought home, though of course how he got the money to buy them back is an open question. Others too had fled abroad in despair (ibid., lines 10–12): these also were brought home. Whether Athenians had been legally enslaved inside Attica (ibid., lines 13–15) is doubtful, to my mind—if early Athenian law insisted on something corresponding to the Roman sale of a citizen trans Tiberim (‘beyond the Tiber’), then Solon’s doyl
iZ (literally ‘slavery’), will be used in a loose sense, for ‘a condition of bondage’. 5. The main part of the Seisachtheia is rightly represented by Aristotle and Plutarch as a cancellation of debts and a prohibition against ‘lending on the body’—as I shall explain presently, it was this latter provision which was the most important, as Aristotle and Plutarch insist—Aristotle with special emphasis in 6.1, where he ’ puts it in front of the xrevn & apokop
Z [‘cancellation of debts’], and in 9.1, where he gives it as the prvton & kai m
egiston [‘first and most important’] of the tr
ia dZmotik
vtata [‘three elements most favourable to the demos’], to the exclusion of the cancellation of debts. 4 [¼ Jones, The Roman Economy (1974), 293–307, also reprinted with addenda by D. Crawford in M. I. Finley, ed., Studies in Ancient Society (1974), 288–303.]

Horoi and Hektemoroi

115

I agree with you that the plucking up of the Horoi must have dealt with ‘the central core of Solon’s problem’—or at any rate with part of it. As regards the peasant proprietors (according to my guess, the majority of the distressed debtors), there is no problem: the Horoi were wooden pillars recording the fact that the lands they stood on were what we should call ‘mortgaged’, and of course the destruction of the Horoi accompanies and symbolises the cancellation of the mortgages. 6. But what about the Hektemoroi? You say we can be ‘virtually certain’ that the plucking up of the Horoi ‘rescued the Hektemoroi’. Why? I hope you would now agree that there can hardly have been no (or even very few) indebted peasant proprietors among the & p
enZteB=dZmoB [‘the poor/the people’], and consequently that not all of them could have been Hektemoroi. Aristotle could not have known what proportion of the poor were Hektemoroi. His kai ’ ‘ Zmoroi [‘and they were called pelatai ekalo &ynto pel
atai kai ekt
and hektemoroi’] can quite naturally be taken to refer to the substantial class of share-croppers. After that he avoids using the term again, probably because he realised that not all the debtors were Hektemoroi. Personally, I think it is rather improbable that the Horoi had anything to do with the Hektemoroi, for this reason. Horoi in the Greek world had one of two functions: to be boundary stones, or to record encumbrances. The reason for the first needs no discussion. The reason for the second is also clear, but perhaps needs to be stated: because the ‘mortgagor’ normally remained in possession, the lender put up a Horos to give notice that the land was already encumbered, to prevent the mortgagor from borrowing again on the same security, representing it as unencumbered—Greek lenders (on bottomry as well as mortgage) always seem to have wanted to stop further borrowing on the same security. I dare say most Greek states used these mortgage-horoi, although stone ones are known only from Attica and four of the islands in the fourth and third centuries. But why on earth would a Horos be desirable in a case of share-cropping tenancy? One could easily discover whether a man was a proprietor or a sharecropper, and the fact that he was one or the other did not need advertising, nor did the terms of his tenancy, especially if they were of a standard variety. I can see no answer to this. You say that Plutarch (Sol. 13) ‘speaks with an uncertain voice’ and ‘contradicts himself’; but this is a product of your own

116

Athenian Democratic Origins

artificial distinction between rent in arrear and other kinds of debt. The relevant passage reads as follows. ‘The whole demos was in debt to the rich. For either they worked the land for them, paying them a sixth of its produce, being called Hektemorioi and Thetes, or incurring debts on their persons, they became liable to seizure by their creditors, some of them being enslaved on the spot, and others being sold abroad.’5 Now what Plutarch does make clear from the first is that he regards the Hektemoroi as a class of debtors: ‘ oxrevB, indebted. It is characteristic like all the others, they are yp
of Plutarch to be imprecise in the use of technical terms and to get into a bit of a muddle when trying to draw an antithesis, but if we allow for this it is easy to see what he is trying to say. All the demos was in a state of indebtedness to the rich, either as Hektemoroi or (without being Hektemoroi) as having given their bodies as security. From ‘they became liable . . . ’ onwards he is referring to both groups. This is certain: Plutarch had Aristotle in front of him, and in Ath. Pol. 2.2 it is the people who did not pay their rents (misu
vseiB), i.e. the Hektemoroi, who became liable to seizure. F. You take it for granted that there was something inherently bad in the very status of the Hektemoroi, in just being a sharecropper paying one-sixth, and that Solon must have passed a law ‘abolishing the system’. There is not the least support for this in any ancient source, and inherent probability seems to me dead against it. There is nothing necessarily unpleasant about being a sharecropper (it all depends on the precise terms), and one-sixth would be an altogether exceptionally low rate! Share-cropping exists in many places in the modern world, and there is plenty of evidence from antiquity, both from the Near East and from the Roman world—by no means only the Later Empire: note e.g. Pliny, Epistle IX 37.1–4, and the inscription of Henchir Mettich of A.D. 116–17 (FIRA2 I 100, p. 486, I 19–31: one-third of most produce to go to the lessor, plus the operae mentioned in IV 23–7). And see the inscription of Souk el-Khmis: FIRA2 I 103, p. 497, III 8.6 Even

5
yp
‘ oxrevB t vn ’ g ’ vrgoyn eke
’ inoiB & ar ege
a‘
paB men g ar o‘ dZmoB Zn & ploys
ivn: Z ‘ ’ xr
ea & e
‘ kta t vn & gignom
envn telo &ynteB, ektZm
orioi prosagorey
omenoi kai uZteB, Z
’ i to&iB s
’ vgimoi to&iB dane
izoysin Zsan, ’ &y lamb
anonteB ep vmasin ag
oi‘ men ayto ’ i t doyle
yonteB, oi‘ d’ ep Zn j
enZn piprask
omenoi. 6 [See too CSAGW, 214–17, 257.]

Horoi and Hektemoroi

117

on land which is far from fertile, a proportion of one-third to onehalf or more to the landlord is very common, and often it is twothirds or more: I have known 80% and above. In only an insignificant proportion of the cases I have come across has the proportion due to the landlord been lower than one-third, except perhaps for special crops. What is more, even around the half-and-half level the system can work well. There is some evidence of share-cropping being regarded as desirable by tenants and undesirable by landlords in some conditions, as indeed in Pliny’s case. Sharecropping is particularly common where the landlord supplies part of what the Roman lawyers called the instrumentum (as he generally does in me´tayage):7 here indeed it is a socially valuable institution, especially where plough-animals or tools are scarce, or there have been several bad harvests in succession, or irrigation or similar works are necessary—tenants then need help, if they are to live at all, and the natural consequence is share-cropping. Note especially Gaius in Dig. XIX 2.25.6: partiarius colonus quasi societatis iure et damnum et lucrum cum domino fundi partitur, ‘the sharecropper shares both profit and loss with the farm’s owner, by a kind of principle of partnership’. No, there is nothing wrong about share-cropping in itself; and if the Attic Hektemoroi kept the very high proportion of five-sixths of what they produced, even if their landlords did not provide any instrumentum, they ought to have been better off than most tenants. Solon would have been foolish indeed to forbid such an institution, and I see no reason to believe he did, especially as no one says he did. The fact that there is no evidence for share-cropping in Attica thereafter means little. There is no evidence at all about the nature of rents in Attica for (I think) well over two centuries after Solon—and when evidence does become available we find money rents virtually all over the Greek world, without there being any reason to think share-cropping had been forbidden. The development from share-cropping to money rents seems to be a natural one in a healthy economy, just as the reverse development, which we see happening in Italy and elsewhere in the early Principate, is a sign of economic decline. 7

[Me´tayage: ‘a system of land tenure in Western Europe and the US, in which the tenant pays a proportion (usually half) of the produce (as rent) to the owner, who furnishes the stock and seed or a part thereof’ (OED).]

118

Athenian Democratic Origins

What was wrong with the Hektemoroi is made quite clear by Aristotle: any default led to the tenant-debtor becoming liable to seizure. And—a point which is constantly overlooked—this meant that any contract between landlord and tenant was really at an end, because there was now no limit to what the landlord could screw out of the tenant by the threat of enslavement, except of course the danger of the tenant’s running away! The landlord might say, ‘You’ll pay me not one-sixth now, but whatever I choose: onehalf or even two-thirds’—and the tenant would either have to comply or be sold up. Hence what was needed to free the Hektemoroi was exactly what was needed to free other debtors: (a) to forbid the most vicious element in the existing situation, namely the liability to enslavement (‘securing loans on bodies’), and (b) to cancel existing indebtedness (in ‘rent’ or ‘loans’). And this is exactly what Solon did: note the wording of Ath. Pol. 6.1 and 9.1. He may have mentioned share-cropping; but that is the most you are entitled to conclude from Pollux VII 151. The most illuminating parallel for the Seisachtheia is still Nehemiah V 1–13, though I have yet to meet an Oxford ancient historian who uses it, and some have not heard of it. Almost everything in the Solonian situation is there, and the only extraneous element in Nehemiah is the tribute in v. 4. Cf. also II Kings iv 1, for debtbondage. Another vital Old Testament text is I Kings xxi 1–3, which shows how one can knock the bottom out of the theory that Attic peasants would never have mortgaged their lands and pledged their bodies if they could have sold their land—a notion which I seem to remember was the original foundation of the ‘inalienability’ theory. I am sure, by the way, that you would like T. H. Aston, ‘The Origins of the Manor in England’, in TRHS, Ser. 5, Vol. 8 (1958) 59–83, where you will find powerful arguments against the prevailing view about the ‘free ceorl’ as the basis of early English society, and emphasis on a considerable degree of peasant dependence from the beginning. As he says, ‘A large proportion of the known Old English tenantry show no sign of having ever been, as a whole, unattached landowning peasants.’ And what we need is more ‘recognition of the seignorial idea as a primitive force in the organization of rural society’. Almost your own words about early Attica—with which, subject to the disagreements set out above, I beg to agree!

Horoi and Hektemoroi

119

THE 1968 LETTER The trouble is, in my opinion, that we cannot make a consistent and plausible picture from the evidence exactly as we have it: I am sure there is something wrong somewhere! My instinct is to try to find some point at which our sources (and here that really means Aristotle) may have gone wrong, inadvertently and for some plausible reason. I believe that I have found one such point, and that if we take Aristotle to have made just one wrong assumption (quite natural on the evidence available to him), all the knots unravel. In my earlier letter (see D 2 init.) I did not yet dare to mention this theory: at that time it was no more than a hunch, which I occasionally tried out on my better pupils. May I explain briefly how I now see the agrarian situation in 594, sticking as closely as possible to Solon and Aristotle. Forgive me if I go over rather a lot of very familiar ground—I want to give a picture which is complete in essentials as far as the evidence goes. A. The great evil of the existing system arises from the very harsh law of debt: the fact that defaulting debtors become liable to seizure and can be enslaved. Solon’s remedy (precisely adapted to this situation) is threefold: (i) cancel existing debts, (ii) release those already sold as slaves (even bringing back those sold abroad, though how this was done is a real problem—perhaps just a few were brought back, by a public subscription?), and (iii) forbid pledging the body in future. I should like to stress that all Solon’s known remedies specifically concern debt. This is certainly the basic part of Aristotle’s picture. It is partly confirmed, and not contradicted, by the fragments of Solon, of which the most relevant are West’s fragments 4.23–25 and 36.8– 10, 10–15. It is worth noticing that we have 11 lines devoted to the debt situation: the enslavement of the poor, the flight of some and the sale abroad of others, the bringing back of these, and the freeing of them and of those in slavery (doyl
iZ) at home. B. Solon himself begins the iambics in which he celebrates his own achievement by appealing to the Black Earth in support of his claim that in removing the many Horoi he freed the land which until then was ‘in servitude’ (pr
osuen . . . doyle
yoysa, fr. 36.5–7). Aristotle (Ath. Pol. 12.4) quotes this passage as being ‘about the cancellation of debts and those who were formerly in servitude but were freed by the Seisachtheia’ (that is to say, he took Solon to be

120

Athenian Democratic Origins

writing here too about the debt situation); and Plutarch (Sol. 15.3– 6) also appeals to it in support of the view taken by ‘most writers’, that the Seisachtheia was a cancellation of debts, against the theory of Androtion that Solon merely reduced the rate of interest. (Of course I would be the last to defend Androtion’s fancy picture, but it is worth noticing that for him too the trouble lay entirely in debt and usury.) C. In Aristotle, Ath. Pol. 2.2 (and Plutarch, Sol. 13.4) but nowhere else except in the lexicographers (if I may include Pollux among them), we meet those who were ‘called Pelatai (¼clients) and Hektemoroi’ (‘Hektemorioi and Thetes’, in Plutarch). The Hektemoroi, according to Aristotle and Plutarch, are share-cropping tenants paying a rent (m
isuvsiB) of one-sixth of their produce to their landlords, who own what we should call the freehold—this
[‘(the land) was in the ’ igvn Zn last point is clear, because of di’ ol
hands of a few’] in Ath. Pol. 2.2 and 4.5 and tvn & ploys
ivn toyB ’ agro
yB [‘the fields of the rich’] in 2.2, which together make it certain that for Aristotle (rightly or wrongly) the rich owned the land let to the Hektemoroi. According to Aristotle (Ath. Pol. 2.2, lines 11–12 OCT: there is no corresponding passage in Plutarch) the Hektemoroi, when they failed to pay their rents, became liable to seizure (as did all defaulting debtors). Unfortunately neither Aristotle nor Plutarch has a single word more to say about the Hektemoroi, nor does either author say precisely how they were affected by the Seisachtheia. A cancellation of debt would of course wipe out any arrears of rent, but presumably would not of itself affect the payment of future rent, unless perhaps the ‘rent’ was a disguised form of interest on a loan, such as we find in Ps.-Demosthenes XXXVII (c. Pantaen.) 5 ff.: here Pantaenetus, who owed Nicobulus and Euergus 105 minae on a silver-refinery and 30 slaves, took a lease of the property from them at a rent equivalent to interest at 12% p.a. on the debt (which seems to have been roughly half the full value of the property: cf. §31)—the ’ words used are misuo &ytai . . . m
isuvsiB . . . emisu
vsato. (Euergus subsequently took possession of the property, evidently alleging non-payment of the rent/interest.) D. But Aristotle, with the words kai oi‘ daneismo
i [‘and ’ xr
ea loans . . . ’] in Ath. Pol. 2.2 (as Plut. Sol. 13.4 at Z lamb
anonteB [‘or incurring debts’]), introduces debt having a different origin from that of the Hektemoroi, in straight borrowing

Horoi and Hektemoroi

121

rather than defaulting on rent. I suppose Aristotle’s Hektemoroi could have borrowed from someone other than their landlords, to pay their landlords; but surely he is thinking here of different people: peasant freeholders and/or landless men. (Plutarch, Sol. 13.4, who will have had no better sources than Aristotle, is a little ‘ oxrevB, clearer about the two ways in which the demos became yp
although he does not specifically show how the Hektemoroi, his first category, became ‘indebted’; doubtless he took Aristotle’s picture for granted.) This seems to be all the evidence that matters, apart from one or two passages which it is more convenient to mention later. Now to explain it. 1. We have to begin by facing a serious difficulty. It has been usual to connect the Horoi with the Hektemoroi. But if we must accept the Hektemoroi as share-croppers (because this is what Aristotle actually says, and because it is the most ‘natural’ explanation of the term), then I cannot accept the association of Horoi and Hektemoroi, for two reasons: (a) Horoi would be quite pointless in the case of share-croppers, whether they were paying their rents in full or not. As I said in my other letter, D §6, the fact that a man was a share-cropper did not need advertising, nor did the terms of his tenancy, especially if they were of a standard pattern. Nor can I see any point in setting up a Horos to indicate that the Hektemoros had fallen into arrear and become liable to seizure. A Horos would, however, be useful as a warning that the land of a peasant proprietor was subject to some ‘ ’ i l
ysei [‘sale kind of charge (ypou
ZkZ [‘mortgage’], pr &asiB ep subject to redemption’], or something more primitive)—a real form of ‘servitude’ of the land. (b) That the Hektemoros was a share-cropper paying one-sixth of his produce as rent to his landlord could not possibly be a ground for regarding the land as being ‘in servitude’. Again, this is true whether or not the Hektemoros was in arrear with his rent and therefore in the condition of a ‘debtor’. One might try to weaken the force of this objection by claiming that Solon could legitimately speak of ‘servitude’ because the land ‘really’ belonged to the Hektemoros: he and his ancestors had owned it, as peasant proprietors, but he had then got into debt and borrowed on security of it, and it was only when he defaulted that ‘the mortgagee foreclosed’ (as we might say) and became the

122

Athenian Democratic Origins

owner, and the peasant became a Hektemoros; hence the ‘servitude’ of the land, burdened with heavy payments to the rich lender, now in the position of owner of the land. (And you yourself might wish to add that Solon removed this ‘servitude’ by giving the ownership of the land back to its original owner, now the Hektemoros.) Against this I would urge that the one-sixth is an inconceivably low rate of me´tayage for such a situation as I have just depicted (cf. my earlier letter). A foreclosing lender would be mad to give his defaulting borrower such easy terms. Nor can I believe that Draco, or any other legislator in the aristocratic society of pre-Solonian Attica, could possibly have chosen a rate as low as one-sixth as the maximum permissible me´tayage. If the Hektemoros was a payer of one-sixth of his produce, the land must have genuinely belonged to the landlord (the surprisingly low rate of me´tayage being attributable, perhaps, to a general shortage of agricultural labour, forcing landlords to let on terms very favourable to tenants). To this argument a possible counter might be that Aristotle (Ath. Pol. 2.2: clear implication) and Plutarch (Sol., 13.4: definite statement) are just wrong in saying that the Hektemoroi paid onesixth, and that in fact they were share-croppers who received onesixth—in that case, I suppose, the fearfully severe me´tayage terms could have arisen as a result of indebtedness, default and ‘foreclosure’. But five-sixths is as surprisingly high a rate of me´tayage as one-sixth is a low one, and I do not think many people would want to accept that way out. 2. The conclusion the argument points to so far is either (a) that the Horoi have to do, not with the Hektemoroi, but with our people in D 4 above, the peasant freeholders, or (b) that the Hektemoroi were not share-croppers. I suppose (a) is quite possible, and until recently it has been my usual way of resolving the difficulties. But (i) Solon begins his account of his achievements with his removal of the Horoi, and seems to put great emphasis upon it, and (ii) Aristotle seems to be conceiving at least a high proportion of his debtors as Hektemoroi. So I am reluctant to accept (a), which involves treating the Hektemoroi as a not very important class. As an alternative, let us explore (b). 3. Aristotle shows no great knowledge of, or interest in, the Hektemoroi. As you say, ‘he forgot the Hektemoroi once he had mentioned them in his introductory sketch’! And Plutarch adds nothing, nor do the lexicographers. Surely this is because there was

Horoi and Hektemoroi

123

not any evidence about them? In particular, I think we can say with confidence that Solon can never have said anything about them— or anyway nothing illuminating, nothing worth quoting—in his poems or his laws. Probably their name alone was remembered in tradition. Now I should like to emphasise that except for just three lines in Aristotle (Ath. Pol. 2.2, lines 9–11 OCT) and less than two lines in Plutarch (Sol. 13.4, lines 16–17 Teubner) there is absolutely nothing but debt anywhere. Could not Aristotle be simply wrong about the significance of the name Hektemoroi—could he not be jumping to an ‘obvious’ but false conclusion? I would suggest that the hektfrom which the word hektemoroi came did not have anything to do with share-cropping at all, but referred to a rate of interest on loans. In antiquity, of course, payment of interest at a rate per cent per year is virtually unknown: roughly speaking, one either paid (a) at a rate calculated by the month (as normally with mortgages etc.) or (b) with a specified increment, whatever the period of the loan (as normally, I think, with bottomry bonds). In seventh- and sixthcentury Attica, then, you borrow corn from a local nabob, most commonly when your own corn gives out before harvest, and you pay t o e
‘ kton m
eroB ¼ interest at a hekteus per medimnos (hence the curious sixth), either (a) per month or (b) for the whole period during which the loan lasts, i.e. normally until harvest. If (a), then (because 12  16 2=3 ¼ 200) you are paying at what we should call ‘200% p.a.’; and if you repaid in, say, 3 months, you would have to pay back the loan plus 50%; if in 6 months, the loan doubled. If (b), then our notion of a rate per cent per annum is not applicable—the ‘rate’ varies according to the period of the loan. If you repaid in 3 months, plus one-sixth, you would be paying at what we would call ‘a rate of 66 2/3% p.a.’ (four times 16 2/3%); if in 6 months, at ‘33 2/ 3% p.a.’ (twice the 16 2/3%); if in 12 months, at ‘16 2/3%’. In either event we have straight ‘debt’. Of course you give everything you have got as security: your land, your body and those of your family (note kai t a t
ekna kai ai‘ gyna&ikeB [‘and their children and their wives’] in Ath. Pol. 2.2; cf. II Kings iv 1; Nehemiah v 1–13). If you fail to pay principal and interest by the due date you become liable to seizure, with your family. If you adopt this theory about the nature of Hektemorage, you have nothing but debt, as we have seen—and then Solon’s remedies, his only known remedies, are exactly right: cancel existing

124

Athenian Democratic Origins

debts, free those already enslaved for debt, and forbid enslavement for debt in future. I suggest that by the fourth century B.C., from which our earliest evidence comes, there is likely to have been better evidence for what Solon did, the reforms he actually carried out, than for the situation he had to deal with, which would have to be reconstructed by inference, from Solon’s poems and laws and from tradition. 4. There is just one pair of texts you might want to bring up against this theory: ’ imortoB de gZ& par ‘ ep ’ i m
erei (a) Pollux VII 151: ep
a S
olvni Z ’ o tvn gevrgoym
enZ, kai mort Z t o m
eroB t o ap & gevrgvn & [‘In Solon, share-farmed land is called epimortos, and the portion paid by the farmers is called a morte’]. Pollux, I would emphasise, does not mention the Hektemoroi here, though at IV 165 he has ‘ Zmoroi d’ oi‘ pel
atai par a to&iB ’Attiko&iB [‘among the Athenekt
ians the pelatai are called Hektemoroi’]. This suggests to me that if ’ imortoB=mort
Pollux found ep
Z in Solon’s laws, he did not find the Hektemoroi in the same context—or indeed at all: his words at IV 165 could easily derive from Aristotle. ’ imortoB (no. 4985 Latte): l
egetai oy
’ tv (b) Hesychius, s.v. ep
’ ii m
erei ergaz
’ ’ kai o‘ hep omenoB: mort Z g ar t o m
eroB ekale& ito kai ‘ Zmoroi oi‘ t ekt
o e
‘ kton telo &ynteB [‘epimortos: this is also the name for the man who works for a share. For the portion was called a morte, and those who pay a sixth Hektemoroi’]. These texts are most interesting, because of the very great rarity of references to share-cropping in Greece. I would very much like to know the contexts in which Pollux found epimortos and morte, especially because these words do not seem to have continued in use—perhaps because the practice of share-cropping soon died out? I am glad you would now agree that Solon did not pass, or anyway did not necessarily pass, ‘a law which abolished the system’. Share-cropping may, in appropriate circumstances, be in the interests of tenants in general: see F in my earlier letter, especially the quotation from Gaius. Everything depends on the size of the landlord’s share and the penalties for default. 5. But I feel no compulsion to suppose that the Hektemoroi held share-cropped land. Aristotle, and perhaps others before him, thought they did, and of course the lexicographers followed suit. But what evidence did Aristotle have? If the Hektemoroi were in fact indebted peasant freeholders of the kind I have suggested,

Horoi and Hektemoroi

125

would it not be the easiest thing in the world for Aristotle to assume from their name, wrongly, that they were tenants holding share-cropped land and paying a sixth of their total produce? Those texts in Pollux and Hesychius which I have quoted actually help me, because they show how easy it would have been for Aristotle to have made his mistake—he would know that sharecropping did exist in Solonian Attica! 6. I would claim that my theory fits all the evidence very well, if I am allowed to assume one perfectly natural error by Aristotle. We can now think entirely in terms of debt, as all our evidence suggests we should. Aristotle was quite justified in ‘forgetting’ the Hektemoroi once he had mentioned them at the beginning, because they were just a particular class of debtor—but a very numerous and important class, as we want them to be. There is no problem about the Horoi: they can be on the lands of the Hektemoroi (again, as we want them to be), as a warning to intending purchasers that the lands were encumbered. And Solon’s remedies are exactly the ones that were needed. Moreover, we do not have to face your very awkward problem whether the Hektemoroi now had their former ‘leaseholds’ turned into ‘freeholds’ (as you want them to have been), because they were always freeholders. On your view, I think you are in a real dilemma: certainly there were lots of small Athenian landowners in the fifth and fourth centuries, so you need to turn your Hektemoroi into freeholders; but there is not a scrap of evidence for this, although if real it would have been a tremendously important reform, a major act of confiscation and redistribution. Surely it is quite out of character for Solon? And in my view Solon fr. 34 (West) is some evidence against it. Surely (against what you say) ‘the rapacious characters addressed in tetrameters’ (esp. fr. 34) are the demos, in the sense of the poorer classes in general? It is in reply to the ‘rapacious characters’ of fr. 34.1–3, who were annoyed with what he did not do (fr. 34.4–5), that Solon says he did not want ’ the esulo
i to have an equal share (i’ somoir
iZ) of the land with the ’ kako
i (where esulo
i=kako
i ¼ ‘good/bad’ must be used in the social rather than the moral sense, just as in fr. 36.18 (West): ’ agau
oB=kak
oB). So surely it was the kako
i ¼ the lower classes & who had been vainly demanding a redistribution of land (gZB ’ anadasm
oB: cf. Ath. Pol. 11.2; Plut. Sol. 16.3). See also Solon fr. 37 (West), where (after Solon’s reforms) there are two sides

126

Athenian Democratic Origins

(lines 9–10), the ‘greater and stronger’ of line 4, and the demos of lines 1 and 7, whom Solon claims to have restrained, and who were evidently complaining, although they now had ‘more than they could have dreamt of’ (lines 1–3). So, I submit that everything falls into place! A few loose ends remain. 7. From the use of the words epimortos and morte, presumably in Solon’s laws, we can infer the existence of a class of share-croppers; but these need not have been either very numerous or very afflicted. Although being a share-cropper is not as nice in good times as being a tenant at a fixed rent, because the absolute size of the landlord’s share increases in proportion to the crop, yet in very bad times the share-cropper is likely to be rather better off, because the absolute size of the landlord’s share is reduced in proportion to the poor crop—people tend to forget this. If the plight of the Athenian poor in 594 was due partly to a succession of bad harvests, as is so often supposed, then indebted peasant proprietors might well be worse off than share-croppers. It need not surprise us, therefore, if we find share-croppers existing in Solon’s time but not playing any noticeable part in the crisis. 8. As for your point about the Thetes, I agree that the name should indicate a class at least the majority of whom are nonfreeholders; but I would go further and suggest that the Thetes of the post-594/3 period were mostly still not freeholders—after all, the name certainly ought to have some significance for the condition of the class not before its creation but after it had come into existence. As I see it, the Solonian Thetes would include some poor freeholders (ex-Hektemoroi perhaps), but they would mostly be leaseholders (including at first some share-croppers), agricultural labourers, artisans, and traders. I would see evidence of a class of agricultural labourers in Attica (who hardly appear at all, incidentally, in the fifth and fourth centuries) in Solon fr. 13.47–48 (West), where there is an interesting reference to the man who ‘serves (latre
yei) those who have to do with curved ploughs, ’ cutting up the much-wooded land ei’ B eniayt
on (for the year, or year by year?)’: this is part of the long list of occupations in lines ’ 43–62—where the straight freeholding (aytoyrg
oB) farmer, by the way, does not appear in his own right! For me, lots of Hektemoroi, once their debts were cancelled, might rise at once, or very soon, into the zeugite class. Indeed, I wonder whether the need to get

Horoi and Hektemoroi

127

more hoplites may not have been one of the main reasons why the Athenian upper classes consented to let Solon reform the State; cf. the situation in early Republican Rome, where the Plebeians exploited such a need to their own profit. 9. Finally, I would like to emphasise that I have put forward only a tentative theory, to which there may well be a better alternative. Especially after reading that article by Trevor Aston which I mentioned near the end of my earlier letter, I would be perfectly willing to consider an alternative theory which began with ‘the seignorial idea’ and attributed to the Hektemoroi some peculiar dependent status, surviving from the Dark Ages. But at present I do not see how this can be done without making suppositions which are in much more serious conflict with the evidence of Solon and Aristotle than those I have sketched above.

AFTERWORD Several novel theories about the Hektemoroi have been published of late: that they were small freeholders who were obliged to provide periodic labour service to large landowners in exchange for a sixth of the produce (Gallant, 123–4); that ‘the land which was enslaved and ‘ Zmoroi was public land’ (Rihll); that marked with o‘
roi and worked by ekt
the ‘sixth’ was a form of ‘gift’ or protection money exacted from poor farmers by the rich (Harris); that the Hektemoroi were ‘tenant farmers, who rented the land at a fixed rent and paid one-sixth of the rent in silver, when the contract was made, and the rest, either in silver or in kind, at harvest time’ (Stanley, 203). But these are isolated voices, and in English-language scholarship (contrast Welwei, Athen 155–6) the position advocated by Andrewes has become very widely accepted (so e.g. Rhodes, CAAP 94–5; Osborne, GM 223; Murray, EG 192, points out that the central thesis goes back to Fustel de Coulanges). Thus the view against which Ste. Croix is arguing remains a real target. As for Ste. Croix’s specific new theory about the Hektemoroi, the following sentence in Andrewes, 379, is apparently a response to it: ‘the low rate of a sixth tells against borrowing, and so does the uniform rate; it is hard enough to see why any rich man should lend to the poor for so slight a return, and still harder to see why so many should have adopted an identical rate that their debtors acquired this specific designation.’

128

Athenian Democratic Origins

Andrewes, A., ‘Solon’, in The Cambridge Ancient History 2, III.3 (1982), 375–91 Fustel de Coulanges, La cite´ antique (1864) Gallant, T. W., ‘Agricultural Systems, Land Tenure, and the Reforms of Solon’, BSA 77 (1982) 111–24 Harris, E. M., ‘A New Solution to the Riddle of the Seisactheia’, in Mitchell/ Rhodes, DPAG 103–12 Rihll, T. E., ‘EKTHMOROI: Partners in Crime?’, JHS 111 (1991) 101–27 Stanley, P. V., The Economic Reforms of Solon (Pharos, Studien zur griechischro¨mischen Antike, XI, St. Katharinen 1999), 174–203

4 Cleisthenes I: The Constitution

(I) CLEISTHENES IN HERODOTUS AND ARISTOTLE It is the reforms of Cleisthenes, rather than the man, which have a claim to lasting historical importance; but one may naturally feel an interest in the shadowy figure whose name has become permanently linked with the constitution under which Athens flourished for nearly two hundred years. As Wade-Gery has remarked,1 ‘Kleisthenes did not dominate popular imagination. The founder of Democracy, in popular thought, was not Kleisthenes but Solon: the destroyer of tyranny, not Kleisthenes but Harmodios.’ After the Atthidographers Cleisthenes is scarcely ever mentioned and never treated as an important figure, except once or twice by the learned Plutarch. ‘De tous les grands personnages de l’histoire athe´nien, il est sans doute le moins souvent e´voque´’ [‘Of all the great figures of Athenian history, he is without doubt the least often mentioned’].2 What do we really know of the man’s political attitude? The picture of Cleisthenes which is invariably presented by modern writers—namely, that of a man who turns himself into a popular leader at a late stage, in order to achieve personal power— derives ultimately from one source: Herodotus (V 66–72). Aristotle of course had other sources3 for his account of Cleisthenes’ consti‘ vmenoB de ta&iB tution, but his statement in Ath. Pol. 20.1, Ztt
‘ ’ & etaire
iaiB o‘ Kleisu
enZB prosZg
ageto t on dZmon, apodido yB tfiv& 1 EGH 135. But Cleisthenes was the founder of the democracy for Herodotus (VI 131.1) and Cleitophon (Arist., Ath. Pol. 29.3), and the founder or second founder for Aristotle (see esp. Ath. Pol. 22.1; Pol. VI 4, 1319b21–2) and Isocrates (VII 16, 20; XV 232, 306; XVI 26–7). 2 Le´veˆque and Vidal-Naquet, CA 122. (See the whole passage on ‘le destin posthume de Clisthe`ne’, CA 119–22.) 3 If only the institutions of Athens in his own day and intelligent inferences therefrom: see pp. 137–8 below.

130

Athenian Democratic Origins

pl
Zuei t Zn polite
ian [‘since he had been worsted in political alliances, Cleisthenes brought the people over to his side, handing over the politeia to the masses’], must be a direct elaboration of Herodotus: Cleisthenes and Isagoras, the two men who at that time ’ edyn
asteyon [‘held the leading positions in the state’] (V 66.1), ’ ‘ estas
iasan peri dyn
amioB, esso
ymenoB de o‘ Kleisu
enZB t on & dZmon prosetair
izetai [‘were rivals for (political) power; and since he was being worsted, Cleisthenes took the people into partnership’] (66.2). Here I entirely agree with Wade-Gery:4 Aristotle had no narrative source except Herodotus. And see Herodotus V ‘ g ’ & 69.2: vB ar d Z t on ’AuZna
ivn dZmon pr
oteron apvsm
enon t
ote ‘ p
antvB pr oB t Zn evyto &y mo&iran proseu
Zkato (sc. o‘ Kleisu
enZB) [‘then he thoroughly brought over to his side the Athenian people, who had previously been disdained’], introducing his reforms,
te t & prosu
emenoB pollfiv& kat
yperue tvn & whereupon Zn on dZmon ’ antistasivt
evn [‘having won the support of the people he was much more powerful than his rivals’]. The whole familiar picture of Cleisthenes the astute politician who, to serve his own ends, suddenly (as it were) turns demagogue, derives entirely from these passages. This picture may conceivably be true; but it did not satisfy Grote,5 and it should not be meekly accepted without examination. Herodotus was apparently not much interested in the constitution of Cleisthenes, although his reference to Kleisu
enZB . . . ‘ aB wyl aB kai t Zn dZmokrat
iZn ’AuZna
ioisi katast
ZsaB o t [‘Cleisthenes . . . who established the tribes and the democracy for the Athenians’] (VI 131.1) shows that he regarded him as the founder of that Athenian democracy which he evidently admired. At any rate, his account of the reforms of Cleisthenes (V 66–67.1; 69) is—to put it mildly—very unenlightening. In the first of the two passages in which he alludes to the reforms (V 66.2) he mentions only the change in the number and names of the tribes. As we have just seen, the motive Herodotus attributes to Cleisthenes for introducing his reforms in general is factious rivalry with Isagoras for political power; but in regard to one particular feature of the new constitution, the change in the names of the tribes, he offers it 4

EGH 136–8. Against Wade-Gery’s view of one correction of Herodotus by Aristotle, I would accept that of Le´veˆque and Vidal-Naquet, CA 7 n. 3. 5 HG III 348–50, ch. xxxi.

Cleisthenes in Herodotus and Aristotle

131

’ i, V 69.1) that Cleisthenes was motivas his opinion (dok
eein emo
ated by contempt for Ionians: it was in order that the Athenians should not ‘have the same tribes as the Ionians’, in Herodotus’ ’ i, 67.1), that the Athenian lawgiver ‘imitated opinion (dok
eein emo
his maternal grandfather, Cleisthenes of Sicyon’ (67.1; 69.1), reorganising the Athenian tribes and abandoning the old Ionian names, just as the tyrant Cleisthenes had renamed the Dorian tribes of Sicyon. This is such feeble stuff that at least one of Herodotus’ more determined admirers has been driven to wonder whether he may be speaking ironically!6 But Herodotus refers to the Athenian Cleisthenes as imitating the Sicyonian twice over, and it is hardly likely that such a skilled writer would so labour his irony. In his comparison between the two statesmen, of course, Herodotus is not giving the motive for the Cleisthenic tribal reforms as a whole, but only for the change of names. But our complaint against Herodotus is precisely that he chose to lay all the stress on the change of names, for this was a trivial part of the whole reform; and in concentrating upon it Herodotus, who ‘sometimes writes for children and sometimes for philosophers’,7 is hardly writing for philosophers. In his second passage (V 69.2) Herodotus says, t aB wyl aB ’ iZse ple &ynaB ej ’ elass
’ metvn
omase kai epo
onvn. d
eka te d Z ’ i tess
ervn epo
’ iZse, d
eka (?d
ekaxa) de kai toyB wyl
arxoyB ant ’ t d
ZmoyB kat
eneime eB aB wyl aB [‘he renamed the tribes and increased their number. He also established ten phylarchs in place of four, and distributed the demes ten in each tribe (or: among the tribes in ten groups)’]. Here again he begins with the names and number of the tribes. Then we have a reference to the change in the number of the phylarchs—a statement which is so trivial, taken literally, that some commentators have wanted to make it refer to the creation of the Strategoi.8 The essential new feature of Cleisthenes’ constitution, of course, was the basing of citizenship for the first time upon membership of a deme;9 but only at the very end of Herodotus’ account are the demes so much as mentioned, and then all we are told is that Cleisthenes ‘distributed the demes among the tribes in ten groups’, or ‘distributed the demes ten in 6

J. L. Myres, Herodotus Father of History (1953) 84: ‘ironically or not’. Edward Gibbon, The Decline and Fall of the Roman Empire (ed. J. B. Bury, 1935) II 495 n. 54 (ch. xxiv). 8 9 This I cannot accept: see pp. 223–5 below. See pp. 138–40 below. 7

132

Athenian Democratic Origins

each tribe’, according to which reading we adopt in V 69.2, d
ekaxa or the d
eka of the manuscripts.10 It would be difficult to think of any statement about the demes which could say anything less significant than this. And the trittyes, a particularly ingenious and unusual feature of the Cleisthenic tribal system, are not so much as mentioned. Herodotus was not writing a constitutional history, and there was no compelling reason why he should have included any details about these particular constitutional reforms, any more than about Solon’s, on which he is entirely silent, although he refers to Solon several times. But since he does give us a brief account of the reforms, we are entitled to examine it, to see if he grasped the nature and significance of Cleisthenes’ work. Surely the superficiality of his picture and his almost frivolous concentration upon inessentials suggest that he never received a full account of the reforms and did not properly understand them. And if he did not, he is hardly likely to be a reliable witness to the attitude of the man who introduced them. Herodotus may or may not have believed that among the events which precipitated Cleisthenes’ ‘conversion’ to the democratic cause was the election of Isagoras to the archonship for 508/7. It is not Herodotus but Aristotle (Ath. Pol. 21.1) who tells us that the archon in the year of Cleisthenes’ reforms was Isagoras, and whether Herodotus knew this fact or not we cannot tell. Believing as he did that the Alcmeonids had been in exile during the whole of the Peisistratid tyranny,11 Herodotus evidently did not realise12 that Cleisthenes had been eponymous archon in 525/4 and therefore could not have been a candidate for the archonship in 508. Before the publication of the archon-list, M/L 6 ¼ Fornara 23, revealed this interesting fact for the first time, Herodotus’ modern readers ‘ vmenoB . . . ta&iB often assumed—in spite of Aristotle’s phrase, Ztt
‘ etaire
iaiB [‘worsted . . . in political alliances’]—that the struggle between Cleisthenes and Isagoras peri dyn
amioB [‘for (political) power’] was indeed for the archonship.13 I do not believe that Aristotle thought this, and the words I have quoted suggest to me that he at any rate did not understand Herodotus in that sense. The archonship was of course a political prize, and 10 11 12

See Endnote i (p. 172 below). Hdts VI 123.1; cf. misot
yrannoi [‘tyrant-haters’] in 121.2. 13 See Wade-Gery, EGH 164–5. See How and Wells, CH II 32.

Cleisthenes in Herodotus and Aristotle

133

Cleisthenes may well have coveted it for a relative or an associate, Alcmeon for instance. Especially in these early days, when some leading clan or family might have greater political influence than any comparable group in the late fifth or fourth century, each faction would naturally try to obtain the principal offices, especially the eponymous archonship,14 for its own members. Herodotus must surely have had in mind a failure by Cleisthenes to secure that leading political position in aristocratic circles to which he would have had reason to aspire as the head of the Alcmeonid house; but he may have seen the election of Isagoras (if he knew about it) as the last straw, the fact which finally made it plain to Cleisthenes that a change of tactics was necessary if he was to secure political predominance. In one respect many of his modern readers may have misunder’ stood Herodotus. By the words pr
oteron apvsm
enon in V 69.2 Herodotus need not have meant ‘formerly rejected by himself’, as has sometimes been supposed.15 I believe Herodotus probably meant that the demos had until now been generally disdained, kept out of political power. But even if Herodotus did mean ‘disdained by Cleisthenes’ I would not give his statement much weight, because there is no reason to think he had reliable information about Cleisthenes’ earlier political attitude or his motives in introducing his reforms. There is some very strong circumstantial evidence which must make us distrustful of the portrait of Cleisthenes as a man who becomes a popular leader solely in order to advance his own political power. This is the fact that the constitution of Cleisthenes reserved no special place whatever for Cleisthenes (who had been eponymous archon already and therefore, presumably, could hold none of the nine leading offices of the State) and indeed looks as if it was deliberately designed to prevent any one individual from achieving personal power! If Cleisthenes had really been eager above all for personal power, is it such a constitution as this that he would have advocated? Is not Herodotus’ Cleisthenes, on the contrary, just the sort of portrait that might be expected to emanate from aristocratic Athenian informants of the historian who shared 14

Cf. Arist., Ath. Pol. 13.2 (fin.). ’ Any reader would naturally take pr
oteron apvsm
enon here as passive, and I therefore think it would be perverse to see it (with Kru¨ger) as a middle, meaning ‘which had previously rejected him’. 15

134

Athenian Democratic Origins

the cynical opinions about the proper behaviour of men of their class in a democracy which are put by Thucydides (VI 89.4–6) into the mouth of Alcibiades—‘Oh yes, Cleisthenes certainly became a prost
atZB to &y d
Zmoy [‘champion of the people’]; but of course he played that role not from any love for democracy but because the circumstances obliged him to do so, and for perfectly respectable motives: the advancement of his own political influence and that of his family.’ It must have been the Alcmeonids who were responsible for Herodotus’ false statement that the family had been in exile throughout the Peisistratid tyranny, and surely he will also have obtained from them the information that he gives about Cleisthenes. Since they could be so misleading on a question of fact, it would be foolish to place any reliance upon their representatives about the motives that had inspired Cleisthenes’ reforms. We cannot know what Cleisthenes’ motives were in 508, the evidence being so unreliable; but I suggest that the circumstantial evidence from the nature of the reforms should be given at least as much weight as the statements of Herodotus, which are hardly likely to be very well founded. I think we should keep open two different possibilities. One is that Herodotus may have been misinformed, and Cleisthenes may have been the leader of a group of genuine reformers who desired to bring about a real democratisation, in the sense of the effective participation of a much larger proportion of the citizen population in the administration of the State (including its constituent local units), as well as probably broadening the citizen body itself to some extent. Undoubtedly ’ eneto Z ‘ polite
ia & now dZmotikvt
era poly tZB S
olvnoB eg
[‘the constitution became much more democratic than Solon’s’] (Ath. Pol. 22.1).16 It is not necessary to suppose that Cleisthenes was an ‘idealist’, except perhaps in the same sense as Solon— may not Cleisthenes, like Solon, have been doing what he thought was best for Athens, as well as for himself? Apart from a handful of aristocrats who may have wished to return to something like the pre-Solonian constitution, there is no reason why any Athenian should have objected to the excellent new constitution of Cleisthenes. It is often misleading to translate dZmotik
oB by ‘democratic’: see my CAE 22–4. But in this case I think we can safely say, ‘the constitution became more democratic than Solon’s’. 16

Cleisthenes in Herodotus and Aristotle

135

The alternative possibility is that Herodotus was right in the main about Cleisthenes himself: his attitude may have been purely or largely selfish, and he may have seen the introduction of the new constitution entirely or at least primarily as a means to the consolidation of his own political power and that of his family. If we adopt this second alternative, however, we must make certain qualifications. The new constitution bears all the marks of having been carefully thought out and designed with quite remarkable skill and brilliance for at least the immediate ends we know it did achieve— it is hardly possible that anyone at this stage foresaw the further reforms of the fifth century. This constitution is no hastily devised makeshift, put together by one individual to procure his own advancement. If Cleisthenes himself was not a sincere reformer, interested in creating ‘national unity’ and in providing Athens with a new constitution which would secure such wide and enthusiastic support as to allow her now great potentialities to develop freely, then he was not the real architect of the constitution which has come down to us under his name, but merely gave his powerful and doubtless essential support to a construction produced by others, who may have been working out for some years what kind of regime they would try to bring about when the tyranny ended. In assuming the leadership of the progressive cause Cleisthenes may have been in a position to stipulate for certain advantages for himself or his family; but it must not be assumed that he could do as he liked here and ‘fiddle’ the constitutional arrangements just as he pleased. It is essential to remember that he was no Solon, personally entrusted by special commission with the task of enacting laws, given carte blanche, and able to pass, within the limits of practicability, whatever laws he himself thought best, but merely the leader of a faction, who had freedom to act as he wished only to the extent that he could induce a majority of Athenians to back him. It is no doubt due to the Herodotean picture of Cleisthenes the astute politician, aiming primarily if not solely at power for himself, that so many modern writers have wanted to look for ‘fiddles’ in his constitution. This constitution, they feel, cannot be as straightforward as it looks; there must have been something special in it for Cleisthenes or the Alcmeonids. The result has been a whole series of ingenious speculations, which it is now time to discard. Once we have ceased to regard the constitution as having been designed above all to bring increased personal power to

136

Athenian Democratic Origins

Cleisthenes we shall not, of course, close our eyes to the possibility of there being certain Cleisthenic ‘fiddles’, but we shall no longer demand an explanation of the main features of the constitution in these terms. Indulgence in such fantasies is likely to lead to our overlooking the real nature and purpose of the reforms of Cleisthenes, t Zn dZmokrat
iZn ’AuZna
ioisi katast
ZsaB [‘establishing the democracy for the Athenians’].17 I shall for convenience refer to the reforms of 508/7 as those of Cleisthenes and may sometimes speak as if he alone were their author; but when I say that ‘Cleisthenes’ did or intended this or that, I must usually be understood to refer to Cleisthenes and his associates, who drafted the new constitution and procured its acceptance by the demos.

(II) THE CONSTITUTION OF CLEISTHENES The constitution voted in 508/7 B.C. which we know as that of Cleisthenes1 has been described many times, but usually in such a way as to overemphasise particular elements of it at the expense of the main features, and it is the latter above all which I wish to clarify. Two valuable recent works, by C. W. J. Eliot and D. M. Lewis2 (writing from very different points of view), have 17

Hdts VI 131.1. For the factual details it will be sufficient to refer to Busolt-Swoboda, GS II 868–87; Hignett, HAC 124–58. For the chronology I would follow Cadoux, AAKH 113–16, rather than C. W. Fornara, in CQ 57 ¼ n.s. 13 (1963) 101 ff; G. V. Sumner, AF6 84–6; and in CQ 55 ¼ n.s. 11 (1961) 35–7; Eliot, CDA 145–7; or R. Sealey, in Historia 9 (1960) 175–7. (On the last, I agree with Le´veˆque & Vidal Naquet, CA 46 n. 1.) 2 Eliot, CDA, esp. 136–58; Lewis, CA. These works rest upon a whole series of patient investigations, of sites as well as literary and epigraphic sources, which began to take shape in the late 19th century, in the work of Loeper, Milchho¨fer, von Schoeffer, Szanto and others: see the bibliography in Eliot, CDA 160– 9; add & V. von Schoeffer, in RE V I (1903) 1–33, 35–122, s.v. dZmoi; B. Haussoullier, La vie municipale en Attique (1884). During the present generation further advances have been made in the study of demes and trittyes: the most useful works, apart from those of Eliot and Lewis, already cited, are A. W. Gomme, The Population of Athens in the Fifth and Fourth Centuries B.C. (1933), 37–66 (with a map at the end of the book); H. Hommel, ‘Die dreissig Trittyen des Kleisthenes’, in Klio 33 (1940) 181–200 (with map, p. 197, and names of trittyes, pp. 198–9; cf. Eliot, CDA 147–58, esp. Table I, p. 157); Hommel, in RE, 2te Reihe, VII I (1939) 330–70, s.v. Trittyes; A. Philippson (& E. Kirsten), Die griech. Landschaften, I iii (1952), with the map at the end of I ii (1951: explanation of the numbers on this map at iii 1065–8); E. Kirsten, ‘Der gegenwa¨rtige Stand der attischen Demenforschung’, in 1

The Constitution of Cleisthenes

137

made it easier for us to understand the essential nature of the reforms, by clearing away some widely held misconceptions about the arrangement of demes and trittyes in the Cleisthenic system3— misconceptions upon which earlier writers had founded a variety of false inferences about the nature and purpose of the reforms. If we are reasonably well informed about the new constitution, even in details, it is because we can use, to supplement the evidence of the literary sources, the facts of the Athenian constitution as we know them in the late fifth and fourth centuries, from inscriptions even more than from literary sources. There is often a danger that we may be failing to allow for unrecorded changes made between 508/7 and the time from which our evidence mainly comes. But I believe that the main conclusions I shall reach here are not such as to be liable to falsification by the existence of such changes: the essential points seem to me clear and certain, although there are many individual questions of detail which for the time being remain open. *** The only literary source which even professes to give details of the new constitution is Aristotle, Ath. Pol. 21.4 I do not see how we can confidently decide whether the actual laws of Cleisthenes were available to and consulted by Aristotle.5 The much discussed ‘rider of Cleitophon’ of 4116 is inconclusive. Even if (as is possible) Atti del III [Roma, 1957] congr. internaz. di epigr. gr. e lat. (Roma, 1959) 155–71 & Tav. XXVI. Among other studies, see S. Dow, ‘The Attic Demes OA and OE’, in AJP 84 (1963) 166–81; W.E. Thompson, ‘TrittyB t vn & pryt
anevn’, in Historia 15 (1966) 1–10. 3 The essential maps for ordinary use are the admirable ones by Kirsten in Stier’s WAW I (1956): the main one on p. 13 (the essentials of which are reproduced with little change by Le´veˆque & Vidal Naquet, CA 15, Fig. 1) and three smaller ones (with more details of demes in the City trittyes) on p. 12. The more recent outline map given by Eliot, CDA 139, is described by the author as ‘a cautious representation of current opinion based on Gomme and Kirsten but incorporating minor corrections of them’. And see the map (based on Eliot’s) in Forrest, EGD 192. (This last work appeared only after this essay had been finished, and I have been able to make only one or two references to it.) 4 I am leaving the question of ostracism, dealt with by Arist., Ath. Pol. 22.1, to a later section of this essay. 5 My own feeling is that they probably were not. Here (as over his belief that Cleitophon was interested in Cleisthenes’ procedure) I disagree with Wade-Gery, whose discussion of the constitution of Cleisthenes (EGH 135–54) is one of the few illuminating treatments of the subject as a whole. 6 & Arist., Ath. Pol. 29.3: prosanazZtZsai . . . kai toyB patr
ioyB n
omoyB o y‘B Kleisu
enZB e
’ uZken o‘
te kau
istZ t Zn dZmokrat
ian [‘seek out in addition the

138

Athenian Democratic Origins

Cleitophon’s proposal was sincere, all it proves is that men then thought a search7 might reveal the laws of Cleisthenes and make them available for examination, not that the laws were actually available and were duly examined. And there is no other reference anywhere to consultation of ‘the laws of Cleisthenes’. Although I would not go as far as Hignett, who insists that the description of the reforms of Cleisthenes in the Ath. Pol. ‘cannot have been taken from an official record’,8 I see nothing in Ath. Pol. 21 which need be based, or is even likely to be based, upon any source other than intelligent inference from the constitution existing in Aristotle’s own time.9 Aristotle knew that a certain number of elements in the developed Athenian constitution were post-Cleisthenic, and the rest could safely be attributed to Cleisthenes. Aristotle’s historical method was perfectly valid in this case, and his picture will be substantially true unless there were important post-Cleisthenic reforms he did not know about. Everyone nowadays, I believe, will accept nearly all the constitutional details given by Aristotle, and the question whether he had documentary sources is therefore of no very great importance. It might be settled in the negative if we could decide in favour of Eliot’s contention that the trittyes could not have been allocated to the tribes by lot,10 as Aristotle says they were;11 for if Aristotle was able to read the laws of Cleisthenes he could hardly be wrong on this point. But Eliot’s case cannot yet be regarded as proved,12 and the question must still be treated as an open one. *** First and most important, the constitution of Cleisthenes represents a change from classification by kinship (largely fictitious ancestral laws which Cleisthenes enacted when he established the democracy’] (the remaining words are presumably from Aristotle or his source). 7 & The precise implications of prosanazZtZsai are difficult to define; but surely ’ ’ the words anazZte& in, anaz
ZtZsiB always contain an element of ‘searching for’ or ‘discovering’. Cf. Ar., Lys. 26; Thuc. II 8.3; VIII 33.4; Hdts I 137.2; Plato, Critias 110a; Apol. 18b; Laws III 693a; II Alcib. 140d; OGIS 267.9. At any rate, I have not come across a case in which these words are used in reference to the examination of something that is already known or easily available. 8 HAC 130 (my italics). 9 It cannot be assumed that the pre-Aristotelian Atthides contained much information on constitutional matters: see pp. 286–307. 10 CDA 140–5. See p. 156 below. 11 12 Ath. Pol. 21.4. See p. 156 below.

The Constitution of Cleisthenes

139

kinship), through the phratries and the four old ‘Ionian’ tribes, to a territorial classification in terms of ten new tribes, made up of demes. We speak of Cleisthenes’ ‘tribal’ reform, and of the tribes as being ‘divided into demes’; but we must never forget that it was the deme which was the basic unit, and that in reality the tribes were not ‘divided into’ but built up from the demes, in units known as Trittyes—thirdings or Ridings. A man became a citizen of Athens, and a member of a particular tribe, by being entered in a deme register, its lZjiarxik on grammate&ion, and (subject to appeal to a dicastery) he lost his citizenship if his deme decided to erase his name from its register. There was no general citizen register, nor (as far as we know) were there tribal registers:13 there were simply the collective registers of the various demes. It seems certain, and is now (I think) generally agreed, that the demes were created as far as possible out of natural units of habitation, which however had previously possessed an existence only as such, and had not formed recognised political units before Cleisthenes, who may in some cases have had to draw definite boundary lines where only vague ones, or none at all, had formerly existed, and even create some of the demes, especially in the city itself.14 By the early second century B.C. there were 170–174 demes,15 and there were probably almost as many from the time of Cleisthenes.16 The change to a system of organisation in territorial rather than kinship groups—in tribes (wyla
i) which were topika
i [‘by location’] rather than genika
i [‘by descent’], to use the convenient language of Dionysius of Halicarnassus17—occurred elsewhere in the Greek world, and there is reason to suppose that such a step was characteristically part of a democratisation,18 even if it is true 13

As apparently at Syracuse in the late 5th century: see Plut. Nic 14.6. See Arist., Ath. Pol. 21.5. 15 Polemo, ap. Strab. IX I 16, p. 396. On Hdts V 69.2, see Endnote i (p. 172 below). 16 [Recent enquirers favour a slightly lower total, of 139 or thereabouts: cf. Whitehead, DA, 21.] 17 Ant. Rom. IV xiv 2, dealing with the ‘Servian reforms’ at Rome. 18 See Busolt, GS I 256 ff., esp. 257, 264, 272 (’Massgebend waren fu¨r die Einteilung der ganzen Bu¨rgerschaft nach rein territorialen Prinzip ohne Ru¨cksicht auf den gentilizischen Zusammenhang meist die Grundsa¨tze der Demokratie’); also A. Andrewes, ‘The Patrai of Kamiros’, in BSA 52 (1957) 30–7, esp. 31–2, 36–7; W. G. Forrest, ‘The Tribal Organisation of Chios’, in BSA 55 (1960) 172–89, esp. 180–1. 14

140

Athenian Democratic Origins

that there was nothing necessarily and intrinsically democratic about it.19 (It is also possible to draw a parallel between the Cleisthenic reform and the institution of the territorial tribes at Rome).20 Many of those who have recognised this aspect of the reforms as the basic one have not made it clear why it was such a great improvement upon the earlier system or why Cleisthenes introduced it, and others have given unsatisfactory explanations. Recently, for example, D. W. Bradeen,21 expressing views which seem to have long been widely held, has attributed the fundamental reform to Cleisthenes’ ‘desire to facilitate the admission of new citizens and to break the political power of the Eupatridae in the phratries’. We shall see that the phratries must have controlled admissions to the citizenship but that there is no evidence of any kind that other activities which can be described as ‘political’ were carried on by or through the phratries. No one will doubt that in 508 the leading Athenian political families possessed very great power and influence. We know little about them, but it seems very probable that by the end of the tyranny they will not have been by any means exclusively eupatrid, as they were at the beginning of the sixth century, for under Solon’s timocracy some wealthy non-eupatrids will have wanted to play their part on the political stage, and the powerful influence of Peisistratus and his sons will have brought some of them into leading positions, if only to buttress the tyranny against eupatrid resentment.22 It would be a mistake to conceive the situation in 508 in anything like the same terms as in 594. The ‘leading political families’ in 508, then, although doubtless they will have had a strong eupatrid core, were not identical with the pre-Solonian eupatrids. There is no harm in our calling them ’ eneia [‘nobility’] as arxa& ’ ‘the aristocracy’. Aristotle defines eyg
ioB ’ plo &ytoB kai aret
Z [‘ancient wealth and excellence’].23 The leading 19

See Busolt-Swoboda, GS I 264; II 877. That a change from political classification in terms of descent to that in terms of locality is a common feature of early societies was well known to Niebuhr and Marx. 20 See H. Last, ‘The Servian Reforms’, in JRS 35 (1945) 30–48, esp. 38–9, 48. I think Last rather overemphasises the enlargement of the citizen body effected by the ‘Servian’ tribal reforms, at the expense of other factors; but the article is a valuable one. 21 TCR 22. 22 See pp. 80–3 above. 23 Pol. IV 8, 1294a21–2; cf. V 1, 1301b3–4, and Eurip., fr. 22 (Nauck); contrast Pol. III 13, 1283a37.

The Constitution of Cleisthenes

141

position of the late sixth-century Athenian aristocracy in every sphere of social and political activity was indeed due to their inherited wealth and the fact that they had far easier access to ’ aret
Z than other Athenians, for they alone received from their families, from their youth onwards, the one kind of training for political leadership which was then available: this, when success’ ful, produced aret
Z. (Foreign connections with ‘the right people’ in other states, an essential part of the equipment of the aspiring statesman, were naturally transmitted with the plo &ytoB and the ’ aret
Z. Xenia and proxenia,24 in particular, were normally hereditary.) These leading families no doubt ‘controlled the phratries’, just as they ‘controlled the main cults’ by monopolising the priesthoods; but there is no reason to think they exercised their political power through the phratries or through the cult associations: their ’ ’ political power, based on their arxa& ioB plo &ytoB kai aret
Z [‘ancient wealth and excellence’], was exercised directly through the holding of high office (including membership of the Areopagus) and predominance in the Assembly and the Council of Four Hundred. It was, I believe, partly the absence of political bodies at lower levels which made it so easy for the leading aristocratic families ‘ (oi
koi) and associations (etaire& iai) to exercise political power— subject, of course, to the overriding control of the Peisistratid house during the tyranny. To a reformer like Cleisthenes there was nothing intrinsically objectionable about the phratries: we must recognise this, and not jump to the conclusion, for which I see no support in the ancient evidence, that the reforms of Cleisthenes ‘must have been’ aimed in some way against the phratries or the cult associations, or at least the position of the aristocrats in those organisations.25 All these, apart from the removal from the phratries of their one great political privilege (control of admission to citizenship), were left entirely untouched by Cleisthenes: t a de g
enZ kai t aB wratr
iaB kai t aB i‘ erevs
ynaB ’
‘ astoyB kat e’
iasen exein ek
a t a patria [‘he permitted everybody to keep their clans and phratries and priesthoods in accord with tradition’] as Aristotle (Ath. Pol. 21.6) explicitly says. What 24 Xenia is usually translated ‘guest-friendship’. For proxenia (the accredited right of representation of another city in one’s own), see S. Perlman, ‘A Note on the Political Implications of Proxenia in the Fourth Century B.C.’, in CQ 52 ¼ n.s. 8 (1958) 185–91, with references to earlier work; add Busolt-Swoboda, GS II 1246–9. 25 Cf. pp. 142–3 and 160 below.

142

Athenian Democratic Origins

Cleisthenes did above all was to set up a new kind of political classification, functioning at more than one level, which immediately provided a new sphere of political activity—for the aristocracy, of course, as the people most competent at the time to provide the necessary leadership, but also (and this was something quite new) for the ordinary citizen who had the necessary leisure. The nature of the units in and through which this took place was allimportant: as we shall see presently, at lower levels (deme and perhaps tribal assemblies) they favoured the participation, and provided scope for the initiative, of the non-aristocrat; and at higher levels (the Assembly, the state magistracies, the Heliaia and above all the Council), even though leadership remained largely in the hands of aristocrats, the whole atmosphere gradually changed in such a way as to allow the mass of citizens to express their will. How far some of these developments were foreseen in 508, of course, is arguable. I suggest that quite apart from any democratisation which may have come about through the reorganisation of the Council we must see in the new system three quite different changes for the better, operating from the very first: 1. Politics and war were the two spheres in which the new classification was most important. (We are not likely to forget the political aspect of constitutional reforms; but it is easy for us to overlook their military significance, and we should remind ourselves of Aristotle’s advice:26 ‘It is essential that a constitution be constructed with an eye to military strength.’) When the Athenians fought they were brigaded in tribal divisions, and the Council, that essential organ of State, was divided into tribal prytanies—at least from the time of Cleisthenes, in my opinion,27 and perhaps even from Solon’s. Now the old ‘Ionian’ tribes must, by the late sixth century (and indeed long before) have become scattered more or less indiscriminately all over Attica. Even many of the older clans had already come to consist of widely separated units28—a fact (too 26

27 Pol. II 7, 1267a20–1. See p. 167 and n. 138 below. See Wade-Gery, EGH 107, on the Brytidai and Amynandridai: the wide distribution among demes, as Wade-Gery points out, must have existed in Cleisthenes’ time. The situation is the same with many other clans, e.g. the Eumolpidai and the Kerykes. It is just possible, I suppose, that membership of some of these clans had later been acquired by men whose ancestors had not been members in Cleisthenes’ day; but such privileges are not likely to have been bestowed often. 28

The Constitution of Cleisthenes

143

often overlooked) which leads to the important conclusion that the phratries, a fortiori, most often have been scattered over different parts of Attica. This would have made nonsense of the old system of classifications: in each locality there would be members of all four tribes, and the tribal divisions would thus have become quite arbitrary; and the phratries, which may once have had some importance as political units, and even as military formations,29 would surely have lost all real meaning except as organisations celebrating religious cults in common and controlling access to citizenship. The troubled situation in the seventh century, which culminated in the reforms of Solon in 594/3, and the great change which evidently came over the face of Attica during the tyranny, will doubtless have brought about a considerable redistribution of the population, with a diffusion of many phratries over different areas of Attica. Already in the second quarter of the sixth century the political groupings we hear of were given names having a local character.30 Under the new system, based upon ‘local units which with few exceptions were natural and well established’ (as Hignett puts it),31 a man would fight, and do politics—at any rate in his prytany and his deme and tribal assemblies—with his own immediate neighbours, whom he would usually know, from almost daily contact, so that he could tell whom to trust, and listen to (or shout down), and choose as his commander or spokesman. The Greeks were well aware how desirable it was, given their form of society, for the members of each inhabited area within the State to know each other.32 (Indeed, some would have extended this principle to the whole State, and would have limited the number of citizens with 29

See A. Andrewes, ‘Phratries in Homer’, in Hermes 89 (1961) 129–40. See p. 154 and n. 83 below; and cf. pp. 160–1 below. HAC 142. Recent work on the demes (and trittyes: see pp. 148–54 below) has given further proof of this: see esp. Eliot, CDA. R. S. Young, in Hesp. 20 (1951) 135 ff., at pp. 140–3, has suggested that the boundaries of even the city demes were determined to some extent by important roadways. 32 ’ yB ’ en p
’ ’ gnvr
imoyB ayto See Plato, Laws V 738de: oyffl me&izon oyd olei agau
on, Z ‘ iB ei
nai [‘there is no greater benefit for the state than for the people to be known ayto& to one another’]. (Plato hoped the desired association would be achieved through cults.) Cf. VI 771de; also Arist., Pol. VII 4, 1326b14–18, where Arist. shows he believed not only that citizens ought to know personally the candidates they elected to office but even that dicasts ought to have personal knowledge of the litigants whose cases they tried. (Arist. of course is thinking of the city as a whole, not just the individual tribe.) 30 31

144

Athenian Democratic Origins

this end in view.)33 It is true that since men were registered for ever after in the demes and tribes of which their ancestors had been members in the time of Cleisthenes, the principle I have laid down here would gradually have become less and less applicable, as demesmen changed their place of residence. This, however, was partly a necessary consequence of an overriding principle, to which most if not all Greeks, and incidentally the Romans, held as firmly as the Athenians: that a man must belong to the same division of the State as his father and his more remote ancestors before him. It may also have been partly due to a desire to preserve approximate equality among the tribes in case of future movements of population between demes in different tribes: these would not affect the numbers included in tribes based on the ancestral principle adopted by Cleisthenes. I am assuming here that the intention of Cleisthenes was to set up approximately equal tribes, which would then be likely to preserve their rough equality for the foreseeable future, because of the hereditary nature of their membership. I see no way of proving this assumption, but I think it is perverse to doubt it,34 if only because the elections of numerous magistrates, as well as the councillors, must always have been made on the basis of one or more per tribe, and this surely implies approximate equality of membership among the tribes. But in practice, of course, it would be exceedingly difficult (in the absence of good maps or accurate lists of citizens) to ensure anything like proper equality, and there may well have been appreciable differences in size from the very first. ‘The tribes obviously were intended to be approximately equal in population, but naturally could not be and were not exactly equal.’35 2. Before Cleisthenes’ time citizenship must have been claimed through membership of a phratry, although we have no evidence what the criteria of admission to phratries were, and it seems to me that different phratries are very likely to have had different rules— a possibility which is often overlooked.36 Admission to the citizen33

e.g. Plato, Laws V 737e etc.; Arist., Pol. VII 4, 1326a25–b25. W. E. Thompson, TTAH 402–8, appears to have a very different view, but this is largely because he exaggerates the position of his opponents (mainly Eliot, CDA 140–5): see p. 156 & n. 88 below. 35 Larsen, RGGRH 6. 36 Since the masterly article by A. Andrewes, PP, it has become more than ever true that the evidence for membership of a phratry as a qualification for citizenship in 34

The Constitution of Cleisthenes

145

ship, through the phratries, will then have been to a great extent in the hands of the most prominent members of the phratries, probably the heads of the leading clans within them,37 who in the early times will have been overwhelmingly eupatrid and may have been predominantly so even at the end of the sixth century. Since the phratries were in theory kinship organisations, and moreover religious, in the sense that their members participated in common cults, some of them no doubt very ancient, there might be genuine feelings of antipathy towards newcomers who had managed to gain admission, especially to those who had perhaps become members not through aristocratic patronage38 but through the intervention of the tyrants on their behalf. After the fall of the Peisistratids the position of these newcomers would become precarious, and no doubt many were expelled by their phratries, and therefore lost their citizenship in or just after 510/09. This may possibly have taken place as the result of a formal diacZwism
oB [‘revision of the citizen register’] as stated by Aristotle (Ath. Pol. 13.5);39 but if any such order was issued by the State, it could hardly have been anything more than an instruction to the phratries (analogous to that given to the demes in 346/5) to carry out a scrutiny of their lists and expel intruders who had obtained entry irregularly—the power of decision in individual cases would surely have remained with the phratries. early Athens is mainly negative: the fact that there is really no alternative. But I think it is now universally accepted. Jac., ii 142 n. 14, thinks we can be ‘quite certain . . . that before Kleisthenes those who did not belong to a phratry were not citizens’. It is perhaps worth remarking here that in some other states admission to membership of some phratries might evidently be gained more easily than to others: see IG XII v 716. 8–9; 717. 7–8; 720. 3–5 (Andros); cf. viii 267. 9–10 (Thasos, p
atrZ). Most illuminating here is the famous inscription, SIG3 921 ¼ IG II2 1237 [reedited with full commentary by C. W. Hedrick Jr., The Decrees of the Demotionidai (1990)], containing resolutions of the Dekeleieis, a phratry within which the Demotionidai (a constituent genos) had a position of special importance—here I follow Wade-Gery (EGH 116–34) and Andrewes (PP, esp. 3–5), against the interpretation of Wilamowitz (AuA II 259–79), which used to be generally accepted. [The view accepted by Ste. Croix has recently been defended by P. J. Rhodes, CQ 47 (1997) 109–20 and contested by S. J. Lambert, CQ 49 (1999) 484–9.] 38 ’ See e.g. FGrH 323 (Cleidemus) F 14; 356 (E ypatrid vn & p
atria) F 1; Jac., ii 142 n. 14. 39 Jac., i 158–60 (on Androt. F 52), argues that no such diacZwism
oB is likely to have taken place in 510–08, and that both it and the alleged increase of foreign b elements in the citizen body referred to in Pol. III 2, 1275 34–7 (see pp. 168–71 below) are ‘political inventions and counter-inventions of the closing fifth century and of the fourth’. Contrast Wade-Gery, EGH 148–9, 151. 37

146

Athenian Democratic Origins

When we turn to what happened in 508/7 onwards we must distinguish two successive situations: first, that under the actual legislation of Cleisthenes, which must have contained some definition of those entitled to become demesmen ab initio and vote at the first assembly of the deme, and secondly, that created by the decisions of the demesmen so constituted—who might, at once or subsequently, alter their own composition, by admitting some claimants whose title to citizenship might be disputed, and rejecting others. Of course, we cannot say for certain what the order of events was; but surely before the first general Assembly of the Athenian State took place under the new constitution, each deme will have met at least once, and at its first meeting, presumably as the first item on the agenda, settled the list of members of the deme: it would then be the collective demesmen who would be entitled to attend the first general Assembly of Athens. At the initial stage (the actual creation of the demes) it is difficult to believe that the definition of those entitled to be registered as demesmen in the first place could have included any but existing citizens, namely phraters. (I see no real alternative to this. I suppose it is just conceivable that the law enacting the new constitution provided for the first deme registers to include all those who were then or had once been phraters; but even this is very unlikely, and any wider measure—including, for example, ‘all free residents in Attica’—seems to me impossible.)40 It is surely at the second stage, by votes of the new deme assemblies already constituted, that any ‘new citizens’ will have been admitted, whether those who had been disfranchised since 510 or those who had never yet obtained citizenship at all. The demes are very likely to have been more liberal in granting admission to their ranks than most of the phratries, for they were devoid of any elements of kinship or of ancient cult, and their votes would be less easily controlled by aristocratic families. One may guess that those who had helped to resist Isagoras and Cleomenes might be treated with special favour.41 And there is an important point to add: it would have been futile 40 Jac., i 160 (lines 17–24) is confused: he does not distinguish sufficiently between ‘the existing citizen body’, i.e. the phraters, and ‘the free population of Attica’—he seems to treat the latter merely as ‘a more neutral term’. 41 Cf. what happened on the return of the democracy in 403: see Tod II 100 ¼ Harding 3 (¼ IG II2 10); P. Cloche´, La restaur. de´mocr. a` Ath. en 403 av. J.-C. (1915), 447–68.

The Constitution of Cleisthenes

147

for Cleisthenes to enfranchise a number of new citizens who were not acceptable to their demes, or to the citizen body in general,42 for they could easily have been excluded at a later date, even if it had been possible to have them enrolled at the initial stage. It is clear, then, that if we speak of ‘Cleisthenes’ as having admitted new citizens, we must remember that this came about as a result, certainly, of his reforms, but probably not directly (to any appreciable extent, anyway) by his actual legislation itself. I shall deal later43 with the question whether Cleisthenes had the deliberate plan of enrolling a substantial number of new citizens. 3. The third great change introduced by Cleisthenes may seem to us more obviously ‘democratic’ in character, though it is hard to say how far this would have been apparent in 508. It consisted in the provision of an entirely new political organisation at local level,44 through the demes, which had their own assemblies and appointed their own demarchs and other officials. As far as we know, there had been no explicitly political organisation before 508 in which the ordinary citizen could have had any chance of making his voice heard and receiving a training in politics and selfgovernment. Clans and phratries were not (as far as we can tell from our evidence) specifically political organisations,45 except in so far as it was the phratry through which a man obtained admission to the citizen body; and with their strongly traditional character they would inevitably be dominated by their most aristocratic members. The old tribes, each roughly a quarter of the whole citizen population, could have offered no field for public activity except to those eupatrids who might aspire to be their Phylobasileis. In some sense the place of the demes before Cleisthenes was taken by the naucraries.46 Of these we know scarcely anything, and very different explanations of their functions have been given in modern times.47 They certainly had financial duties, and perhaps naval and even military ones as well, and their Naukraroi, treated by Aristotle (Ath. Pol. 21.5) as analogous to the demarchs in the 42 We do not know whether the right of appeal to a court of law against removal from a deme register, or refusal of entry to it (as in Arist., Ath. Pol. 42.1), existed from the first. There is no early evidence. 43 Pp. 168–71 below. 44 See p. 141 above on the absence of anything of this kind before Cleisthenes. 45 See pp. 140–2 above and 160–1 below. 46 Arist., Ath. Pol. 21.5. 47 [A projected Endnote on naucraries seems not to have been written.]

148

Athenian Democratic Origins

Cleisthenic constitution, may have been important figures;48 but the essential point is that there is nothing to suggest that the naucraries, a creation of the aristocratic state, resembled the demes in having popular assemblies making decisions by majority vote, in which the ordinary citizen could play a part, nor is it likely that the Naukraroi were elected from below, as the demarchs must surely have been from the first49—doubtless for a year at a time,50 and perhaps (although this may have come later) with restriction of re-election. No doubt many of the demarchs were themselves men of birth and wealth, especially in the early days; but one great democratic principle which we can later see at work in every nook and cranny of the whole Athenian political system must surely have been applied from the beginning, if with less rigour than in later times, to the office of demarch (and minor deme posts): accountability, liability to ey
’ uyna [‘examination of an official’s conduct on retirement’].51 *** The trittys-system was another essential feature of the reforms. Each tribe was made up of three tritt
yeB, one from each of the three districts into which Attica was divided for the purpose: ’
sty, paral
ia, mes
a ogeioB, to which I shall refer as City, Coast and Inland, with the warning that the ‘City’ group of trittyes was many times larger than the area within the walls—it extended as far 48 It is doubtless they, under the style of pryt
anieB t vn & naykr
arvn [‘prytanes of the naucraries’] who are conceived by Herodotus (V 71.2) as ‘administering Athens’ in the late 6th century. 49 We do not seem to know whether demarchs in the late 5th and 4th centuries were appointed in general by vote or by lot: see Busolt-Swoboda, GS II 966–7, esp. 966 n. 4. By c. 300 Eleusis had a demarch chosen by lot (IG II2 1194. 3); and in the 320s the demarch of Peiraeus was also appointed by lot (Arist., Ath. Pol. 54.8), but he was a State magistrate as well as a deme official. [Whitehead, DA 115, regards sortition as certain for the 4th c.; there is still no evidence for the 5th.] It is interesting to note that Euxitheus, the speaker of Dem. LVII, though apparently not of a distinguished family, had been both demarch (§§ 63–64) and phratriarch (§ 23)— but this of course was in the mid-4th century, and the situation may have been different at the end of the 6th. 50 I know of no evidence before the 4th century (see Busolt-Swoboda, GS II 966 n. 4); but this is not surprising, for evidence could hardly be expected. 51 There is very early evidence of ey
’ uynai and ey
’ uynoi in a deme: IG I3 244 (Scambonidai), lines A 5–7, B 7 ff., B 17 ff., an inscription which dates from the first half of the 5th century, probably before the reforms of Ephialtes. For the 4th century, see IG II2 1174 (367/6 BC); 1183 (after 340), lines 16–27; add 1216 (3rd cent.), lines 7, 10.

The Constitution of Cleisthenes

149

as Halimous on the south, Hymettus on the east, Cholargos on the north, and Aigaleos and the headland opposite Salamis on the west.52 Each trittys consisted of a group of demes, which were contiguous in the great majority of the Coast and Inland trittyes—much less can be said of those in the City district.53 In a few cases a single deme constituted a trittys.54 How far the trittyes were intended from the first to have, and how far they ever obtained, any real importance in their own right are questions I need not examine now. Some historians would treat them as little more than links between demes and tribes. Lewis,55 on the other hand, has argued that they were intended from the beginning to have independent importance: he cites most of the texts, to which I would add Aeschines III 30, showing that trittyes (like demes and tribes) might elect men from their own number, apparently to administer public funds in connection with work assigned to the tribes, like digging ditches or building triremes; but the text is 52

See the excellent map in WAW I 13. The division lines ‘break down notably in the city. Why the city is organised as it is we cannot tell, and our questions will produce few profitable answers’ (Lewis, CA 30). Indeed, for the city ‘we are still almost without evidence to determine the principles on which the deme-system was arranged’ (CA 27). But certainly for the Coast and Inland trittyes the statement of Hammond, HG 189, that ‘the demes of a single trittys were for the most part not contiguous with one another’ is the reverse of the truth—which has been known since the time of Loeper and Milchho¨fer (see Eliot, CDA 47–9). The exceptions which everyone must admit nowadays to the rule I have stated in the text, that the demes of a trittys are normally contiguous, are as follows: (1) The Coast trittys (named Myrrhinous) of III Pandionis included the deme of Probalinthos, far to the north, beyond the Coast trittys of II Aigeis. (2) Halimous, on the coast south-east of Phaleron, is a long way from the remainder of the territory of the tribe IV Leontis to which it was attached: according to Kirsten, op. cit. (in n. 2 above) 161, and in Philippson, op. cit. (in n. 2 above) I iii 984, this was the Inland trittys; but Lewis, CA 32, is surely right in following Loeper and attributing it to the City trittys (Skambonidai), a view apparently taken also by Eliot, CDA 137. Kirsten, in his map in Stier’s WAW I 13, shows Halimous as part of the City area. There are other possible irregularities, some of which may be regarded by some people as probable or even certain: for example, (a) Eliot sees the Coast trittys of Leontis as divided into two quite separate parts (see n. 75 below); and (b) some might be willing to accept the likelihood of a gap rather than assume a trittys of very peculiar shape, such as that which Kirsten gives to the Inland trittys of Leontis (with an extension towards the east to make it meet the deme of Hekale which certainly belonged to it). 54 There are two examples which are absolutely certain (Acharnai and Phaleron) and one which is virtually so (Aphidna: the other two demes associated with it are known to have existed before the late 4th century): see S. Dow, in TAPA 92 (1961) 66 ff., at p. 71; Eliot, CDA 151–3, 155, 158 n. 62. 55 CA 35. See also A. E. Raubitschek, ‘The Gates in the Agora’, in AJA n.s. 60 (1956) 279–82. 53

150

Athenian Democratic Origins

unsatisfactory at this point. For my present purposes, however, I need only consider trittyes as a means of grouping demes for inclusion in the tribes. This function is stated by Aristotle (Ath. Pol. 21.4) in the clearest possible terms: Cleisthenes created the ‘ astZ [sc. wyl trittyes o
‘ pvB ek
Z] met
exfi Z p
antvn tvn & t
opvn, ‘in order that each tribe might share in every region (of Attica)’.56 Some modern scholars have paid too little attention to this statement: Bradeen ignores it altogether57 and others have been unwilling to believe that it means exactly what it says. Taking too seriously (as I have claimed)58 the Herodotean picture of Cleisthenes the power-hungry politician, thinking only of himself, most recent historians have wanted to look for ‘fiddles’ in his constitution, in his own interests or those of the Alcmeonid house to which he belonged.59 The trittys-system has been a favourite point of attack. A good example of this attitude is provided by Walker, who believed, for no good reason, that ‘a system more artificial than the tribes and trittyes of Cleisthenes it might well pass the wit of a man to devise’,60 and invented a most ingenious explanation of the trittys-system as a cunning attempt by Cleisthenes to secure ‘that in each of the ten tribes there should be a compact body of voters who were his own special adherents, and owed their position in the body politic to his reforms’—that is to say, the members of the City trittyes, where Walker felt sure the majority of Cleisthenes supporters would be found.61 The fallacy in this argument, which invalidates it completely, is that such a device could have been available only if the Athenians had had a system of group voting, comparable to the Roman one.62 In fact they voted not by tribes but individually, so that it would not have mattered to Cleisthenes 56

Cf. the very similar phrase in Arist., Pol. VII 10, 1330a 15–16. TCR, esp. 23 & n. 7, and 28. 58 See section (i) of this essay. 59 The best explanation of what the Alcmeonid ‘house’ (oi’ k
iZ) was, to my mind, is that given by Wade-Gery, EGH 106–8; but there may also have been a clan known as the Alcmeonidai. [See now Parker, AR 318–19.] 60 E. M. Walker, in CAH IV 143. Cf. Beloch’s statement, referred to on p. 153 below, and the remark in Busolt-Swoboda, GS II 874, made in support of the view that Herodotus believed there were originally 100 demes: ‘Kleisthenes besass offenbar eine Vorliebe fu¨r ‘‘die logisch-arithmetische Konstruktion’’ ’—surely the very reverse of the truth. 61 Walker, ibid. 146–8. The similar theory of Sealey (see n. 63 below) is approved by Day & Chambers, AHAD 119, 179. 62 This has already been pointed out by Hignett, HAC 141, cf. 135. 57

The Constitution of Cleisthenes

151

where his supporters were registered: their votes would have counted exactly the same whether they were members of all ten tribes or of one only. In the assemblies of individual tribes, of course, the distribution of voters might make a difference; but ’ there is no indication that tribal assemblies (agora
i) were ever of any importance63 except in connection with the choice of those few ’ officials (mainly tribal overseers, epimelZta
i) who were elected by their tribes and not—like the tribal taxiarchs and phylarchs, and the Strategoi and other State magistrates—by the whole body of Athenians according to tribes (kat a wyl
aB), that is to say, one (or so many) from each tribe.64 I have already65 expressed disapproval of views which attribute to Cleisthenes the ability to carry out extensive ‘fiddles’ of this sort.66 He may conceivably have been able to manipulate the constitution to some small extent in the interests of himself and his clan and adherents, if only by having the boundaries between demes, or between trittyes (and therefore tribes), drawn to their advantage, or to the detriment of their leading political opponents—if indeed the drawing of such boundaries could be made to serve ‘party political’ purposes at all: no one has ever yet given a satisfactory explanation of how it could, though many have assumed it.67 I do not believe it is in our power to decide whether a particular noble clan or family, Alcmeonids or Philaids or Bouzygai, would prefer the areas in which it mainly resided (or owned land, or exercised influence in one way or another) to be concentrated in one tribe or divided up among two or more tribes. The former seems usually to be taken for granted; but no valid argument has ever been produced in its favour. On the contrary, it 63 R. Sealey (in Historia 9 (1960), at pp. 173–4), who has a theory resembling Walker’s, had to admit that ‘the tribal system did not affect the public assembly’, and so was driven to suppose that the functions of the tribal assemblies ‘may well have been greater at first than later’—hardly a secure basis on which to build a theory about the purpose of the trittys-system. 64 See p. 224 n. 34 below on the meaning of kat a wyl
aB [‘according to tribes’] in Arist., Ath. Pol. 22.2 65 Pp. 133–6 above. 66 A recent statement which exemplifies this fault is that of Thompson, TTAH 406, where the Herodotean picture is accepted and used as a basis for inference that Cleisthenes may have been ‘aiming at political advantage and control of the boule through unfair distribution of population among the phylai and unfair allotment of seats in the boule’. 67 Cf. pp. 158–64 below.

152

Athenian Democratic Origins

might be more advantageous to a distinguished house to have representatives in as many different tribes as possible, because it might then hope to have more than one of its members elected to any given collegiate office the holders of which were chosen kat a wyl
aB [‘according to tribes’]—above all, of course, for the first twenty years the archonships, and thereafter the Strategia. At any rate, on these points there is no evidence, and the fact that the overwhelming majority of Athenians evidently approved of the reforms and acted with remarkable energy and enthusiasm in defence of their new constitution68 suggests that any purely Cleisthenic or Alcmeonid ‘fiddling’—if such there was, apart from measures in the interest of the majority of the Athenians— is not likely to have been substantial. I wonder, indeed, whether the whole conception of the allocation of areas with the deliberate aim of obtaining ‘party-political’ advantage may not be a crude anachronism, quite inappropriate to the time of Cleisthenes. Again, it has often been supposed that the contiguity of the Inland and Coast trittyes within certain tribes must have some special significance. According to Aristotle, Ath. Pol. 21.4, the trittyes were allocated by lot. Some scholars69 have begun by accepting this statement, and have drawn inferences accordingly. Others have rejected it, because it does not fit the conceptions they have arrived at on other grounds, either (like Eliot)70 on the basis of circumstantial evidence, or (like Beloch)71 through conjectural reconstructions of the purpose of the trittys-system. Let us first get the facts right as far as we can. Contiguity of Inland and Coast trittyes has been alleged at one time or another for the tribes IX Aiantis, II Aigeis, V Akamantis and III Pandionis. The last we can dismiss at once: the Coast (Myrrhinous) and Inland (Paiania) trittyes of Pandionis touch, if at all,72 only at a point;73 and, as Lewis has shrewdly noticed, ‘the centre of population of the inland trittys is at Paiania, far to the west’.74 The position of the Coast and Inland trittyes of Akamantis is disputed 68

See Hdts V 72–9; Arist., Ath. Pol. 20.3–4. Most recently Bradeen, TCR 28; Lewis, CA 36. 70 CDA 136–48 (with n. 16). 71 GG I2 ii 327–32. 72 Lewis, CA 35, agrees with Kirsten’s map in Stier’s WAW I 13: the two trittyes just fail to touch. 73 Thus Eliot, CDA 144, with the map on p. 139. 74 Lewis, CA 35. 69

The Constitution of Cleisthenes

153

and doubtful.75 The Coast and Inland trittyes of Aiantis and Aigeis respectively are certainly contiguous over some distance in each case; but Lewis76 has shown that it is easy to attach a wrong significance to this. In the case of Aigeis, he points out that the two trittyes are quite different in character, the Inland trittys being mostly hill country, while that of the Coast consists of a plain, with large villages. The territory of Aiantis, which looks on the map like ‘a solid north-eastern block’, is also a combination of two districts of very different character, with Aphidna, the only important centre of population of the Inland trittys, situated ‘well up in the hills, a long and tedious climb’ from the plain of Marathon, the principal part of the Coast trittys (Tetrapolis). At a time when less was known about the composition of the trittyes Beloch77 could actually argue that the whole reform (which he described as an example of ‘electoral geometry’)78 was the work of Peisistratus. He started from the examples of contiguity (real and supposed) which I have just described, and imagined they were intended to create solid blocks of territory dominated by Peisistratus and his party. It is true that some of the coastal areas, notably Brauron and Marathon (in the Coast trittyes of Aigeis and Aiantis respectively) were ‘Peisistratid territory’; but we have already seen that the corresponding Inland trittyes were quite different in geographical character, and Lewis has also pointed out that Aphidna was the deme of Callimachus, the polemarch at Marathon, and of the Gephyraioi, the family of Harmodius, and therefore not at all likely to be a centre of Peisistratid influence.79 And of course, as I have already shown,80 there is no reason to think it would be an advantage for a leading family to be concentrated in a single tribe (or as few tribes as possible). 75 This is because of uncertainty about the precise location of Phrearrhos. As a result, Lewis (CA 35) agrees with Kirsten in seeing Thorikos, the Coast trittys of Akamantis, as separated from the Inland trittys (Sphettos) of the same tribe by quite a wide belt of the Coast trittys (Phrearrhos) of IV Leontis; whereas Eliot (CDA 91–2 n. 58, 144, and the map on p. 139) has the Coast trittys of Leontis divided into two distinct halves, separated by one block of the Coast trittys of Akamantis, with another piece of the latter trittys just touching that block at one corner and continuing in a north-westerly direction to meet the south-east boundary of the Inland trittys of Akamantis—a shape quite different from that of any of his other trittyes. 76 77 CA 35–6. GG I2 ii 327–32. 78 Cf. p. 150 & n. 60 above. 79 80 Lewis, CA 35–6. Pp. 151–2 above.

154

Athenian Democratic Origins

Bradeen81 has managed to find quite a different kind of significance in the phenomena which interested Beloch. He too sees the areas concerned as Peisistratid strongholds; but, accepting the statement of Aristotle that the trittyes were allocated by lot, he infers that Cleisthenes ‘cared little how the coast and inland trittyes fell’, and therefore must have been primarily interested, not in those two districts, but in the remaining one, the area of the City trittyes. This of course is a non sequitur anyway, and we shall see in a moment that Aristotle’s statement about the use of sortition may not be correct as it stands. *** If we refrain entirely from over-ingenious speculation and simply look at the facts that stare us in the face, we shall find a perfectly natural explanation of the trittys-system. Cleisthenes was confronted with two contradictory desiderata. On the one hand it would obviously be an advantage for each tribe to be as homogeneous as possible, and consist of a compact block of citizens in a single area.82 On the other hand, a certain amount of heterogeneity was highly desirable, for two reasons. First, it would be inconvenient from the military point of view to have most of the hoplites concentrated in a few tribes, and politically it would be most unfortunate if successive sets of tribal prytaneis in the Council had quite different political complexions, with a substantial number of prosperous landowners in one, of small peasants and herdsmen from the hill country in the next, fishermen in the one after, and so on. Secondly, and perhaps more important, it was wise to bring together, within each main unit of the citizen population, men from different parts of Attica who might otherwise have little contact with each other, and thus create ties between them: this would also insure against the development of regional rivalries such as Attica had experienced in the first half of the sixth century. Whether the old div’ to &y ped
ioy [‘those from the plain’] (Pedieis or isions of oi‘ ek Pediakoi), Paraloi (Paralioi) and Hyperakrioi (Diakrioi, Epakrioi) had survived the tyranny to any extent we do not know;83 but of 81

82 TCR 24, 28, cf. 30. Cf. pp. 143–4 above & n. 32. I prefer to use the term given by Hdts I 59.3. How far this threefold division (highly schematised in Arist., Ath. Pol. 13.4–5; cf. Plut., Sol. 13.2; 29.1) really corresponds to the facts of political life in Attica in the second quarter of the 6th century, I feel very uncertain. Too much confidence, in my opinion, is displayed in 83

The Constitution of Cleisthenes

155

course there was always a danger that Attica might once more be divided according to local loyalties, if not on this then on some other basis, unless the tribal system was deliberately designed to prevent it. By means of the trittyes, Cleisthenes introduced the necessary minimum of heterogeneity: in this way he ensured that each tribe would be a fairly representative cross-section of the whole Athenian population, for both political and military purposes. This and nothing else was the essential purpose of the trittys system. The heterogeneity, however, was indeed kept to a minimum. After a full and careful examination of the evidence for three Coast trittyes (those of VII Kekropis, I Erechtheis and X Antiochis, stretching south-east from the edge of the City area almost to Cape Sunium), Eliot concludes that they were ‘natural geographic entities’, having ‘boundaries that followed natural limits’; and on the basis of a more general (but necessarily much more superficial) study of the Cleisthenic units as a whole, he thinks it ‘fair to conclude that most of the remaining coastal and inland trittyes were natural units with easily defined and easily understood boundaries’.84 Here Eliot seems to me to have sufficiently proved his case. In addition he offers two interesting suggestions, which, however, have been strongly challenged. First, he thinks that far from avoiding the contiguity of trittyes within a tribe, Cleisthenes ‘preferred to have the coastal and inland trittyes contiguous within a phyle where it was topographically possible and where these combinations did not run contrary to the creation of ten approximately equal phylai’.85 Secondly, he believes that since the various trittyes even within each of the three groups (City, Coast and Inland)86 the reconstructions of virtually all recent writers except R. J. Hopper, ‘ ‘‘Plain’’, ‘‘Shore’’ and ‘‘Hill’’ in Early Athens’, in BSA 56 (1961) 189–219. It has been argued by F. R. Wu¨st, ‘Zu den pryt
anieB t vn & naykr
arvn und zu den alten attischen Trittyen’, in Historia 6 (1957) 176 ff., at pp. 182–9; and ‘Gedanken u¨ber die attischen Sta¨nde—ein Versuch’, in Historia 8 (1959) 1–11, accepted by J. H. Oliver, ‘Reforms of Cleisthenes’, in Historia 9 (1960) 503–7, that the older classification into Eupatridai, Geomoroi (Georgoi, Agroikoi) and Demiourgoi was still alive in Cleisthenes’ time; but I cannot see that he has any evidence, and his case seems to me entirely unconvincing. 84

CDA 136 ff., at p. 138. CDA 143–4. 86 If the lot was used in the distribution of trittyes, it would not be necessary, in order to ensure equal tribes, for all trittyes to contain an equal number of citizens; but those within each of the three areas (City, Coast and Inland) would have to be roughly equal—every City trittys would have to have about the same number of citizens as every other City trittys, and similarly with the Coast and the Inland areas. 85

156

Athenian Democratic Origins

were clearly far from equal in area and will have been very unequal in population, and since Cleisthenes surely intended the men of each tribe to be roughly equal in number, Aristotle must be wrong in asserting that the allocation of trittyes to tribes was made by lot—Eliot thinks that on the contrary there must have been a careful matching up of, for example, large trittyes with small ones and medium-sized trittyes with others of similar size.87 But both these suggestions have been strongly contested by W. E. Thompson,88 with good arguments as well as some bad ones; and before either of them can be taken as proved, a great deal more work will have to be done, for other Cleisthenic trittyes, of the kind which Eliot has done so well for his three. Meanwhile, I think we should suspend final judgement. If we come to the conclusion that Aristotle (Ath. Pol. 21.4) was wrong about the way in which the lot was used, there is an easy explanation, if we believe (as I do) that apart from Herodotus’ narrative Aristotle is probably relying here on nothing but tradition and inference, or possibly on the statements, similarly based, of one of the early Atthidographers. We could then suppose that the lot will indeed have decided the allocation to particular tribes of the ten divisions of Attica produced by the grouping of trittyes, or (to put it the other way round) the distribution of tribe names to the ten divisions.89 Tradition will have preserved the use of sortition, but at some stage, when the details were not well remembered, this will have been conceived as the allocation to tribes of trittyes instead of combinations of trittyes—a natural enough mistake.90 In some cases the boundary lines of a trittys, as Eliot notes, were drawn with no regard to natural geographical features: this was particularly so ‘in the large plains where the areas contained by natural boundaries were too large for inclusion in single trittyes’.91 Sometimes we find what may be deliberate attempts to draw the boundaries of trittyes (that is to say, of tribes) either in such a way as to cut across certain well-established religious associations between areas which now became Cleisthenic demes, or to give effect to some other purpose at which we can only guess. Lewis92 has 87 88 89 90 91

CDA 140–7, esp. 145. TTAH 402–8. See Endnote ii (p. 173 below). See Arist., Ath. Pol. 21.6, where there is no mention of allocation by lot. Cf. Eliot, CDA 146 n. 16. 92 CDA 138, 140. CA, esp. 30–6.

The Constitution of Cleisthenes

157

detected what he believes to be a whole series of such attempts. Thus, for example: (1) Probalinthos, one of the four constituent parts of the old tetrapolis (all ‘Peisistratid territory’), was put in a different trittys from the other three, which became the Coast trittys (Tetrapolis) of Aiantis; (2) the deme of Hekale was placed in a separate trittys from the nearby ones of the tribes Aigeis, & Kekropis and Antiochis, containing oi‘ p
erij dZmoi [‘the surrounding demes’] which celebrated the Hekalesia at Hekale (Plut., Thes. 14.2); and (3) the four demes of the Tetrakomoi (Peiraeus, Phaleron, Xypete and Thymaitadai, according to Pollux IV 105), were divided up among three different tribes. In each of these cases we find the boundary lines of the trittyes drawn in such a way as not to coincide with pre-existing religious units. Again, (4) Halimous, even if it belongs to the City rather than to the Inland trittys of Leontis,93 is certainly far from any other Leontid deme; and as there is ‘a very primitive Demeter cult on Cape Colias’, Lewis suggests that ‘perhaps here too there is a desire to detach a cultsite from its neighbourhood.’94 Lewis sees in these arrangements a set purpose, ‘an attack on organisations which held a locality by religious ties, some of them in areas attached to political opponents of Cleisthenes’; he thinks ‘the trittyes were constructed in a deliberate attempt to create units which would be sufficiently distinct from existing local units to compete with them and to destroy the influence which they gained from possessing a common cult in a common locality’.95 Lewis perhaps does not say in so many words that what he sees as Cleisthenes’ ‘attack’ on the local cult-organisations was also an attack on the influence obtained by certain important families, who were political opponents of Cleisthenes, from their control 93

See n. 75 above. I have not mentioned the division of the demes involved in the cult-league of Athena Pallenis, on which see Lewis, CA 33–4 and the Additional Note, CA 39. Gargettos and Pithos at least were not included in the same trittys as Pallene; but, as Lewis shows, it is very doubtful what the original extent of this league was, and of course the more demes there were in it, the less significant a division among tribes would be. [On the league see now G. R. Stanton, BSA 79 (1984) 292–8.] 95 Lewis, CA 37 and 34–5 (my italics). Cf. CA 36: local cults ‘survived, because Cleisthenes failed with them’. General assent to Lewis’s position seems to be given by Le´veˆque and Vidal-Naquet, CA 17–18. Since in what follows I entirely reject many of the inferences drawn by Lewis from his material, I should like to emphasise that I have found his contribution most valuable, in clearly establishing many facts out of a confused mass of evidence. 94

158

Athenian Democratic Origins

of these cults (a control derived from holding their priesthoods); but this is certainly implied, and he does say explicitly that the details of Cleisthenes’ settlement ‘could be and were manipulated to the disadvantage of traditional opponents’.96 In my view this whole ingenious theory is misconceived. At first sight it looks most plausible, but it rests upon false premises, and the logic of the argument has not been properly thought out and will not bear close examination. In the first place, it is silently assumed that ‘control of a local cult’ could be an important foundation for the political power of a leading family. Lewis can speak of ‘a family which depended for influence on the control of a local cult’,97 contrasting it with ‘a family of land and wealth’ like the Alcmeonids.98 But I cannot see how in sixth-century Attica a family without ‘land and wealth’ could possibly have had any political influence at all, even if it controlled a local cult. (Such a family would indeed be hardly likely to control a cult of any importance—unless perhaps it had acquired that control at an earlier time when it had been much more prosperous.) And it cannot simply be taken for granted, without the production of evidence of any sort, that the control of cults itself necessarily gave rise to political power. Priesthoods of important cults would surely have been held from the first by families already distinguished, and they might no doubt increase their own prestige and renown (themselves political assets) by exercising such priesthoods; but how could political power be founded 96 Lewis, CA 39. When he adds, ‘The nature of the Alcmeonid position was such that it could not be affected by this settlement’, he is evidently harking back to an earlier passage (CA 37) in which he has distinguished the Alcmeonids from some of their opponents on the ground that they ‘possessed no important local cult of their own that we know of’. A very similar view, based on the material produced by Lewis, is taken by Forrest, EGD 198–9. 97 Lewis, CA 39 (my italics again). To such a family, he thinks, a division among different tribes ‘might well prove fatal’. But what he has been discussing is the division among more than one tribe of the members of the demes which celebrated the cults in question, not of the families which controlled the cults—and splitting the former among different tribes of course in no way involves the splitting of the latter! 98 The Alcmeonids, Lewis thinks (reasonably enough), might welcome the opportunity to have a hand in the affairs of several tribes. Alcmeonids are known in Alopeke (X Antiochis), Agryle (I Erechtheis) and Xypete (VII Kekropis). Leukonoe (IV Leontis) has been suggested by Sealey, but the evidence is not good: see Lewis, CA 23 & nn. 8–11. Eliot (CDA 74 n. 21 and 145 n. 15) believes the Alcmeonids still had ‘influence, though not necessarily residence’ in Aigilia (Antiochis again).

The Constitution of Cleisthenes

159

upon the holding of such priesthoods? Such assumptions may have their place in dealing with the Roman Republic, where religion could often be put directly at the service of politics; but they are not appropriate to the Greek city. No evidence has yet been produced for Athens or any other state of Classical Greece, as far as I know, that the control of cults by individual families was an important source of their political power; and I doubt if any such evidence exists. Even if some such evidence were forthcoming, the essential point I wish to make would remain. It is that whether the units celebrating the cult, or the family controlling it, were placed in one or several of the new tribes, no important difference would be made to any political power or influence which the family in question had (or had not) been accustomed to derive from its control of the cult— which would continue exactly as before, entirely unaffected by the new constitution! Lewis’s assumption that Cleisthenes intended to attack such cults is groundless: there is nothing in any of the new arrangements which would be likely to weaken any local cults, and it is a mistake to see their unimpaired survival as an example of ‘failure’ on Cleisthenes’ part.99 Let us imagine a noble genos who provide the priest of a cult celebrated by four proto-demes (if I may call them that) in one of the country districts of Attica, and let us suppose that in the Cleisthenic reorganisation each of the four resultant demes is given to a different tribe. Would the political position of the genos be affected; and if so, how? I cannot see why the new arrangements should make any important difference. It is not as if all members of each of our group of demes might have been expected to belong before 508 to even a single tribe, let alone a single phratry—as will be clear from what has been said above,100 it is most unlikely that they would.101 Since we can be reasonably certain that the early Athenian regiments were tribal ones, the members of our noble genos (all ex hypothesi 99

100 As Lewis does, CA 35, 36, cf. 22. Pp. 142–3. Forrest, EGD 198–9, has evidently failed to realise this, for when he says that under the new rules many of those affected by one of the cults ‘would belong, politically, to an entirely different unit than those who controlled it’, or speaks of the man whose ‘former master now commanded a different tribal regiment, was represented by a different tribal committee in the Council’ (a situation which he thinks ‘must have transformed the feelings that he once had had for him’), he is gratuitously assuming that under the old regime the great man and the common people would all have been members of the same political and military unit. 101

160

Athenian Democratic Origins

members of one Ionic tribe), will not have been hereditary generals, so to speak, of all the men in the four demes or (probably) even the majority of them; and if and in so far as they were hereditary political bosses of the area they would be so, not because they were members of the same political units (tribes and phratries), but because their wealth and social distinction, reinforced no doubt by their priestly functions, enabled them to exert special influence in the area—and certainly nothing in the Cleisthenic reforms would affect their ability to go on doing so exactly as before. In short, there is no reason why a family controlling a local cult should necessarily have lost one atom of its influence (political or other), either because its own members were divided among different tribes or because the units making up the religious association now belonged to two or three or more new tribes instead of three or four old tribes and even more different phratries! Indeed, might not the acquisition by a leading citizen of influence over men now associated with him politically for the first time in the new Cleisthenic units have compensated, and sometimes more than compensated, for any severance of former ties? Moreover, it is a necessary assumption of any theory such as the one we are examining (although this has evidently not been realised by its proponents) that, but for the allocation of the demes of e.g. the Tetrapolis to two different tribes, the noble family controlling the cult would have belonged, under the new constitution, to the same tribe as the members of all the demes concerned. This assumption is obviously false: the ‘noble family’ would most probably reside in the ‘City’ area,102 and unless there were a special ‘fiddle’ the odds would be five to one against their deme finding itself in the same tribe as either of the two tribes to which the Tetrapolis demes were allocated. The whole idea that Cleisthenes wished ‘to persuade the Athenians that phratry-membership was of no importance’103 also seems to me a great exaggeration. I see no evidence at all that Cleisthenes had any wish to ‘attack’ the phratries, any more than the other cult associations,104 except to the extent that the phratries performed an important political function, namely, 102 103 104

See pp. 163–4 and n. 120 below. Lewis, CA 34–5. Cf. p. 141 and 158–9 above.

The Constitution of Cleisthenes

161

controlling admission to citizenship: this of course had to be withdrawn from them, but otherwise they were entirely unobjectionable.104a It seems to be tacitly assumed nowadays that political activities were organised through the phratries and that the leading men exercised their influence through them. Control of citizenship apart, I very much doubt this, and before accepting it we should ask for some specific evidence. The scattering of phratries over Attica, which we have already noticed, would have made them unsuitable means for the exercise of political functions. My own feeling is that the great men would have exerted political power above all in the districts in which they lived or owned land, and otherwise mainly through their family connections and their ‘ etaire& iai [‘associations’]—compare the situation in the time of Cleisthenes and Isagoras, as represented by Herodotus and Aristotle.105 If they did make use of State divisions, apart from the Assembly, the Council and the tribes, it was perhaps the naucraries which they would most likely have employed.106 Some of the statements in Aristotle, Pol. VI 4, 1319b19–27, to which Lewis appeals,107 must refer more to Cyrene108 than to Athens, for there is no suggestion anywhere that Cleisthenes created any new phratries; it does not seem to be true of Cleisthenes’ Athens that ’ iga kai t a tvn & i’ d
ivn i‘ ervn & [‘private rites’] were reduced ei’ B ol
koin
a [‘to a few public ones’]; and as far as Athens is concerned, ’ . . . ai‘ . . . syn
the words kai p
anta sowist
eon o
‘ pvB an Zueiai diazeyxuvsin & ai‘ pr
oteron [‘and everything possible is to be devised in order to . . . break down the former intimacies’] are satisfied by the breaking down for all important purposes of the old tribal system. Of the clans and phratries and priesthoods Aristotle expressly says (Ath. Pol. 21.6) that Cleisthenes let them be. Evidently in Aristotle’s opinion Cleisthenes had no need to ‘attack’ the phratries. It seems to me that most of the ‘anomalies’ which Lewis has so acutely pointed out in the Cleisthenic trittys-system are likely to 104a [But for the view that phratry membership remained a prerequisite for citizenship even after Cleisthenes see S. D. Lambert, The Phratries of Attica (1993), ch.1.] 105 Pp. 129 ff. above. 106 I have in mind particularly Hdts V 71.2. 107 CA 39–40. 108 See Endnote III (p. 173 below) for a reference to the view of Miss L. H. Jeffery on Hdts IV 161.3.

162

Athenian Democratic Origins

have been due primarily if not entirely to the difficulty of even roughly equalising the membership of the tribes without doing some rather drastic redrawing of the more obvious boundaries. For the Tetrapolis, at any rate, Eliot has a simple and plausible alternative: he demonstrates that splitting up the old four units would have been necessary anyway, for Rhamnous to the north could hardly be left on its own, and with it not more than three of the old units (now Cleisthenic demes) could be combined without creating an unduly large trittys:109 so that Probalinthos (apparently a deme of some size)110 had necessarily to be detached from the other three demes of the Tetrapolis. Even if we feel that this and similar explanations of individual cases, concentrating on purely administrative factors, are insufficient, and we need to find some more serious and positive intention underlying the ‘anomalies’ in the trittys-system, we must be careful how we formulate our theory. All we have a right to say is that some of the boundary lines seem to have been drawn in such a way as not to coincide with existing religious associations—sometimes, perhaps, merely because that seemed the best way of obtaining trittyes of the right size and complexion, and sometimes, perhaps, because it was desirable to avoid creating political units which already had some kind of cohesion among themselves (if not of a political character) and might therefore become the nuclei of separatist tendencies, or of local factions. The unity of Attica was an overriding consideration. To go beyond this, in my opinion, is to indulge in mere guesswork. Lewis may possibly be right in thinking that the giving to a deme of a patronymic name such as Boutadai or Philaidai, which was already the name of a distinguished genos, was a deliberate political blow at the clan in question, since the deme would certainly not be made up entirely of clansmen of that genos.111 (By the mid-fourth century, at any rate, the clan of the Boutadai had renamed itself the Eteoboutadai, ‘the real descendents of Boutes’.)112 However, this seems to me very doubtful. As Lewis admits, we hear from 109 Eliot, CDA 144–5. It remains a puzzle why Probalinthos was placed not in the trittys immediately to the south, that of Aigeis, but in the trittys whose other demes lie beyond that, belonging to Pandionis. Eliot (CDA 144–5 & n. 14) offers a possible explanation. In any event, no better explanation is available on Lewis’s thesis, which demands only that Probalinthos be separated from the other three demes of the Tetrapolis. 110 It had 5 councillors in the 4th century: see Lewis, CA 31 & n. 88. 111 112 CA 26–7. See Dem. XXI 182.

The Constitution of Cleisthenes

163

Plutarch113 that Philaios (the son of Ajax) traditionally settled at Brauron, the site of the Cleisthenic deme of Philaidai; and I should prefer to see the origin of the name of the deme in this tradition. Other demes with patronymic names may have received them for similar reasons. Aristotle certainly believed that some demes were named after those who founded the settlements there.114 Again, I very much doubt whether Lewis is right, except to a very small extent, in saying that Cleisthenes could arrange his geographical units in such a way as to bring ‘distinct advantages for himself and his family’, and that ‘the details of his settlement could be and were manipulated to the disadvantage of traditional opponents’.115 Even if Cleisthenes would have been glad to gain such advantages for himself and his family in so far as he could, we must remember that he was probably not in a position to do very much manipulation of this sort without forfeiting popular support;116 and even more important, it is far from clear how political advantage could be gained in this way.117 Yet another view which one sometimes encounters is that Cleisthenes intended to weaken the political influence of the leading aristocratic clans (the gene). First, in so far as these had exercised political influence through the phratries, by procuring the exclusion or rejection of aspirants to citizenship of whom they did not approve, their political position will automatically have been weakened, as we have seen,118 by the withdrawal from the phratries of the sole right to create new citizens. Secondly, in so far as the new constitution worked better and more democratically than the old one, the influence of the aristocrats (or at any rate of those aristocrats who were not willing to accept an increased measure of democracy) will certainly have diminished somewhat, although individual aristocrats—a Cleisthenes, a Cimon or a Pericles— might still exercise great authority as long as they held the confidence of the demos. Otherwise, I cannot see that Cleisthenes, himself an Alcmeonid and an aristocrat, can have had any desire to impair the position of the aristocratic class as a whole. Bradeen,119 indeed, has asserted the very opposite of this. Noticing 113 114 115 117 119

Sol. 10.3. Ath. Pol. 21.5. [A marginal addendum asked ‘And would name weaken genos?’] 116 Lewis, CA 39. See pp. 133–6 above. 118 See pp. 151–2 and 158–61 above. Pp. 144–5 above. TCR 28–30.

164

Athenian Democratic Origins

that most of the eupatrids lived in Cleisthenes’ ‘City’ area,120 he has seen the main purpose of the trittys-system as the placing in each tribe of a good number of those nobles who for nearly a century yet were to be the leading figures at Athens in politics and war.121 This view derives partly from assumptions about the significance of the contiguity of the Coast and Inland trittyes in one or two tribes which we have already seen to be mistaken.122 Like so many theories about the reforms of Cleisthenes, it overemphasises one particular element in the constitution. Furthermore, to speak of those men from the City area who certainly provided the great majority of the leading political and military figures of the fifth century as if they were all ‘eupatrids’ is wrong. During the tyranny, many non-eupatrids will have risen to prominence, and even if they had not originally lived in or near the city they would surely have moved there as soon as they began to play a regular part in politics, and would have been enrolled in City demes under Cleisthenes’ constitution. But it may well be true that the large size of the ‘City’ area was partly due to the fact that it was thought desirable to have a strong infusion of the principal ‘political families’ in every tribe. To this extent only is Bradeen’s theory justified. *** The significance we attach to Cleisthenes’ creation of a Council of Five Hundred123 depends to some extent upon whether we accept the existence of a Solonian Council of Four Hundred.124 In any event, the new Council had important probouleutic, executive and judicial functions. We know that the judicial powers of the Council 120 ’ ’ ’ o t ’
Cf. Bekker, Anecd. Gr. I 257: eypatr
idai ekalo &ynto oi‘ ayt o asty oi’ ko &ynteB [‘those inhabiting the city itself were called Eupatridae’], a statement which is true only in a very loose sense. One might expect the ‘top people’ in most ’
Greek states, for political and social reasons, to live in or near the asty—cf. Arist., fr. 558 (Rose), ap. Athen. VIII 348ab: most of the wealthy (e y
’ poroi) of Naxos lived ’
in the city (asty), the rest in villages (kat a k
vmaB). 121 To the references given by Bradeen, add R. Sealey, in Proc. of the African Classical Associations. 1 (1958) 61 ff., at pp. 71–8. 122 See pp. 152–4 above. 123 The most complete treatment of the Council and its multifarious duties is that by Busolt-Swoboda, GS II 1021–54. A great deal of information is given clearly and concisely by Jones, AD, ch. V (‘How did the Athenian Democracy Work?’), esp. pp. 105–8, 111–22. See also Hignett, HAC 148–53, 237–44, and Index, s.v. ‘Council of Five Hundred (Kleisthenic)’; P. Cloche´, in REG 34 (1921) 233–65; 35 (1922) 269–95; [Rhodes, AB]. 124 See pp. 83–9 above.

The Constitution of Cleisthenes

165

were wider originally than in the late fifth and fourth centuries, and the same is likely to have been true of its other duties.125 If we think the Council had no existence before Cleisthenes, then its introduction certainly represents a very considerable degree of democratisation of the constitution. Even if we believe in a Solonian Council, as I do myself, there will have been at least one innovation making for a democratisation (choice of councillors by demes), and probably two or three others: (a) Choice of councillors from and by the individual demes.126 Previously, the elections were presumably done by tribes. Some subdivision of the tribe for electoral purposes is possible, but the only subdivisions we ever hear of for any purpose are trittyes, phratries and naucraries,127 and it is hard to believe that any of these gave any scope to their humbler members. Now, with members chosen in the demes, by what seems to be the earliest system of proportional representation for which we have any evidence, there would in many cases be a greater chance for men of relatively humble position to gain access to the Council.128 (b) Restrictions on re-election. It seems to me virtually certain that re-election to the Council was restricted from the first, perhaps always according to the principle for which there is actual 125 For the judicial powers of the Council, see p. 87 above. In favour of a generally larger sphere of competence of the Council in its early years, see e.g. De Sanctis, A2 347 ff.; Hignett, HAC 151–3; and esp. Larsen, RGGRH 5–18 (and 19 for a comparison with the council of the Chian inscription). Larsen (see esp. RGGRH 15–17) believes that a major reconstruction of the powers of the Council took place in 501/0 B.C., to which year he dates the original enactment preserved for us by the fragmentary inscription, IG I3 105. I cannot see how this speculation can be either proved or refuted: it remains a pure hypothesis and should be regarded as such until further evidence turns up. 126 [See Whitehead, DA 266–70 (with some reservations about ‘selection by the deme’, 369–73.] 127 I confess I can make no sense of these alleged subdivisions. The preCleisthenic trittyes (see Hignett, HAC 71–2) are now known to be as real as the phratries and naucraries, but they cannot be identified with the phratries, as in late lexicographic sources claiming to derive from Aristotle (Ath. Pol., fr. 3 Kenyon, Sandys ¼ fr. 5 Oppermann ¼ fr. 385 Rose3 : see Hignett, HAC 59, 71–2), and I am doubtful whether the naucraries were subdivisions of the tribes or had any direct relationship to the phratries or the trittyes: cf. Hignett, HAC 67–74. 128 ’ tZB & wylZB & o‘
lZB [‘from the whole Other officials chosen in the 4th century ek tribe’] were earlier (pr
oteron) selected in the demes, according to Arist., Ath. Pol. 62.1. We do not know who they were, or how far back Aristotle’s pr
oteron takes us, except that Ath. Pol. 22.5 speaks of the sortition of archons from 487 onwards as being made from pr
okritoi [‘pre-selected candidates’] elected by the demes, and this perhaps suggests that election of these other officials in the demes, outright or from pr
okritoi, may go back to Cleisthenes.

166

Athenian Democratic Origins

evidence only from the fourth century, that no Athenian could be a councillor more than twice,129 and perhaps not in successive years. However, it has been argued (by Kahrstedt, for instance)130 that before the middle of the fifth century, when the citizen population greatly increased, such a severe restriction would have made it impossible to find enough men able and willing to be councillors each year, especially if the minimum age limit of thirty always existed, as seems very likely.131 Kahrstedt has suggested for the earlier period a rule similar to that which Athens laid down for councillors at Erythrae in 453/2:132 no iteration within four years. However that may be, a restriction on re-election to the Council is very likely to have been first introduced by Cleisthenes, or made relatively more severe than under Solon’s constitution; and if so, the effect would have been a further democratisation, because the class of councillors would thereby have been widened. (c) Sortition of councillors. It cannot be proved that councillors were chosen by lot as early as the time of Cleisthenes; but there is not a hint in any source that sortition was introduced at a later date. Only twenty years after Cleisthenes’ reforms the archons were being chosen by lot,133 and by 453/2 Athens was ordering the sortition of councillors at Erythrae,134 surely according to her own model. Selection of councillors by lot from the time of Cleisthenes therefore seems very probable. Some modern ’ prokr
itvn writers135 have assumed that this was sortition ek [‘from pre-selected candidates’], from a larger number of candidates previously elected by the demes; but this seems to me unlikely, and there is certainly no trace in any source of there having been two stages at any time in the process of selection of councillors, as there is of archons (Arist., Ath. Pol. 22.5, with 8.1). I would suppose that selection was from those who offered to serve; but of course this remains uncertain. 129

130 Arist., Ath. Pol. 62.3. UMA 135–7. It is now known to have existed as early as about the 480s, for men specially chosen to supervise a festival of Heracles: see IG I3 3, lines 8–10. For the actual Council of Athens it cannot be proved before Xenophon’s time: see Mem. I ii 35; also Dem. XXII, Hypoth. i. Cf. Arist., Ath. Pol. 4.3; 30.2; 31.1. But the fact that a minimum age of 30 years was prescribed by Athens for the Erythrae council in 453/2 (see the next note) strongly suggests, even if it does not prove, that the rule was already in force at Athens. In Aristotle’s day, and very probably from the first, there was a minimum age limit of 30 for dicasts: see Arist., Ath. Pol. 63.3; Poll VIII 122. 132 133 See M/L 40 ¼ Fornara 71 (IG I3 14), lines 8–12. See pp. 215–28 below. 134 135 See n. 132 above. e.g. Hignett, HAC 150, 227. 131

The Constitution of Cleisthenes

167

(d) Admission of Thetes. It is not known for certain whether in the late fifth and fourth centuries the Thetes were eligible in theory for the Council. There is no evidence that they were not, and I think it is generally assumed that they were. This may be right;136 but if so, whether they were made eligible by Cleisthenes or at a later date remains doubtful. The fact that we hear of no later admission of the Thetes must be taken as evidence—though not very strong evidence—in favour of their having been qualified in theory from the first. The admission of the Thetes would not have seemed such a radical step in the time of Cleisthenes as it may appear to us at first sight, for until political pay was introduced very few Thetes could ever have offered themselves for election, let alone attended regularly, in the unlikely event of their being elected; and it may have seemed unnecessary to make any special rule excluding them. It is sometimes denied that Cleisthenes’ Council had a prytanysystem from the first, and Kahrstedt137 would date its introduction as late as 462/1. There is, however, evidence—which seems to have been largely overlooked—that the system existed at least as early as 490;138 and if then, surely from the inception of the Cleisthenic Council139—even conceivably under Solon’s. *** 136 Jones, AD 105–6, supposes that councillors ‘had, like other magistrates, to be of at least zeugite census’. But I know of no evidence that in the late 5th and 4th centuries the Thetes were excluded even in theory from any magistracies except the archonship and the post of Treasurer of Athena: the limitation of all magistracies to the top three classes is specifically Solonian (Arist., Ath. Pol. 7.3), and we hear nothing of it later. 137 BGRF 10–12; cf. UMA 88, 103–5. His arguments seem to me to have no force. 138 Plut., Mor. 628 e-f (¼ Quaest. Conv. I X 3): o‘
ti kai t o c
Zwisma kau’ o‘ toyB ’AuZna
ioyB ej
’ Zgagen [to Marathon] tZB & Ai’ ant
idoB wylZB & prytaneyo
ysZB grawe
iZ [‘that the decree by which he led out the Athenians (to Marathon) was passed during the prytany of the tribe Aiantis’]. I think Plutarch may very well have read this decree in Craterus, as suggested a century ago by Lugebil (see pp. 647–8 of the work cited in n. 21 on p. 219 below). The authenticity of the decree (first mentioned in Dem. XIX 303; cf. Arist., Rhet. III 10, 1411a 10) has been accepted by some and denied by others. I will only point out that the fact that Aiantis was the tribe in prytany at the time is one of four points made by Plutarch in the Quaest. Conv. I x concerning the role of that tribe at Marathon, and that the other three points are correct (see Macan, Hdts IV–VI II 218–19). [See now Rhodes, AB 17 n.4; Rhodes inclines to ascribe the prytany system to Ephialtes.] 139 Here I agree with most recent writers, including De Sanctis (A2 444–5), Hignett (HAC 150 ff.), Larsen (RGGRH 9) and Meritt (AFD 153; AY 203 with n. 4).

168

Athenian Democratic Origins

I wish to deal briefly with the popular but mistaken view that Cleisthenes was particularly concerned to enfranchise a substantial number of new citizens, and that this was actually one of his principal aims. The only authority for anything like large-scale enfranchise’ eteyse ments is Aristotle, Pol. III 2, 1275b34–7: polloyB ewyl
j
enoyB kai do
yloyB [meto
ikoyB], ‘He enrolled in the tribes many foreigners and slaves’140—which I would see, with Jacoby,141 as a reflection of later anti-democratic propaganda. In the only other passage in the Politics which mentions Cleisthenes, namely VI 4 1319b 19–27, there is no reference to new citizens. In the Ath. Pol. the first (and the only explicit)142 reference to the introduction of new citizens comes towards the end of the account of Cleisthenes’ reforms: we are told in 21.4 that Cleisthenes ‘made the inhabitants ’ Zlvn), so of each of the demes fellow-demesmen (dZm
otaB . . . all
that they would not, by using patronymics, expose the new citizens (toyB neopol
itaB)’. These passages provide no authority for the view that an intention to extend the franchise was a main inspiration of Cleisthenes’ reforms. Herodotus has nothing whatever to say on this subject. Those who attach more weight than I do to what Herodotus says about Cleisthenes143 may find his silence on this point decisive. The notion that Aristotle did conceive it as a major part of Cleisthenes’ purpose substantially to enlarge the circle of citizens seems to derive entirely144 from a misunderstanding of Ath. Pol. ’ 20.1, where Cleisthenes is described as apodido yB tfiv& pl
Zuei t Zn 140 The only alternative to bracketing meto
ikoyB—accepted, e.g., by Wu¨st, in Historia 6 (1957) 176 ff., at p. 186—is to assume that the order of the words has been disturbed, and read j
enoyB meto
ikoyB kai do
yloyB. Other modifications of the text which have been suggested (see, e.g., Weil, AH 258) seem to me impossible. 141 Jac., i 159; ii 143. Cf. pp. 144–7 above, and see below. [An endnote attacking J. H. Oliver’s treatment of this passage in Historia 9 (1960) 503–7 has been abolished: Oliver has not found followers.] 142 The explanation I adopt of the origin of the expression m Z wylokrine&in given at the end of Ath. Pol. 21.2 does involve the new citizens: see Endnote IV (p.173 below). 143 Cf. section (i) of this essay. 144 ’ Zsai & Surely no one will wish to treat the words boyl
omenoB ayj t Zn dZmokrat
ian in Arist., Pol. VI 4, 1319b 21–2, as indicating a desire on Cleisthenes’ part to increase the number of citizens: they can only mean ‘increase the strength of the democracy’, and in all the translations I have looked at they are taken in that way.

The Constitution of Cleisthenes

169

polite
ian [‘handing over the politeia to the masses’], and 21.2, & polite
iaB [‘so that more people o
‘ pvB met
asxvsi ple
ioyB tZB should share in the politeia’]. The mistake lies in taking the word politeia in these passages to mean civitas (citizenship) and not respublica (in the sense of having a share in the control and administration of the State).145 In both cases the translation ‘civitas’ makes nonsense of Aristotle’s statement. In the first passage (20.1) this is obvious enough, for no one could possibly have said & that t o plZuoB [‘the masses’] lacked the citizenship and were given it by Cleisthenes. In the second (21.2), the whole clause beginning o
‘ pvB [‘so that . . . ’] depends upon syn
eneime p
antaB ei’ B d
eka ’ i tvn ’ wyl aB ant & tett
arvn, aname& ijai boyl
omenoB, and of course ‘dividing them all into ten tribes instead of four with the aim of mixing them up’ will not support the explanation, ‘so that more would share in the citizenship’. The meaning is clearly, ‘so that more would take (an effective) part in affairs of state’—because the few would now be less clearly distinguished from the many and would have less chance of exercising the kind of control that goes naturally with a conspicuously privileged position in the State. It is extraordinary that several scholars have refused to recognise the plain meaning of this passage.146 I see no other ground for believing that the extension of the citizenship was among Cleisthenes’ principal objectives. As I have already shown,147 it is very likely that one result of the new constitution was the admission to citizenship by the demes of many recent arrivals who either had managed to get themselves enrolled in phratries during the tyranny but had been struck off the phratry registers on the fall of Hippias, or had never had citizen rights at all. These men will have been technically foreigners (j
enoi), and a few may even have been descendants of slaves—there may be this small measure of truth in Arist., Pol. III 2, 1275b 34–7, quoted above. The result of the reforms will (as I think) have been foreseen and accepted as highly desirable by Cleisthenes, but there is no reason at all to think it was his major objective: the reforms can be 145 See Wade-Gery, EGH 147–9; cf. Wu¨st, op. cit. (in n. 140 above), at pp. 184–5. Other clear examples of the use of polite
ia in the Ath. Pol. to mean ‘control of the State’ are in 5.1; 15.3; 29.1, 3; 34.1; 41.1; cf. 9.1 (fin.). 146 Most recently, Day & Chambers, AHAD 111–12, and the translation of the Ath. Pol. by von Fritz & Kapp, ACA 90, and 162–3 n. 47. 147 Pp. 144–7 above.

170

Athenian Democratic Origins

satisfactorily explained without any special emphasis upon this incidental if important consequence. Even if Cleisthenes himself and other progressive politicians were far-seeing enough to realise the advantages that might accrue to Athens from an increase in the number of citizens, is it likely that the bulk of the Athenians of 508 shared their enlightened views? We must remember that only the existing citizens would have had the right to vote on the proposals to change the constitution. How was Cleisthenes able to persuade them? It has recently been suggested148 that in this ‘revolutionary situation . . . , whether the neopolitai were legally entitled to vote or not, there was nothing to prevent them carrying clubs and standing round or even on the Pnyx to shout ‘‘All power to the ten tribes’’ ’. This may well be true—but certainly not, it seems to me, if enfranchisement was a main issue. No one will pretend that those who now lacked the franchise were numerous and powerful enough to force their will on those who remained citizens: it was the voice of the existing citizens which was going to be decisive, and for my part I should have thought that violent action by the disfranchised and others would be anything but likely to commend the constitution of Cleisthenes to the existing voters. If there was a ‘revolutionary situation’ even before Cleomenes intervened, it will have been created by the fact that the principal magistrate (supported no doubt by other leading citizens) was in a position to block reforms desired by the great majority of the existing citizens, by refusing to put them to the vote or otherwise. If we put the question of the ‘neopolitai’ in its true perspective, as an important individual issue which was hardly in the centre of the picture, there is no reason why the ‘neopolitai’ should not have joined the main body of citizens in whatever demonstrations may have taken place to put pressure on a hostile archon. They can certainly be seen as potential supporters of Cleisthenes under the new regime, once it had been established—as Cleisthenes will hardly have failed to realise. That it was Cleisthenes who admitted to full citizenship for the first time men who owned no land149 is a theory which seems to me unlikely in the extreme. As an alternative, it has been supposed150 that the landless were first admitted by Solon. In fact there is no 148 150

149 By Lewis, CA 38. Hignett, HAC 143. e.g. by Busolt, GG II2 273 n. 1; Kahrstedt, SSA 59.

The Constitution of Cleisthenes

171

evidence that the exercise of full political rights at Athens ever depended upon ownership of land, and it is best to refrain from speculation on the point.151 Only citizens, of course, could own land in Attica, in historical times at any rate, but there is no trace of any limitation of the political rights of citizenship to landowners.152 As will be clear from what I have said in an earlier essay in this book,153 I see no reason to believe that Cleisthenes altered the qualifications of the Solonian census classes. *** I have said nothing here about the part played in the Cleisthenic reforms by the watchwords Isonomia154 and Isegoria.155 Whether every citizen who wished to speak in the Assembly was now permitted to do so,156 and whether provision was now made for more frequent meetings of the Assembly than had been the rule under Solon’s constitution and the tyranny,157 and who presided in the Assembly and Council,158 are questions about which we can do no more than guess and advance what seem to us ‘probabilities’ on this side or that. As a result of the political reforms of Cleisthenes, dZmotikvt
era ’ eneto Z ‘ polite
ia [‘the constitution became & S
poly tZB olvnoB eg
much more democratic than that of Solon’] (Arist., Ath. Pol. 22.1).159 Not only had there been considerable economic and social changes during the second half of the sixth century, which helped the democracy to begin its growth to maturity: the very fact that the new constitution of Cleisthenes was a very good one, and not—as so many Greek constitutions seem to have been—a positive brake on further development, was also an essential factor. The constitution 151 Although the contrary is often asserted (e.g. recently by Ehrenberg, GS 40), I do not think it can be supposed with any confidence that Greek cities in early times, apart from colonies, normally demanded the ownership of landed property as a condition of full citizenship. Cf. Last, op. cit. (in n. 20 above) 20–2, summarising and strengthening the demolition by Fraccaro of Mommsen’s theory that the ‘Servian’ tribes at Rome originally contained only landowners. 152 A proposal to this effect by Phormisius after the restoration of the democracy in 403 was defeated: see Hypoth. ad Lys. XXXIV. 153 Pp. 51–6 above. 154 [A footnote refers to a long note which was never written or has been lost.] 155 See G. T. Griffith, ‘Isegoria in the Assembly at Athens’, in ASI (Ehrenberg) 115–30. 156 157 Ibid. 123 ff. Ibid. 124, with 136 n. 43. 158 159 Ibid. 123, with 136 n. 42. See n. 16 on p. 134 above.

172

Athenian Democratic Origins

of Cleisthenes worked, and it proved capable of further democratisation as ever broader sections of the citizen population became capable of political activity. For many of the poorer citizens the giving of political pay made a reality of some rights which Cleisthenes would doubtless have been surprised to find them actually exercising. And we must never forget that the ordinary citizen was now provided with opportunities for activity of a broadly political nature not merely in the supreme organs of state but also within his tribe and—more especially—his deme. As Hopper has said,160 ‘A man who knew his deme business would not be lost in State business. Above all, a man would know his fellows.’

ENDNOTES i (from p. 132 n. 10). Hdts V 69.2, d
eka or d
ekaxa If we read d
eka [‘ten’] in Hdts V 69.2, with the MSS, the obvious interpretation is that Cleisthenes (in the opinion of Herodotus) created 100 demes, ten to each tribe; but the Greek is rather awkward and scholars long ago doubted (see e.g. Grote, HG III 351–2; cf. Busolt, GG II2 405 n.3) whether Herodotus meant to say anything so inaccurate. Then in 1889 Lolling suggested the emendation d
ekaxa [‘in ten groups’], which most people now accept—there is good epigraphic testimony for its use in the required sense: see e.g. M/L 94 ¼ Fornara 166. 33–4. In 1926, however, Busolt-Swoboda, GS II 873, still wanted to read d
eka and even to suppose that Cleisthenes really did create 100 demes; and in 1954 Jacoby could still prefer d
eka (i 66, adding that the passage ‘can only mean that Herodotos’ authority gave him the total of 100 demes’) and speak of ‘the original 100 (?) demes’ (Id. 313). The fact, emphasised in this connection by Lewis (CA 30), that Aiantis seems never to have had more than six demes is decisive for me against there having been originally 100 demes. ‘It is indeed theoretically possible’, as Lewis says, that splitting will have increased the number of demes, but the reverse position will have had to apply to Aiantis—and I find any such major reorganisation incredible. On the 160 Hopper, BAD 17. Pp. 14–17 of this work provide an admirable brief description of the scope of these deme and tribal activities, which may have meant much more to the less exalted Athenian than most scholars have realised.

Endnotes

173

subsidiary question whether Herodotus wrote d
eka or d
ekaxa I have an open mind. ii (from p. 156 n. 88). Thompson v. Eliot It is a pity that Thompson exaggerates the case he is attacking, consistently speaking as if it attributed to Cleisthenes a desire for strictly ‘equal’ tribes; he even says, ‘It would be surprising if Kleisthenes found a way to make representation exactly proportional to population’ (my italics). Eliot, on the contrary, makes it quite clear that he is thinking only of approximate equality (CDA 141, twice over), and he says in so many words that ‘Kleisthenes was certainly not expecting to create ten exactly equal phylai’ (CDA 143 n. 12). Yet in the end Thompson himself (TTAH 408), with the words ‘If Kleisthenes was seeking only a reasonable degree of equality, he could achieve it by using the lot’ (my italics), seems to abandon his own extreme position for a more sensible one, and can be taken as disagreeing with Eliot essentially on the factual question, whether ‘a reasonable degree of equality’ could or could not be achieved by allocating trittyes by lot. The question does not seem to me to be capable of decision in the present state of the evidence. ii i (from p. 161 n. 108). Jeffery on Cyrene Miss L. H. Jeffery, in Historia 10 (1961), at pp. 142–4, makes the interesting suggestion that the three mo&irai [‘divisions’] of Hdts IV 161.3 (referring to the constitution set up by Demonax at Cyrene, somewhere about the middle of the 6th century) may have been not the three tribes themselves but units one of which existed within each of the tribes. If so, as she says, ‘Demonax followed the principle adopted later by Cleisthenes, of making each tribe a cross-section of the total population; but his three basic sections were social, whereas Cleisthenes’ were geographical.’ Her suggestion would gain plausibility if kai & peri Kyr
ZnZn oi‘ t on dZmon kauist
anteB [‘and those who established the democracy at Cyrene’] in Arist., Pol. VI 4, 1319b21–3 (where Cyrene is bracketed with Athens), could be taken to refer to Demonax and not, as commonly supposed, to a later democratic reform at Cyrene, for it would nicely satisfy the requirement kai p
anta ’ o‘
ti m
’ ’ ZloiB sowist
eon o‘
pvB an alista anameixu vsi & p
anteB all
[‘and everything possible is to be devised in order that everyone may be mingled with each other as much as possible’]. I do not feel that the words of Herodotus can easily bear Miss Jeffery’s interpretation. But

174

Athenian Democratic Origins

if she is right, there would be an interesting parallel between the reforms of Demonax and of Cleisthenes. iv (from p. 168 n. 142). ‘m h wylokrine&in’ (Ath. Pol. 21.2) Aristotle follows up his statement that Cleisthenes ‘divided them all into ten tribes, with the aim of mixing them up, so that more people ’ exuZ might take (an effective) part in politics’, with the words o‘
uen el
’ azein t kai t o m Z wylokrine&in pr oB toyB ejet
a g
enZ boylom
enoyB. I would translate, ‘from this came also the catch-phrase, ‘‘No research into tribes’’, in reply to those who wanted to investigate people’s descent’; but other interpretations have been offered of this curious sentence. First, we have to decide what Aristotle meant. Only then can we deal with the further question whether his explanation of the origin of the phrase m Z wylokrine&in is likely to be true. The only piece of external evidence which is of any value is a passage in a speech of Alcibiades in Thucydides (VI 18.2), where Alcibiades, replying presumably to an earlier reference by Nicias to the non’ Hellenic origin of the Egestaeans (tvn & . . . ’Egesta
ivn . . . , andr vn & barb
arvn: 11.7), deplores ei’ . . . wylokrino&ien oiffl B xre vn boZue&in. This gives us one firm conclusion: by the late fifth century, at any rate, the wyl-element in wylokrine&in was clearly not interpreted in the limited technical sense of wyl
Z, a tribe. Crawley translated wylokrino&ien here as ‘pick and choose’; but this perhaps unduly widens and weakens it, for the word may still have conveyed at least some notion of ‘descent’, and something like ‘enquire into ancestry’ is surely better. The expression also occurs in other sources, conveniently listed by Sandys, ACA2 83 n.; but none of them is at all illuminating. 1. Aristotle’s meaning. There seems to me no doubt that Aristotle is giving what he thinks (rightly or wrongly) to be the origin of a catchphrase, a quasi-proverbial expression later used (as in Alcibiades’ speech) in a wider sense. There are either two or three problems here: (a) is g
enZ to be taken (as I believe) in the general sense of ‘descent’ or ‘ancestry’, or does it have the limited technical meaning of ‘clans’, so that m Z wylokrine&in is a reply to people prying into clan membership? (b) Did the wyl-element in the original sense of wylokrine&in have (as I believe) the technical sense of ‘tribe’, or was this word used from the first in the broader sense of ‘descent’ or ‘ancestry’ which it certainly bore by the late fifth century? And if this question is answered in the way I am advocating, then a third question arises: (c) What were the tribes in question—the old ‘Ionian’ tribes or the new Cleisthenic ones or both sets? I will take these questions in turn. (a) Prying into clan membership is quite a conceivable activity of snobbish persons in the years following Cleisthenes’ reforms. But one

Endnotes

175

would hardly reply to such a probe with the word wylokrine&in, which could hardly refer to anything but membership of tribes or descent in a very general sense—surely not clan membership? A second argument against taking g
enZ in the strict sense of ‘clans’ is provided by the context of Aristotle’s remark: he has just spoken of the new tribes, and that part of the sentence in which we are interested continues immediately with o‘
uen—it is from this that the phrase m Z wylokrine&in originated. We must therefore give g
enZ the general sense of ‘descent’ or ‘ancestry’ in which Aristotle does sometimes use it in the Ath. Pol. and elsewhere: see e.g. Ath. Pol. 1 and 16.10; Pol. III 1276a35. (I may add that Day and Chambers, AHAD 113–14, must be wrong in taking the implication to be that ‘after Cleisthenes one could say ‘‘Don’t ask about clans: membership in a deme is the important thing now’’ ’, for the demes have not yet been even mentioned: they come in only at 21.4. They also believe that the object of the question is to expose the neopol
itZB [‘new citizen’]; but this will not do either, for the neopol&itai too are not mentioned until 21.4.) (b) The fact already mentioned, that Aristotle not only makes his statement about m Z wylokrine&in immediately after his statement about the creation of the new tribes but actually represents that innovation and its consequences (the mixing up of the citizens, so that more could take an effective part in politics) as the source (o‘
uen) of the expression, seems to me conclusive in favour of the supposition that in Aristotle’s mind wylokrine&in indeed referred originally to tribes, rather than descent in a general sense. And there is another argument. Well before Aristotle’s day the expression certainly had a much broader meaning. I can see no point in his giving the origin of the expression in the way he does unless it had specific reference to the post-Cleisthenes situation and was only later broadened, in such a way as to become a kind of catch-phrase for use in a variety of situations. Aristotle knew the catch-phrase and felt, reasonably enough, that it must have been used first of all to discourage distinctions in terms of wyla
i, in the strict sense of ‘tribes’, before it became extended, as a proverbial expression, to cover any situation in which undesirable distinctions were being drawn on grounds of descent between man and man, or even (as Alcibiades’ speech shows) between one state and another. (c) To what tribes, then, did wylokrine&in originally refer? Not the new Cleisthenic tribes, for they were from the first on a completely equal footing, and no one would want to draw distinctions between them. But the Cleisthenic tribes, although they were the only ones that counted in any political sense, were not the only Athenian wyla
i: the four ‘Ionic’ tribes still survived—their wylobasile&iB [‘tribe-kings’] retained certain functions (see Arist., Ath. Pol. 8.3; 41.2; 57.4; with other passages cited by Sandys, ACA2, ad loc.), and sacrifices were

176

Athenian Democratic Origins

offered a hundred years later by at least one of them, the Geleontes, and by one of its constituent trittyes, Leukotainioi: see Hesp. 4 (1935) 21, no. 2, lines 30–59. The essential point here, I believe, is that only superior persons, who could aspire to be wylobasile&iB of the ‘Ionic’ tribes or priests of their cults, would have any reason to keep alive their membership of these tribes. As I see it, the situation conceived by Aristotle is this. If you are an Athenian citizen and some snobbish person who likes to ferret out people’s descent says to you, ‘Are you a Geleon, or what?’, you reply, m Z wylokr
inei, ‘No research into tribes’, for under the constitution of Cleisthenes the only tribes that matter are all equal, and no distinctions on a tribal basis are worth making. In time, the expression comes to be used more loosely until it is equivalent to ‘Don’t make distinctions on grounds of descent’. 2. Is Aristotle right? I do not see how we can tell whether Aristotle’s explanation is true. There is certainly no strong argument against it, but we do not know what authority he had for it. It has been suggested to me that to use wylokrine&in is not at all a natural way of discouraging the drawing of distinctions between two different sorts of tribe. To this I can only reply that to me ‘Don’t make distinctions of any kind between tribes’ seems natural enough in the situation we are discussing: it is not a very specific reply, certainly, but it does include the essential point. If the neopol
itZB were not enrolled in one of the old tribes (and he would surely not be), then, in the years immediately following 507, not belonging to one of the old tribes would distinguish him from the other citizens. (Most if not all new citizens would also be enrolled in phratries, and the existing members of each phratry would know that they all belonged to a single one of the old tribes and that the newcomer did not.) Later on, as some (the most aristocratic) of the descendants of the pre-Cleisthenes citizens were enrolled in the old tribes and some were not, the practice of t o wylokrine&in would have resulted in distinguishing a small minority of upper-class old citizens. In both periods I think the reply would have been appropriate enough. I do not see that we can be confident that m Z wylokrine&in was not used from the first in the broader sense in which we find it in the late fifth century, meaning ‘Don’t enquire into ancestry’. But I think we should give Aristotle the benefit of the doubt. May he not have known an earlier text, not now surviving, in which wylokrine&in was used in the way he describes? I have of course ignored any interpretation which depends upon translating polite
ia in Ath. Pol. 21.2 ‘citizenship’ (and not ‘control of the State’), because such a translation seems to me absolutely impermissible: see pp. 168–9 above.

Afterword

177

AFTERWORD Section (i). On the reputation of Cleisthenes see now Hansen. Ways in which Cleisthenes could have got a secretly partisan reform bill past the assembly are indicated by Andrewes. Ste. Croix appears to assume a short chronology whereby Cleisthenes’ system was ready to be voted in by 508/7. The case for a chronology extending the process over several years is re-stated by Badian. Section (ii). Since 1966, there has been much energetic study of the system that Cleisthenes created. The works cited by Ste. Croix in p. 136 n. 2 on the assignment of demes to trittyes and tribes are now superseded by the two studies by Traill. But it must be emphasised that such assignment remains a very inexact science, and a recent scheme published by a careful scholar differs from Traill’s most recent version ‘in some twenty places’ (Stanton, ‘Trittyes’, 205 n. 163); the maps in Traill’s two books are also significantly different. For this reason we will not attempt to gloss those small passages of Ste. Croix’s text (particularly notes 53 and 75) where he engages in detail with questions of deme location and trittys assignment: a note of caution and the general bibliographical indications just given must suffice. At several points Ste. Croix takes over from the scholarly debate of his time the language of ‘contiguous’ and ‘non-contiguous’ demes and trittyes. This should now be read with the qualification introduced by Thompson and accepted (as we know from conversation) by Ste. Croix: demes were human communities, not stretches of territory marked out whether on the ground or on a map, and ‘contiguity’ must not be interpreted in narrow geographical terms. Ste. Croix’s distinction between trittyes that are ‘contiguous over some distance’ and ‘at a single point’ will not necessarily be significant, if for instance the point of contact at that single point is a channel of communication, a road (see below on the ‘road-trittys’). One reason why opinions on trittys assignment are not converging towards agreement is the controversial hypothesis that the distribution of demes between trittyes in many tribes was not primarily determined by geography at all, but by a concern to ensure that the total of councillors furnished by each trittys should be either 16 or 17 (since 17 þ 17 þ 16 ¼ 50, the tribal total). For an account of the emergence of this hypothesis, and a critique, see Stanton, ‘Trittyes’ (and more briefly in ‘Rural demes’). Ste. Croix would doubtless have rejected it vigorously; if accepted it would, however, support his general thesis that Cleisthenes’ motivation was not partisan. Siewert adopts the theory of trittyes equalised for administrative purposes and combines it with the suggestion that the system aimed at military efficiency: trittyes were organised in a way that would allow contingents from the same tribe to converge

178

Athenian Democratic Origins

en route when summoned to muster in Athens. Siewert’s military explanation for the phenomenon of what he calls the ‘road-trittys’ (Wegetrittys) has not been generally accepted: see e.g. the reviews of D. M. Lewis, Gnomon 55 (1983) 431–6; P. J. Rhodes, JHS 103 (1983) 203–4. Important doubts have been raised, on both demographic and archaeological grounds, as to whether the deme–trittys–tribe arrangements imperfectly known to us from the fourth century can go back in their details to Cleisthenes: see in brief Hansen, AD 46–9; Osborne, MG 302– 3 (the latter mentioning the important case of the deme Atene, apparently all but uninhabited in 508). These doubts evidently help Ste. Croix’s case or at least do not harm it. Badian, 450, ascribes all attempts to get back to Cleisthenes’ motives across a ‘totally blank century’ to Micawberish optimism. Ste. Croix follows the usage of his time in speaking of the groups known as gene¯ sometimes as ‘clans’, sometimes as families, and in seeing them as being comparable in wealth and power to what we would call ‘great families’. Bourriot and Roussel have argued that gene¯, religious corporations, are to be distinguished from oikoi, houses or families, and only the latter are to be seen as politically significant; for an account of their innovative work see Parker, AR 56–67. Religious structures and political power would not then on their view have been so intertwined that Cleisthenes would have been obliged to attack traditional cult groupings in order to undermine the old aristocratic state. This conclusion if anything supports Ste. Croix’s case. Kearns argues strongly in support of Ste. Croix’s view that Cleisthenes was not concerned to attack the phratries or old religious groupings such as the Marathonian tetrapolis (which in fact survived him). Roussel, 269–89, points out that Cleisthenes did not substitute a residence criterion of citizenship for a descent one, since membership in the new demes (as Ste. Croix himself in fact stresses) was hereditary; but his treatment is compatible in general spirit with that of Ste. Croix. Rhodes, AB 209–10, takes a much more minimalist view of the early functions of the Cleisthenic boule than does Ste. Croix. The 1984 article of Stanton is a detailed and well-informed treatment of Cleisthenes embodying the approach which Ste. Croix attacks; see too his ‘Rural demes’. Herodotus (V 67.1) claimed that Cleisthenes of Athens modelled his reforms on those of his maternal grandfather Cleisthenes of Sicyon. The possible influence of a different model, the tribes of Corinth (these too sub-divided), is discussed by Salmon. Endnote I. The conclusion that there were originally many more than 100 demes is now generally accepted: see Whitehead, DA 18–21, who

Afterword

179

adopts a figure of 139. The possibility of extracting the required sense from the transmitted text of Herodotus, by appeal to a non-existent ‘distributive’ sense of the cardinal d
eka, is prudently not considered by Ste. Croix. Andrewes, A., ‘Kleisthenes’ Reform Bill’, CQ 27 (1977) 241–8 Badian, E., ‘Back to Kleisthenic Chronology’, in Polis and Politics. Studies . . . M.H. Hansen (2000), 447–64 Bourriot, F., Recherches sur la nature du ge´nos (1976) Hansen, M. H., ‘The 2500th Anniversary of Cleisthenes’ Reforms and the Tradition of Athenian Democracy’, in Osborne/Hornblower, RFP 25–38 Kearns, E., ‘Change and Continuity in Religious Structures after Cleisthenes’, in Crux, 189–207 Roussel, D., Tribu et cite´ (1976) Salmon, J., ‘Cleisthenes (of Athens) and Corinth’, in P. Derow and R. Parker (eds.,), Herodotus and his World (2003), 218–33 Siewert, P., Die Trittyen Attikas und die Heeresreform des Kleisthenes (1982) Stanton, G. R., ‘The Tribal Reform of Kleisthenes the Alkmeonid’, Chiron 14 (1984) 1–41 —— , ‘The Trittyes of Kleisthenes’, Chiron 24 (1994), 161–207 —— , ‘The Rural Demes and Athenian Politics’, in W. D. E. Coulson and others, The Archaeology of Athens and Attica under the Democracy (1994), 217–24 Thompson, W. O, ‘The Deme in Kleisthenes’ Reforms’, Symbolae Osloenses 44 (1970) 72–9 Traill, J. S., The Political Organisation of Attica (1975) —— , Demos and Trittys (1986)

5 Cleisthenes II: Ostracism, Archons and Strategoi

(I) THE LAW OF OSTRACISM: ITS DATE AND PURPOSE It has been said that ‘there is certainly no device of ancient statesmanship that will strike the modern reader as more curious than that of ostracism’.1 At first sight, ostracism is indeed puzzling. Many modern scholars have stigmatised it as unjust, inefficient and harmful to the State, and have passed a harsh verdict on the Athenians for inventing it. Ostracism was certainly capable of abuse—as the Athenians themselves came to realise, for after using it at infrequent intervals for some seventy years they never resorted to it again, as far as we know, for the best part of another century during which it remained, so to speak, on the statute book.2 I am going to suggest that on several occasions ostracism did fulfil admirably, according to Greek ideas, the exact purpose for which I believe it was originally devised; that this purpose, again according to Greek ideas, was a reasonable and sensible one; and that most modern historians, misconceiving the purpose of the institution, have failed to make sufficient allowance for its merits. Grote,3 as so often, came very near the truth, but he perhaps did not advocate his views as clearly and eloquently as he might have done, and of course he wrote before the discovery of the one surviving papyrus of the Athenaion Politeia of Aristotle, a fact which may partly explain the failure of later writers to consider his arguments seriously enough. I am not going to pretend that if all the arguments for and against ostracism are weighed up, even as 1

E. M. Walker, in CAH IV 151. Arist., Ath. Pol. 43.5, shows that an ostracophoria could still, in theory, be held down to the last days of the democracy. 3 HG III 368–80, ch. xxxi. Cf. also Sandys, ACA2 89 n.; F. Susemihl and R. D. Hicks, The Politics of Arist. (1894), 415 n. 2

The Law of Ostracism: Its Date and Purpose

181

they might have appeared to the ancient Greeks, the institution as a whole deserves to be praised—as I have said, it could too easily be abused. But there is much more to be said on its behalf than has been generally realised. *** The purpose of the law of ostracism and the date of its introduction are problems which are closely related and must be considered together. Let us begin with the question of the date. Until recently the great majority of historians, rightly ignoring some absurd theories found in late sources,4 were content to accept the attribution of the law of ostracism to Cleisthenes which we find in Aristotle, Ath. Pol. 22.1, 3–4, supported perhaps by Philochorus, FGrH III B 328 F 30,5 and certainly (for what he is worth) by the unreliable Aelian, VH XIII 24, who however contradicts Aristotle by making Cleisthenes the first victim of his own law.6 Diodorus (XI 55.1), who is generally (and I think rightly) supposed to represent Ephorus7 but is believed by Raubitschek8 to go back ultimately to Theophrastus, speaks of ostracism as being introdued met a t Zn kat
alysin tvn & tyr
annvn tvn & peri Peis
istraton [‘after the overthrow of the Peisistratid tyrants’] 4 These attribute the introduction of ostracism to the following: (a) Theseus. For the sources concerned see esp. Jac. ii 116 n. 26; Raubitschek, TO 78 n. 3. That Theophrastus was responsible for this theory was generally accepted down to Jac. i 311–12; ii 116 n. 26. Raubitschek, TO 78 n. 3, has maintained that here the word ’ ostrak
izv [‘ostracise’] is ‘evidently used metaphorically’, but this cannot be proved and does not seem at all evident to me. (b) An otherwise unknown Achilleus son of Lyson (Ptolemy Chennos, ap. Phot., p. 152 [Bekker]: see Carcopino, OA2 10 n. 2). (c) Hipparchus, but only in an obviously corrupt excerpt from Heracleides Lembos, Peri politei vn & I 6–7 (see Carcopino, OA2 7–10). The text is emended in FHG II 209. On Heracleides Lembos, see H. Bloch, ‘Herakleides Lembos and his Epitome of Aristotle’s Politeiai’, in TAPA 71 (1940) 27–39. 5 I feel unable to treat as reliable evidence of what Philochorus actually wrote the ’ amenon [sc. t concluding part of what Jacoby prints as his F 30: arj
o e
’ uoB] nomouet
ZsantoB Kleisu
enoyB, o‘
te toyB tyr
annoyB kat
elysen, o‘
pvB synekb
aloi ’ vn kai toyB w
iloyB ayt & [‘(the custom) which began through the passing of a law by Cleisthenes, when he had brought down the tyrants, so that he might also expel their associates’]. This is certainly not a verbal quotation from Philoch. and may not come from him at all. 6 I think we can safely disregard this notion, in spite of Jacoby’s surprising tendency to think there may be something in it (see Jac. i 124). 7 See e.g. Jac. i 120, 315, 317, ii 115 n. 10; R. Werner, ‘Die Quellen zur Einfu¨hrung des Ostrakismos’, in Athenaeum n.s. 36 (1958) 48–89; F. Schachermeyr, in Klio 25 (1932) 334–47. 8 TO, esp. 93–96.

182

Athenian Democratic Origins

and this (for what it is worth) is most naturally taken to refer to the last decade of the sixth century. Beloch,9 however, and a number of other scholars since his time (including De Sanctis, Kahrstedt, Lenschau, Hans Schaefer, Hammond, Hignett, Raubitschek and Werner)10 have preferred to follow what they have supposed to be the view of Androtion, FGrH III B 324 F 6, that the law was passed at the time of the first recorded ostracism, that of Hipparchus son of Charmus, whom we know from Aristotle (Ath. Pol. 22.3) to have been ostracised in 488/7, i.e. in the spring of 487. Jacoby, adopting the same interpretation of the Androtion fragment, and believing that Aristotle was deliberately correcting Androtion, came to the conclusion that ‘our tradition does not allow of making a final decision between the dates of Androtion and Aristotle’;11 but the arguments he advances in favour of what he assumes to have been Androtion’s view are mainly negative and very weak.12 I believe we can unhesitatingly attribute the introduction of ostracism to Cleisthenes. When we take into account the purpose of the law, to be discussed later, we shall find strong circumstantial evidence in favour of a Cleisthenic date. But the essential point is that there is no good authority at all for any other dating. Hammond’s statement13 that a date in 488 ‘rests on stronger authority’ is the very reverse of the truth. As I shall now show, there is no reason to think that even Androtion accepted a date c. 488. *** In deciding what Androtion meant, it is essential to compare his account carefully with Aristotle’s, which of course was written 9

GG I2 ii 332–3. The first four are cited by Jac., ii 116 n. 29. See also Hammond, HG 221; Hignett, HAC 130, 159–64, 185–6; Raubitschek, OO, TO etc.; Werner, op. cit. (in n. 7 above), esp. 88. 11 Jac. i 124. 12 The only positive one is that ‘the law about ostracism in 488/7 B.C. and that about the appointment of the archons by lot in 487/6 B.C. spring from the same democratic attitude of mind and are quite credible for the first years of the ’eighties’. According to Jacoby (ii 117 n. 38, emphasising ‘the connexion between the two laws’), this argument is ‘perhaps the most conclusive’. I shall show later in this essay that there was nothing specifically ‘democratic’ about the change in the method of election of archons in c. 487, and that there is no need to look for any connection between that reform and the invention of ostracism. 13 HG 221. 10

The Law of Ostracism: Its Date and Purpose

183

later.14 Androtion, speaking of Hipparchus son of Charmus, said
Peisistr
(according to Harpocration) that sygg
enZB men Zn atoy ’ to &y tyr
annoy kai prvtoB & ejvstrak
isuZ, to &y peri t on ’ ‘ ostrakism on n
omoy t
ote prvton & teu
entoB di a t Zn ypoc
ian ’ kai stratZg oB vn tvn & peri Peis
istraton, o
‘ ti dZmagvg oB ’ annZsen [‘he was a relative of Peisistratus the tyrant and was etyr
the first to be ostracised, the law concerning ostracism then having been introduced for the first time on account of their suspicion of those connected with Peisistratus, since he had become tyrant through being leader of the people and general’] . Many scholars, including even Jacoby,15 have supposed that Androtion meant to say the law of ostracism was introduced at the very time it was first used, that is to say (if his dating for the ostracism of Hipparchus was 488/7) in or just before 488, a year or two after Marathon. I shall argue presently that the latter part of Androtion F 6, beginning with the genitive absolute, cannot be a quotation from Androtion or even an adequate summary of what he wrote. Aristotle, after saying (Ath. Pol. 22.1) that Cleisthenes passed new laws with the aim of pleasing the masses (stoxaz
omenon to &y pl
ZuoyB), including the law about ostracism, describes the first ostracism, that of Hipparchus: the Athenians, he says (Ath. ’ Zsanto tfiv& n
Pol. 22.3–4), t
ote prvton & exr
omfiv tfiv& peri ’ ta&iB ’ ‘ et
’ euZ di ‘ ian tvn & en t on ostrakism
on, oB a t Zn ypoc
’ dyn
amesin, o
‘ ti Peis
istratoB dZmagvg oB kai stratZg oB vn ’ ’ inoy t
yrannoB kat
estZ: kai prvtoB & vstrak
isuZ tvn & eke
syggenvn & ‘
IpparxoB X
armoy Kollyte
yB [‘then for the first time they used the law concerning ostracism, which was introduced on account of their suspicion of those in powerful positions, since Peisistratus had become tyrant through being leader of the people and general. And the first of his relatives to be ostracised was Hipparchus son of Charmus of the deme Kollyte’]. He then adds that it was primarily because of Hipparchus that Cleisthenes, ’
wishing to drive him out, had introduced (euZken) the law.16 14

15 See pp. 261–2, 283–4 below. Jac. i 119–24; ii 114–18. On the interpretation of the sentence which follows (oi‘ g ar ’AuZna&ioi . . . prfiao
tZti [‘For the Athenians . . . lenience’]), I agree with R. Seager, in CR 76 ¼ n.s. 12 (1962) 201–2: it is not a second reason for the interval between the passing of the law and its first application, but a clarification of Cleisthenes’ alleged reason for proposing the law—the demos, with its habitual lenience, had allowed Hipparchus to stay in Athens. 16

184

Athenian Democratic Origins

It is when we set Harpocration’s version of Androtion side by side with Aristotle that illumination comes. Aristotle ’ Zsanto tfiv& (a) t
ote prvton & exr
‘ et
’ euZ ‘then for the n
omfiv oB first time they used the law which was introduced’ ‘ ’ ta&iB (b) di a t Zn ypoc
ian tvn & en dyn
amesin o
‘ ti Peis
istratoB ‘on account of their suspicion of people in powerful positions, since Peisistratus’ (c) dZmagvg oB kai stratZg oB ’ t
yrannoB kat
estZ vn ‘had become tyrant through being leader of the people and general’ ’ (d) kai prvtoB & vstrak
isuZ ’ inoy syggenvn tvn & eke
&
‘ IpparxoB ktl ‘And the first of his relatives to be ostracised was Hipparchus’

Harpocration (a) to &y n
omoy t
ote prvton & teu
entoB ‘the law then introduced for the first time’ ‘ a t Zn ypoc
ian tvn & peri (b) di Peis
istraton ‘on account of their suspicion of those connected with Peisistratus’ ’ kai (c) o
‘ ti dZmagvg oB vn ’ annZsen stratZg oB etyr
‘since he had become tyrant through being leader of the people and general’ (d) [at the beginning] sygg
enZB
Peisistr
Zn atoy to &y tyr
annoy ’ kai prvtoB & ejvstrak
isuZ ‘he was a relation of Peisistratos the tyrant and was the first to be ostracised’

I believe that anyone who reads the two passages in conjunction with an open mind will see at once that the version in the Ath. Pol. reads better and makes better sense all round. In particular, there is the expression t
ote prvton & [‘then for the first time’], applied in Harpocration’s supposed quotation from Androtion not (as in Aristotle) to the first occasion on which the law was put into effect, but to the law itself, to the passing of which—an event that ocote prvton & are inapplicable. This curred once only17—the words t
was realised very soon after the rediscovery of the Ath. Pol.,18 but ’ By the expression to &y peri t on ostrakism on n
omoy [‘the law concerning ostracism’] Androtion cannot, in my opinion, have meant that the Athenian law of ostracism was passed earlier than similar laws in other states. 18 By G. Kaibel, Stil und Text des Pol. Ath. des Aristoteles (1893), 174–5; Wilamowitz, AuA I 123 n. 3; Busolt-Swoboda, GS II 884 n. 2 (citing also Wendland and Pohlenz); and most recently by Bloch, in his important review of Jac. i and ii, in Gnomon 31 (1959) 487–99, at p. 493 (see below). Kaibel thought 17

The Law of Ostracism: Its Date and Purpose

185

the point seems to have been missed by Jacoby as well as by those who have wanted to make Androtion say that ostracism was introduced c. 488. As Bloch has pointed out,19 ‘the description of the law in Harpocration as ‘‘then given for the first time’’ is senseless (it cannot mean, as Hignett, op. cit. 159 f., apparently believes, ‘‘given only then [that is, not in 508/07 B.C.]’’)’, whereas the phrase t
ote prvton & has a perfectly acceptable meaning in Aristotle, where it forms part of a statement that the law was used for the first time in 488/7. Bloch’s criticism has been taken further by Sumner,20 who finds no parallel for such a peculiar use of t
ote prvton & as the theory we are now examining accepts, and shows in addition that not only this passage but also those given above under (b) and (c) make better sense in Aristotle’s version. All this creates a serious difficulty, and it is impossible to believe that Harpocration’s text gives anything like an accurate impression of what Androtion said.21 Androtion F 5, in which both Aristotle and Androtion are explicitly cited, may provide us with a useful parallel: here, as Jacoby has shown,22 Harpocration or his excerptors have made nonsense of Androtion—unless, I may add, Androtion was a great deal less reliable as a source for Athenian constitutional history than we have all been accustomed to think, and the errors in F 5 are substantially his.23 It is not as if we could be sure that Harpocration had Androtion in front of him—far from it. Jacoby believed that all the fragments of Androtion in Harpocration are ‘secondhand’; he points out that ‘Plutarch cites him once only (F 34), certainly not directly’, and declares that ‘there are no certain vestiges of Androtion having been read in Roman times’.24 that the apparent fragment of Androtion is nothing but a feeble excerpt from Aristotle, Wilamowitz that the citation from Androtion has been confused with one from Aristotle. 19

See the preceding note. AF6 80–2. 21 It is therefore unprofitable (even apart from the fact that Androt. was presumably writing annalistically under an archon-year) to argue that from the standpoint of Androt., writing some 150 years later, t
ote need not have a precise meaning and could easily be stretched to cover 20 years, so that Androt. may equally well have believed the law of ostracism to be Cleisthenic: for this view see Carcopino, OA2 95, followed by D. Kagan, in Hesp. 30 (1961), at pp. 394–5; contrast A. R. Hands, in JHS 79 (1959), at p. 69. 22 Jac. i 117–19, following Wilamowitz, AuA I 52 n. 19. But we must not press the analogy between Androt. F 5 and F 6 too far: see n. 18 above. 23 24 Cf. pp. 288–9 below. Jac. ii 102 n. 130. 20

186

Athenian Democratic Origins

Harpocration himself, therefore, may have had only an already abbreviated and perhaps inaccurate version of Androtion. Is it possible to discover what Androtion said? One plausible explanation has been offered by Dover and by Sumner, writing independently but coming to much the same conclusions.25 This interpretation sees no conflict of opinion between Androtion and Aristotle. On the contrary, Aristotle reproduced the very words of Androtion, and we may ‘regard Ath. Pol. 22.3 as indicating what Androtion really said about ostracism in connexion with the ostracizing of Hipparchos’.26 He had already mentioned the institution of ostracism at the proper place, when dealing with Cleisthenes. Harpocration (or his excerptor), having no occasion to go back to Androtion’s account of Cleisthenes, did not ‘realize that a contrast ’ Zsanto and et
’ euZ was intended or that et
’ euZ had a between exr
‘‘pluperfect’’ sense’—hence his condensed paraphrase t
ote ‘ prvton & teu
entoB, ‘assuming that the purpose of the clause oB ’ euZ ktl. was solely to state the reason for the law. If he asked et
himself what the point of ‘then for the first time’ was, he may have assumed that another law of ostracism was made at a later date, or that Androtion was controverting a view that ostracism had been invented much earlier; these, however, are questions which would not have interested Harpocration.’27 I think this is probably the best explanation available. Its one drawback is the supposition it entails that Aristotle followed Androtion very closely indeed, copying him almost word for word. This may indeed have happened.28 But I think we may also allow for the possibility (suggested at the end of the last century)29 that Harpocration or his excerptor has simply made a mistake in his attribution and is in fact giving a rather poor summary of Ath. Pol. 22, or a condensed conflation of that passage with a parallel one in Androtion.30 25

26 Dover, AO; Sumner, AF6. Sumner, AF6 83. Dover, AO 257. 28 The theory would have the advantage of providing a reason why Arist. in Ath. Pol. 22.4 narrows his usual explanation of ostracism to the fear that a particular man might become tyrant—cf. pp. 192–8 below. 29 See n. 18 above. 30 It is worth noticing that Harp. twice (s.v. d
ZmarxoB and naykrarik
a) cites Ath. Pol. 21.4–5, very close to ch. 22—of course he (or his excerptor) may have looked this up at a later time than when he was dealing with Hipparchus, or have relied on a different excerptor. I cannot agree with Jacoby (ii 114–15) when he draws 27

The Law of Ostracism: Its Date and Purpose

187

Of course it is quite possible that both Androtion and Aristotle were following an earlier source—Hellanicus, perhaps, or Cleidemus. But that is a speculation which will not help us to go any further. There is not the slightest reason to suppose that Aristotle is deliberately correcting Androtion; but it would be wrong, with Raubitschek31 and Kagan,32 to draw any conclusions about what Androtion said from the fact that Aristotle does not explicitly contradict him. We must not expect Aristotle to cite earlier authors with whom he disagrees: even when he held a view very different from Androtion’s on such an important question as the nature of the Solonian Seisachtheia, he did not bother to contradict the specious explanation previously advanced by Androtion.33 Since I am allowing for the possibility that Androtion may have said something entirely unknown to us about the origin of ostracism, I should perhaps just refer to Jacoby’s demonstration that Androtion might have had his own reasons for denying the ascription of ostracism to Cleisthenes,34 just as he denied35 that Solon’s Seisachtheia was a cancellation of debts. Androtion was among the more conservative Athenians,36 and men whose political outlook resembled his liked to assimilate the constitution of Cleisthenes to that of Solon, distinguishing both from the radical-democratic regime of the late fifth century.37 But, as Jacoby says,38 ‘ancient a contrast between Androt. F 5, where Arist. and Androt. were ‘cited for different matters, the former for an Athenian office, the latter for its history’, and F 6, where the lexicographer ‘consulted a historical source only’ and ‘had no reason for consulting also the ’Aup.’. It is a gratuitous assumption that the lexicographers would refer only to the most obvious works. As Jacoby says (ii 230 n. 11), Harp. ‘often puts side by side quotations from Arist. and another (frequently Atthidographic) author’. I fancy there are at least as many quotations by Harp. from the historical section of the Ath. Pol. (with or without another source, such as Androt. or Philoch.) on questions which are historical rather than constitutional—see e.g. Harp., s.v. dek
azvn and tetrak
osioi; Schol. Ar. Ach. 234; Lys. 666, 1153; Ran. 1532; Schol. Lucian., Tim. 30 (p. 116.4 R.). 31

32 OO 221. Op. cit. (in n. 21 above) 395. FGrH 324 F 34, ap. Plut., Sol. 15.3–4; Arist., Ath. Pol. 6.1: see Jacoby’s excellent commentary, i 144–6, cf. 99. 34 35 Jac. i 120. See FGrH 324 F 34: see n. 33 above. 36 See Jac. i 86–106, esp. 96; ii 91–3 n. 86. See N. H. Baynes, Byzantine Studies and Other Essays, 144–67 (1955) (ch. viii) for an admirable analysis of the outlook of Isocrates, which in some ways was closely comparable to Androt.’s. 37 On Arist., Ath. Pol. 29.3, see Jac. i 120; A 206, 384 n. 30. 38 Jac. i 120. 33

188

Athenian Democratic Origins

historians universally regarded ostracism as a typically democratic institution, which took its origin from the suspicion (or the envy) felt by the Demos towards the privileged classes. Anybody seeing in Kleisthenes not the father of radical democracy, but the restorer of the Solonian, the man whose constitution was oy’ dZmotik Z ’ a paraplZs
ia tfi Z& S
all olvnoB [‘not populist but similar to that of Solon’],39 would wish to dissociate him from that radically democratic institution as he dissociated Solon from the revolutionary measure of a complete cancelling of all debts.’ To sum up—it is not at all certain what Androtion said about the date of introduction of ostracism, but at any rate there is no positive reason to suppose he differed from Aristotle, and a very plausible case can be made for identifying their views; and even if Androtion did dissociate the law from Cleisthenes (which to me seems unlikely), he might easily have done so in obedience to his own political beliefs rather than in reliance upon any evidence. Those who feel as uncertain as Jacoby whether Aristotle had any good authority for attributing the law to Cleisthenes must at least admit that, such as it is, ancient authority is strongly in favour of a Cleisthenic date. *** The only other argument I have been able to discover against attributing the law of ostracism to Cleisthenes is the one advanced originally by Beloch40 and repeated by various other writers: that ‘such a weapon is not forged to be left for twenty years in the sheath’. Hignett41 quotes this remark with approval, adding ‘If the law was first applied early in 487, the obvious inference is that it had been passed not long before’, in fact ‘inside the year 488’. Such statements have little persuasive force anyway, and none at all if we decide (as we shall find good reason to do later) that the law of ostracism was not intended primarily for use against Hipparchus son of Charmus or some other ‘prospective tyrant’, but to meet a particular situation which had occurred recently but apparently did not arise again in acute form until after Marathon. To my mind, Beloch’s statement is entirely unwarranted and Hignett’s inference the reverse of obvious. Those who think along these lines should reflect that only twenty years separate the reforms of 39

Arist., Ath. Pol. 29.3.

40

GG I2 ii 332.

41

HAC 160.

The Law of Ostracism: Its Date and Purpose

189

Cleisthenes from the first recorded ostracism, and that after the batch of ostracisms in the 480s there are quite long periods during which there may have been no ostracophoriai at all, or at any rate none in which the necessary quorum of 6,000 votes42 was obtained—above all, 443 to c. 417, a span of 26 years. Why should not the Athenians design an institution for use in particular circumstances which they had recently encountered and might reasonably expect to encounter again, but which did not exist in a sufficiently acute form at the time or for twenty years afterwards? Parallels can be found.43 (Of course, we do not know for certain that there was no single ostracophoria before 487: there may have been, and the quorum of 6,000 may not have voted.) There is no other argument, as far as I know, against attributing the law of ostracism to Cleisthenes. Hignett, indeed, produces none, although he can speak of ‘the shipwreck of all hypotheses which in one way or another have tried to maintain the ascription of ostracism to Cleisthenes’.44 Similarly, there is no argument in favour of a date c. 488, except the worthless one we have just been examining and a faulty interpretation of Androtion F 6. The removal of the alleged conflict between Androtion and Aristotle makes it impermissible to claim that the statements of both writers must have been ‘based on inference and nothing more’.45 Even if it is not likely that the text of the law was seen by Aristotle or the Atthidographers, Aristotle’s statement may rest upon a correct and still unanimous tradition. *** I turn now to consider the purpose of the law of ostracism. Several ancient writers whose works survive have expressed their views on this question. Before I discuss their opinions, however, I wish to mention a theory which entirely lacks ancient evidence in its support but has been advanced in recent times by at least some of those who believe ostracism to have been introduced c. 488: namely, that it was intended from the first as a weapon of ‘party-political’ struggle, by which ambitious politicians could get rid of their 42

Cf. pp. 210–11 below. See e.g. Bloch, op. cit. (in n. 18 above) 493, and Busolt-Swoboda, GS II 884 n. 2. 44 HAC 164. 45 Day and Chambers, AHAD 13 (italics in the original). 43

190

Athenian Democratic Origins

rivals. This is the view, for instance, of Beloch,46 who thought the Alcmeonids were responsible, of Hammond,47 for whom the law was essentially anti-Peisistratid, and of Hignett,48 who believed the law was devised by ‘the anti-Persian leaders’. (I am ignoring some variants which seem to me very improbable and have not found favour, such as that of Raubitschek,49 who would bring Cleisthenes back from retirement after Marathon to introduce the law.) No form of the theory we are now considering seems to me tenable. After the time of Cleisthenes, who probably did not long survive his triumph in 508/7,50 until that of Cimon in the 470s and ’60s, it seems very unlikely that any Athenian was ever in a position of unchallenged predominance—Themistocles, perhaps, for two or three years before the invasion of Xerxes; but the literary evidence and the high proportion of ostraca cast (or intended to be cast) against Themistocles51 among those which have survived to modern times (over a third of the whole, twice the number against anyone else) suggest that the political position of this remarkable man was always insecure. And I entirely agree with Hands52 that ‘whoever was responsible for the institution of ostracism chose a time when he believed that he was strong for its introduction’. Certainly the acceptance of ostracism as a device invented by a leading politician for getting rid of his rivals one by one hardly fits Cleisthenes himself, whose constitution (as I have pointed out already)53 might have been expressly designed to prevent one man from gaining supreme power at Athens. It is understandable, therefore, that those who have held the theory we are discussing should have preferred to date the introduction of ostracism after Marathon, at the time when it was first employed. But let us look at this theory more closely. Its proponents have failed to realise its implications. Those who devised the institution—the Alcmeonids, the anti-Peisistratid or anti-Persian leaders, or whoever—must 46

47 48 GG II2 i 29–30. HG 221. HAC 186. OO; contrast C.A. Robinson, in AJA 56 (1952) 23–6. 50 His parents were married in the 570s, and if he was the eldest son he will probably have been 60 or more at the time his reforms were passed. Modern scholars sometimes speak of his ‘strange disappearance’ or of his ‘fall’. We cannot of course exclude such explanations, but it seems simpler to suppose that he died soon after 507. 51 Even in the 480s: see Hands, op. cit. (in n. 21 above) 77. [But the new Kerameikos material contains 4647 ostraka cast against Megacles; Themistocles comes second with 1696. See the Afterword.] 52 53 Op. cit. 69. Pp. 133–4 above. 49

The Law of Ostracism: Its Date and Purpose

191

have been able to convince the Assembly that it was desirable, and they could hardly do this by giving false reasons for the reform: ostracism is not an institution which automatically recommends itself by any obvious merits. But if the inventors of ostracism had enough popular support to get their measure passed by a considerable majority of votes, then they surely did not need its aid, for they were already in a position to obtain the backing of the Assembly for whatever policy they desired. And if, on the other hand, they were only able to command a narrow majority in the Assembly, they themselves would be in grave danger of falling victims to their own device, should the balance of opinion in the Assembly swing only slightly against them—always a possibility which the Athenian politicians had to reckon with, since the composition of the Assembly might vary greatly from meeting to meeting.54 The hand of Themistocles has often been suspected behind the introduction of ostracism; but that sagacious politician, however willing to make use of ostracism when he himself was in a strong position, would surely have been the last to invent such a danger for himself—unless, of course, he was in a much stronger position than I have imagined; and if he was, he would not have needed to have his chief opponents ostracised. Besides, there is not a suggestion in any of our sources that Themistocles had anything to do with the invention of ostracism; and if he had, it would surely have been remembered and told of him, for it was just the kind of action which would have been thought characteristic of him. Rather than see the ostracisms of the 480s as part of a plan by Themistocles and his political allies to get rid of their chief enemies, I would prefer to adopt a suggestion made by A.R.W. Harrison55 and A.R. Burn56 that the ostracophoriai in question were a series of unsuccessful attempts to get rid of Themistocles, which recoiled on the heads of his opponents. Ostracism, as Harrison points out, was a very tricky and two-edged weapon, much more likely to be advocated by the opponents of an up-and-coming statesman with strong personal views than by such a man himself. The initiator of an ostracophoria would have to be very certain of the discipline of his supporters before embarking upon a procedure where any votes split between two or more of his opponents would increase the chances of his 54 55 56

See Jones, AD 109–10; Finley, AD 10 ff. In an unpublished paper which he has been kind enough to show me. Persia and the Greeks (1962) 290.

192

Athenian Democratic Origins

own ostracism. Is it likely that Themistocles submitted himself voluntarily to this risk five times in six or seven years, when he was himself the most obvious ‘candidate for ostracism’? *** The general consensus of opinion in ancient times about the purpose of ostracism is summarised by Jacoby in the statement quoted on pp. 187–8 above: ‘ancient historians universally regarded ostracism as a typically democratic institution, which took its origin from the suspicion (or the envy) felt by the Demos towards the privileged classes.’ For present purposes we can discount the few sources which speak of envy: they usually give it only as a subsidiary motive.57 And we need not concern ourselves with various allegations of some specific motive as having been at work in the ostracism of a given individual, however true or false they may be in the particular case: for example, that Hyperbolus & p
was ostracised di a ponZr
ian kai ai’ sx
ynZn tZB olevB [‘because of his wickedness and the shame he brought on the city’]58 or Aristeides when Themistocles put it about that he had obtained ’ a monarx
ian adoryw
orZton [‘monarchy but for the absence of a ‘ megalopr
agmvn kai wilot
yrannoB bodyguard’];59 or Damon vB ‘ [‘for being a great plotter and friend of tyrants’];60 or Cimon vB wilol
akvna kai mis
odZmon [‘for being a lover of Sparta and ’ fi Z& tfi Z& hating the demos’],61 or di a paranom
ian, o
‘ ti tfi Z& adelw ‘ eayto &y syn
fivkZse [‘because of his irregular behaviour in cohabiting with his own sister’].62 Two ancient writers whose works survive expressed themselves repeatedly and at some length on the subject of ostracism and the ends it was designed to achieve: Aristotle and Plutarch. Apart from his invaluable description of the procedure of ostracism,63 Plutarch64 has nothing essential that is not also in Aristotle, who can be taken as the representative spokesman of antiquity on the 57 e.g. Plut., Arist. 7.2; Them. 22.4; Alcib. 13.6; Poll. VIII 20. Demetr. Phal., ’ FGrH II B ii 228 F 43, ap. Plut., Arist. 1.2 (epiwu
onoiB), seems to have stressed envy (contrast Plut., Arist. 1.7 etc.). This is the motive emphasised above all by S. Ranulf, The Jealousy of the Gods and Criminal Law at Athens I (1933), 132–42. 58 59 60 Thuc. VIII 73.3. Plut., Arist. 7.1. Plut., Per. 4.3. 61 62 63 Plut., Per. 9.5. Ps.-Andoc. IV 33. Arist. 7.5–6. 64 The principal passages in which Plut. gives his views on the purpose of ostracism are Arist. 1.7; 7.2–3 (cf. § 4); Them. 22.4; Nic. 11.1 (cf. §§ 6–7); Alc. 13.6 (cf. § 9).

The Law of Ostracism: Its Date and Purpose

193

subject—there is indeed virtually nothing in the other ancient writers which is not found also in him. It is commonly said nowadays that Aristotle (not to mention most other ancient writers) thought ostracism was designed to be used against a potential tyrant; and the view thus attributed to him has been accepted by a large number of modern scholars, including Bengtson, Busolt, Calderini, Carcopino, Cloche´, Gilbert, Glotz, Holm, Meyer, Walker and Wilamowitz.65 Most recently Kagan66 has taken up much the same position, except that he persists in speaking of a ‘tyrannist party’, the existence of which depends entirely on conjecture.67 (All the writers I have just mentioned also accept Aristotle’s attribution of the law of ostracism to Cleisthenes.) In fact the notion that ostracism was mainly if not solely intended for use against a potential tyrant is an unfortunate oversimplification of Aristotle’s position. We shall do well to look carefully at everything Aristotle has to say on the subject. In his account of the origin of ostracism in the Ath. Pol., Aristotle says four things directly or indirectly about the original purpose of the law: (1) that Cleisthenes had the aim of pleasing the mass of the people (stoxaz
omenon to &y pl
ZuoyB: Ath. Pol. 22.1);68 (2) that the measure was passed ‘on account of their suspicion of ’ ta&iB dyn
those in powerful positions (tvn & en amesin), Peisistratus having become tyrant through being leader of the people and general’ (22.3, the ‘suspicion’, therefore, being that another might follow the same course); (3) that the law was particularly directed against Hipparchus son of Charmus, whom Cleisthenes wished to drive out (22.4); and (4) that for three years (these will be 65 Bengston, GG4 145; Busolt, GG II2 439–41; GS II 884–6; Calderini, O, esp. V, IX; Carcopino, OA2 , esp. 15–36; P. Cloche´, La De´mocratie athe´nienne (1951), 27–8; Gilbert, CASA 151–2; Glotz, GC 170; HG I 469, 478–9; Holm, HG I 424; Meyer, GdA III2 746; cf. IV2 i 320; Walker, in CAH IV 142, 151–3: Wilamowitz, AuA II 87. 66 Hesperia 30 (1961) 397–8. 67 Androt.’s reference to oi‘ peri Peis
istraton does not course imply the existence of anything that can be called a ‘tyrannist party’; and ostracism is an institution very ill-fitted to attack a ‘party’. 68 The latest writer to defend the accuracy of the account of the origin of ostracism in the Ath. Pol., namely Kagan, Hesperia 30 (1961), esp. 397, has to ignore this statement, which of course is dead against his attempt to represent the lower classes as a probable chief source of support for a new tyrant—the very person against whom the chief law is represented by the Ath. Pol. as being aimed! Arist. is assuming that the Demos was glad to possess ostracism as a means of preventing a tyrant from coming to power.

194

Athenian Democratic Origins

488/7 to 486/5) the Athenians continued to ostracise ‘the friends of the tyrants, on whose account the law had been enacted’, but in the fourth year (485/4) they began to remove ‘anyone else who appeared to be too great’, and Xanthippus was the first to be ostracised of those unconnected with the tyranny (22.6). The impression most people have got from these statements is that Cleisthenes was afraid of Hipparchus and other members of his circle following in Peisistratus’ footsteps, and rising through political and military eminence to a position from which they might aspire to tyranny. This impression has been reinforced by what we know as Androtion F 6, which I would regard as either the source of Aristotle’s statement about ‘suspicion of those in powerful positions’ or derived from it,69 and by Philochorus F 30, the unreliable last sentence of which asserts that Hyperbolus was the one man to be ostracised ‘from among the men of no repute, because of the unpleasantness of his behaviour, not the suspicion of tyranny’ (implying that suspicion of tyranny was the normal reason for ostracism), and goes on to say that Cleisthenes introduced ostracism ‘in order to expel the friends of the tyrants as well’.70 One or two other ancient writers, Diodorus for instance,71 also accepted the view that the function of ostracism was to check the rise of a potential tyrant. But did Aristotle really mean to represent Cleisthenes as designing ostracism for use specifically against a potential tyrant; and if he did, was he right about Cleisthenes’ objective? When we turn to the Politics we find, as we might expect, that Aristotle’s analysis of ostracism is much more subtle. He begins (Pol. III 13, 1284a 17–22) by saying that ostracism is the method by which democratic states, in their pursuit of equality (i’ s
otZB), keep in their place those who appear to have too much political power or ‘ ’ influence (toyB doko &yntaB yper
exein dyn
amei di a plo &yton Z ’
’
tina allZn politik Zn i’ sx
yn); and he adds polywil
ian Z (1284a 36–7, with 30–6) that democracies need to check these ‘ over-powerful men (toyB yper
exontaB), just as tyrants and oli’ aB [sc. garchies do. Even regimes of the right sort (kai t aB oru
polite
iaB]) may need to act in such ways (1284b3–7). And so 69 70 71

See pp. 182–8 above. This may be simply a recollection of Arist., Ath. Pol. 22.6 (cited above). Diod. XI 55.2 (cf. § 3); 86.5; 87.1 (cf. § 2).

The Law of Ostracism: Its Date and Purpose

195

Aristotle comes to the conclusion (1284b15–22) that ‘in cases of palpable undue prominence the argument for ostracism has a ‘ ‘ ’
certain political justice’ (kat a t aB omologoym
enaB yperox aB exei ’ ti d
ikaion politik on o‘ l
ogoB o‘ peri t on ostrakism
on). It would of course be better, he says, so to frame the constitution from the start that it would not need this remedy; but the next best course is, if the need arises, to try to correct the constitution by some such means. This is not, however, what has happened in practice: Greek states have not concerned themselves with what their own particular constitutions needed, but have used ostracism stasiastikvB— & that is to say, in what we might call a ‘party-political’ manner, but with that suggestion of disregard for law which is so often present when the word st
asiB and its compounds are used by Greek writers. Aristotle returns to the subject of ostracism later, when he is considering the causes of st
asiB (Pol. V 3, 1302b 15–21). One way ‘ in which stasis may come about, he says is di’ yperox
Zn [‘through superiority’]: when one man or a group within the State is tfi Z& ’ kat
dyn
amei me
izvn . . . Z a t Zn p
olin kai t Zn d
ynamin to &y polite
ymatoB, by which he means undue power or influence, disproportionate in relation to the State as a whole and the strength of the general body of citizens. Usually, he goes on, the result is ’ dynaste
ia: one-man-rule or a close hereditary olimonarx
ia Z garchy. This is the reason why some states, such as Argos and Athens, make use of ostracism. But it is better, Aristotle again concludes, to see to it from the beginning that no such over’ esontai toso &yton powerful men can exist (o
‘ pvB m Z en
‘ yper
exonteB) than to have to deal with them afterwards. Later again (V 8, 1308b 16–19), after saying once more that it is best so to frame the constitution that no one can gain a position of undue prominence through his associates or his wealth (v
‘ ste mZd
ena ’ ignesuai poly yper
‘ egg
exonta dyn
amei m
Zte w
ilvn m
Zte ’ xrZm
atvn), he recommends banishment (apodZmZtik aB ’ vn) poie&isuai t aB parast
aseiB ayt & as an alleviation; and doubtless he had ostracism particularly in mind.72 Except in one passage (V 3, 1302b 17–19), where monarx
ia ’ dynaste
ia is mentioned as a usual consequence of (i.e. tyranny) Z one man or a group in the State becoming too powerful, Aristotle’s 72

’ As believed by LSJ9 , s.v. apodZmZtik
oB.

196

Athenian Democratic Origins

essential idea about ostracism in the Politics is that it is directed not so much against the possibility of a man’s becoming tyrant as ‘ against yperox Z dyn
amevB,73 which in this context means not merely an abundance but a dangerous excess, a superabundance, of political strength and influence, giving the individual or group concerned a position too powerful for the safety of the constitution.74 Elsewhere in the Politics75 Aristotle remarks, significantly, ’ yperoxa& ‘ ’ that those en iB eytyxZm
atvn [‘with a superabundance of advantages’]—strength, wealth, friends and the like—neither care to obey nor understand how to do so. Other ancient sources referring to ostracism explicitly strike the same note of excess ‘ (yperox
Z).76 Ps.-Andocides IV 35 (although in my opinion a late and very unreliable rhetorical exercise)77 comes back to the vital point made by Aristotle, in representing the law of ostracism as ’ ontvn kai tvn being directed against toyB kre
ittoyB tvn & arx
& n
omvn [‘those who are above the magistrates and the laws’], an expression which puts the central idea very well. Finally, the only surviving writer I can discover earlier than Aristotle who has left us an expression of opinion about the purpose of ostracism, namely Thucydides, evidently had just the same conception, for in saying ’ vmatoB that Hyperbolus was ostracised oy’ di a dyn
amevB kai aji
w
obon [‘not through fear of his power and authority’]78 he implies that the normal reason for ostracising a man was indeed fear of his ’ ivma—what the Romans called his potentia and d
ynamiB and aj
auctoritas. Now it seems to me that although we must pay respectful attention to everything Aristotle says about the nature and motives of ‘ ‘ The words yperox
Z, yper
exein, occur again and again in these passages: Pol. III 13, 1284a 20–1, 36–7; 1284b 15–17; V 3, 1302b 15–18, 19–21; 1308b 16–19. 74 This is clearest in Pol. V 3, 1302b 15–18. 75 IV 11, 1295b 13–16. 76 e.g. Diod. XI 55.3, with 2 (cf. 54.5); XIX 1.1–2, with 3; Plut., Them. 22.4; ’ E.M. 349. 15–23, s.v. ejostrakism
oB. 77 Here I entirely agree with A. R. Burn, ‘A Biographical Source on Phaiax and Alkibiades ([Andok.] IV and Plutarch’s Alkibiades)’, in CQ 48 ¼ n.s. 4 (1954) 138– 42, against A. E. Raubitschek, ‘The Case against Alcibiades (Andocides IV)’, in TAPA 79 (1948) 191–210; cf. Hignett, HAC 395–6. 78 Thuc. VIII 73.3 (cf. Plato Com., fr. 203, cited in n. 85 below). Kagan, op. cit. (in n. 21 above) 393, goes rather far in saying that Thuc. attributed the ‘establishment’ of ostracism to ‘fear and insecurity’. I fancy this may be an unconscious echo of Grote’s statement (HG III 379, chap. xxxi) that ostracism was ‘a product altogether of fear and insecurity’. 73

The Law of Ostracism: Its Date and Purpose

197

ostracism in general, it is unlikely in the extreme that he, or any other writer of the fourth century or later, had any reliable information about the original purpose of the Athenian law of ostracism. In the Ath. Pol. he is making an inference from the way in which ostracism was first used—here I agree with Jacoby’s analysis.79 If, as Dover and Sumner have suggested,80 Aristotle’s statement about the origin of ostracism closely follows Androtion’s, it would be even easier to understand why his explanation is narrowed down to one part of his own general theory, for there is no reason to credit Androtion with any great penetration of mind. We must therefore look particularly at the circumstantial evidence: the known historical facts about individual ostracisms. Even Aristotle, as we have seen, was prepared to approve the use of ostracism as a last resort. I suggest that we must try to see if we can discover some occasions on which ostracism performed a useful function at Athens—because if there were such occasions, and they have common features and betray a consistent purpose, we shall have a starting-point from which we may be able, in the light of Aristotle’s more general analysis, to draw inferences about the motives which led to the employment of ostracism in particular cases, and thus even about the original purpose of the law. It has been said that ostracism, ‘as it was worked after 486 B.C., . . . was, in fact, as injurious to the interests of the state as it was unjust to the individual. To the individual it meant the loss of all that was best worth having during the best years of his life; to the state it meant a fatal impediment to the proper working of the party system. A party unfairly deprived of its leader at some great crisis—and in the Greek democracies the leader counted for much more than he does in our modern popular governments—is not unlikely to have recourse to unconstitutional methods. The answer to the ostracism of Cimon in 461 B.C. was the assassination of Ephialtes.’81 The point made in the last sentence can be disposed of first. The assassination of Ephialtes was not a typical but an altogether exceptional event, almost unparalleled (as far as we know) from the time of Harmodius and Aristogeiton until after the fall of the democracy in 322, except during the oligarchic 79

Jac. i 123. Cf. i 121: ‘Aristotle drew the almost certainly wrong conclusion that ostracism was directed against the menace of tyranny.’ 80 81 See p. 186 and n. 25 above. Walker, in CAH IV 153.

198

Athenian Democratic Origins

revolutions of 411 and 404–3. And, although of course it is very true that ‘in the Greek democracies the leader counted for much more than he does in our modern popular governments’, there was nothing resembling a ‘party system’ in sixth/fifth-century Athens or any other Greek state: above all, the elements of organisation (a list of members, with a chairman, secretary and treasurer) were ‘ entirely lacking, except perhaps inside individual etaire& iai [‘political clubs’], and there was no common programme corresponding to those which unite modern political parties, with each individual member of the party accepting the whole programme. At Athens there were at the most only groups of a more or less fluid nature, the members of which might well be divided differently on different issues, even if they sometimes showed a tendency to share the same opinions in different fields. The ostracism of a prominent politician who had been advocating a particular policy might not convince his followers that his opinion had been wrong, but, unless those followers were very numerous, it would be likely to deter any other man from taking his leading place as an advocate of the same policy, for fear of then sharing his fate. *** There are only nine ostracisms which are certain: first, Hipparchus son of Charmus (487), Megacles son of Hippocrates (486) and a third ‘friend of the tyrants’ (485); Xanthippus the father of Pericles (484), and Aristeides (482 or perhaps 483).82 We then have Themistocles, ostracised in the late 470s;83 Cimon (461), and Thucydides son of Melesias (443). Finally, in 417 or a year or two later,84 Hyperbolus was ostracised, apparently as the result of a bargain 82

The principal authority for all this is of course Arist., Ath. Pol. 22.4–8. Contrast R. J. Lenardon, ‘The Chronology of Themistokles’ Ostracism and Exile’, in Historia 8 (1959) 23–48, with M. E. White, ‘Some Agiad Dates: Pausanias and his Sons’, in JHS 84 (1964) 140–52. 84 The older view, that Hyperbolus, who was assassinated in the spring of 411 (Thuc. VIII 73.3), was ostracised in the spring of 417, is based on the statement of Theopompus (FGrH II B 115 F 96, ap. Schol. Ar. Vesp. 1007) that he was ostracised for six years ( e‘j e
’ tZ). But Theopompus is a very unreliable historian. On the basis of what he himself calls a ‘highly conjectural’ restoration of IG I2 95 [¼ I3 85], A. G. Woodhead, in Hesp. 18 (1949) 78–83, has suggested that Hyperbolus cannot have been ostracised earlier than the spring of 416, i.e. that it must have happened in 416 or 415. Raubitschek’s theory that the ostracism was in the spring of 415 depends entirely upon what seems to me an impossible interpretation of Ps.-Andoc. IV—see n. 77 above. [See now P. J. Rhodes, ‘The Ostracism of Hyperbolus’, in Ritual, Finance, Politics (1994), 85–98, at 91 n. 36; he favours 415.] 83

The Law of Ostracism: Its Date and Purpose

199

struck between the two obvious ‘candidates for ostracism’, Alcibiades and Nicias, each of whom urged his followers to vote not against his principal rival but against Hyperbolus.85 The Athenians learnt their lesson from this and, as far as we know, never held another ostracophoria. Apart from the nine ostracisms I have mentioned there are none which are even probable, and I would not myself accept any—but this raises questions I cannot deal with here.86 We must not, of course, neglect the possibility that there may have been several ostracophoriai at which fewer than the necessary minimum of 6,000 votes were cast.87 *** We know nothing whatever about the motives of the Athenians in voting against the five men ostracised in the 480s. Even if Aristotle was right (as I do not think he was) about the original purpose of the law of ostracism, and Cleisthenes intended it primarily as a weapon to be used if necessary against a potential tyrant, in 85 Plut., Nic. 11; Alc. 13.4–9; Arist. 7.3–4; and esp. the quotation from Plato Com. (fr. 203 K/A) in Plut., Nic. 11.6–7 and Alc. 13.9, ending oy’ g ar toio
ytvn ’
‘ euZ [‘it was not for the sake of such people that ostraka were o y
‘ nek’ ostrax’ Zyr
invented’]. 86 The ostracism of Alcibiades, the grandfather of the great Alcibiades, is just possible. Even this rests on very poor authority: Ps.-Lys XIV 39, which declares, absurdly, that the elder Alcibiades and Megacles were both ostracised twice, and Ps.-Andoc. IV 34, which is almost worthless (cf. n. 77 above. Sec. 34 of the same speech also alleges that Callias son of Didymos was ostracised). E. Vanderpool, in Hesp. 21 (1952) 1–8, thinks Alcibiades was ostracised in 460. It seems to me inconceivable that Damon the ‘intellectual’ was ostracised, despite Arist., Ath. Pol. 27.4; Plut., Nic. 6.1; Arist. 1.7: see Carcopino, OA2 125–42 (although I would not accept all his arguments). It is grotesque to say, as even Glotz and Cohen do (HG II 186 n. 77), that ‘l’ostracisme de Damon est un fait acquis . . . puisqu’on a trouve´ un ostracon marque´ de son nom’: IG I2 912. To this Carcopino (OA2 126) has made the right reply: ‘Pas plus qu’une hirondelle ne fait le printemps, un ostrakon ne saurait suffire a` de´terminer un ostracisme.’ Nor do I think that Raubitschek, in Hesp. 24 (1955) 286–9, has made out a very good case for the ostracism of Menon son of Menecleides, c. 457. Even in the early fourth century ludicrously false statements could be made in public by well-educated Athenian politicians about fifth-century history, sometimes involving ostracisms which certainly never took place—see Andoc. III, esp. 3: kai Milti
adZn t on K
imvnoB ’ ’
’ Xerron
’ o to &yto, pr
vstrakism
enon kai onta en Zsfiv katadej
ameua di’ ayt ojenon ’
omenon onta Lakedaimon
ivn, o‘
pvB p
emcaimen ei’ B Lakeda
imona prokZrykeys
peri spond vn & [‘and we recalled Miltiades son of Kimon, who had been ostracised and was then in the Chersonese, for this very reason, so that since he was a proxenos of the Spartans we might send him to Sparta to open negotiations for a truce’]! [But see Afterword, p. 229 below.] 87 Cf. p. 210 & n. 115 below.

200

Athenian Democratic Origins

particular Hipparchus, it does not follow that the actual reason for the ostracism of Hipparchus was fear that he might become tyrant, for the whole situation had changed very greatly in the intervening twenty years. There has been a great deal of speculation in recent times about the nature of ‘party politics’ at Athens from Cleisthenes to the invasion of Xerxes,88 but most of it is worthless, for we have very little evidence indeed, not nearly enough to enable us to draw even the general outlines of a satisfactory picture. It is a sobering reflection that the two men against whom, after Themistocles, we have the largest number of preserved ostraca,89 namely Callixenus and Hippocrates son of Alcmeonides, are virtually unknown to us from any other source, as is Menon, the next on the list, who however belongs to a rather later period.90 All I would insist upon here is that we must not claim to know why the Athenians ostracised Hipparchus, Megacles or the third ‘friend of the tyrants’, or Xanthippus, or Aristeides. In particular it may be ‘difficult to doubt’ that the ostracism of Aristeides was directly connected with his opposition to the naval building programme of Themistocles (if indeed he was opposed to it!); but such conjectures are harmless only if we do not try to go further and build more conjectures upon them. The ostracisms of the 480s, then, can give us little help when we are looking for evidence about the motives of particular ostracisms. About the background of the last four ostracisms we have more information and can draw reasonably firm conclusions. The ostracism of Hyperbolus presents special features: I shall deal with it later. That leaves us with three occasions, on each of which, I shall maintain, ostracism may be said to have worked thoroughly well, from the standpoint of the fifth-century Athenians, who had different views from ourselves on the subject of the relation between 88 J. A. R. Munro, in CAH IV 230–2, 249–50; E. M. Walker, ibid. 154–6, 167–72, 252–3, 264–7; V 34–9, 46–9, 59–67, 68–72 (and cf. 81–2, 85–7); M. F. McGregor, ‘The Pro-Persian Party at Athens, 510–480 B.C.’, in Athenian Stud. Pres. to W.S. Ferguson (HSCP Suppl. I, 1940), 71–95; C. A. Robinson, ‘The Struggle for Power at Athens in the Early Fifth Century’, in AJP 60 (1939) 232–7; ‘Athenian Politics, 510–486 B.C.’, in AJP 66 (1945) 243–54; A. W. Gomme, ‘Athenian Notes: I. Athenian Politics, 510–483 B.C.’, in AJP 65 (1944) 321–31, repr. in Gomme’s MEGHL 19–28; D. Kagan, Hesperia 30 (1961) 393–401. 89 For figures see H. A. Thompson, in Hesperia 17 (1948) 193–4; E. Vanderpool, in Hesp. Suppl. 8 (1949) 408–12; A. E. Raubitschek, in Actes du IIe congr. internat. d’e´pigr gr. et lat. (Paris, 1952) 59–74; Hands, JHS 79 (1959) 76–9. [But see Afterword, p. 229 below.] 90 See Raubitschek, Hesperia 24 (1955).

The Law of Ostracism: Its Date and Purpose

201

individual and State.91 By each of these three ostracisms—those of Themistocles, Cimon and Thucydides son of Melesias—a potential leader of faction, whose presence threatened the unity of the State in the field of foreign policy, if not home policy as well, was removed from the city for long enough to destroy the resistance he and his followers had been offering to the group which was then dominant, led in the first case by Cimon, in the second by Ephialtes (and possibly Pericles), and in the third by Pericles. With the ostracism of Themistocles, Cimon was able to preserve Athenian friendship with Sparta, which Themistocles had been threatening,92 and bring the whole energy of Athens to bear upon the strengthening of the Delian League and the prosecution of the struggle against Persia, the culmination of which was the victory on the Eurymedon, in 469 or soon afterwards.93 With the ostracism of Cimon, this policy was reversed: Athens broke with Sparta and there followed alliances with Megara and Argos, and the First Peloponnesian war. We hear of one single attempt at treachery in 457 (or 458),94 but otherwise there was evidently a very large measure of unity behind the new set of radical democratic leaders. With the ostracism of Thucydides son of Melesias, open opposition to the imperial policy of Pericles seems to have ceased entirely;95 and even the extreme oligarchs in 411 had no desire to bring the empire to an end until they discovered that it would be impossible for them to maintain both the empire and their personal domination.96 According to our individual preferences we may believe that one or more of these ostracisms represented an error of policy on the part of the Athenians. (Many people might think the ostracism of Cimon a mistake. I myself believe the mistake was the ostracism of Themistocles.) But at any rate, in each case a situation which might easily have led to stasis or to treachery and the betrayal of the city to a foreign foe or at least to a deep division 91 That tiresome abstraction, ‘the individual’, which has bedevilled so much of modern political thought, scarcely appears in Greek political writings, and indeed the contrast between ‘the individual’ and ‘the State’ can hardly be directly translated into Greek except in some such form as t o ’
idion=t o dZm
osion. 92 [See OPW 174–6, 378–9.] 93 For 469, see ATL III 160. In favour of a later date, see M. E. White, op. cit. (in n. 83 above). 94 Thuc. I 107.4. 95 See Plut., Per. 16.3 etc., with 11–12. 96 See Thuc. VIII 90–92.1 (esp. 91.3); cf. 63.4; 70.2–71, with 86.3.

202

Athenian Democratic Origins

which would seriously have weakened Athens was cured, at the expense of ten years out of one man’s political career. I have given three certain examples of ostracisms preserving the unity of the State when it was seriously threatened. The ostracisms of the 480s may also have done precisely that. I have already argued that since we know virtually nothing about the internal situation at Athens in the 480s (or the 490s, for that matter) we can draw no firm conclusions about the reasons behind any of the five ostracisms of that decade. However, no one will dispute that by far the most important single issue confronting the Athenians in this period was the menace of another Persian invasion, more powerful and dangerous than that which the Athenians had beaten off in 490 at Marathon. Almost everyone, I think, would agree that all or at any rate most of the men ostracised in the 480s are likely to have been politicians who were in one way or another opposed to the dominant faction (led, at any rate in the late 480s, by Themistocles), which believed in resisting Persia to the end, without any form of ‘appeasement’, and no doubt had strong ideas on exactly how this could best be done, by naval as well as military action. Not that these particular victims of ostracisms need by any means necessarily have been ‘pro-Persian’, in the sense of actually desiring to subject Athens to Persian rule, either for personal or political motives (like Hippias) or perhaps in the belief that without Persian overlordship Athens would be likely to fall under Spartan domination. Indeed, we have almost conclusive proof that, with the possible exception of Hipparchus,97 none of them could have been so much as seriously suspected of being actually ‘pro-Persian’ (or, for that matter, hostile to the democracy): namely, the fact that when the Persian invasion was imminent the ostracised were recalled.98 At this grave crisis the Athenians would certainly not have recalled men who had been suspected of treacherous dealings with Persia (or of planning to overthrow the 97 Perhaps Hipparchus had already been tried and convicted in absentia, after his ostracism but before 480, on a charge of treason (Lyc., c. Leocr. 117–18). Alternatively, he was not suspected of being ‘pro-Persian’ and was recalled with the others, to be convicted later—perhaps in consequence of a failure to return. 98 Arist., Ath. Pol. 22.8 and other sources, including the famous ‘Themistocles Decree’, M/L 23 ¼ Fornara 55. Among many discussions of the recall of the ostracised in the decree, I may refer to those by D. M. Lewis, in CQ 55 ¼ n.s. 11 (1961) 61 ff., at pp. 65–6; and A. R. Burn, Persia and the Greeks (1962), 351–2. [But against the authenticity of the decree see O. Murray, Early Greece2 (1993), 295–9.]

The Law of Ostracism: Its Date and Purpose

203

democracy), or who had even advocated submission to Persia. It would surely have been quite sufficient to procure their ostracism if some of the five men in question had been willing to ‘appease’ Persia (as Hipparchus son of Charmus may well have been), or at least maintained that resistance must be effected through a hoplite army, as so successfully at Marathon, and not by a large fleet, which may have seemed to some Athenians (especially the more wealthy ones) a foolish extravagance, impossible to maintain in commission without burdensome taxation. The precise form which resistance to Persia ought to take must have been a subject of bitter debate at Athens in the 480s. We are left with only one ostracism, that of Hyperbolus, which is apparently anomalous, in that the motives we have seen at work in the other cases did not apply. In fact, however, if the ostracophoria on this occasion had not been ‘managed’ by Alcibiades and Nicias, and had achieved the end I believe it had been designed to secure, it is very likely that the great Sicilian expedition either might never have been launched, or at least would not have been such an unmitigated disaster. For the ostracophoria of c. 417 can only have been intended, by those who voted for it, to get rid of either Nicias or Alcibiades, the two leading politicians of the day, men who advocated diametrically opposed policies in relation to the war against Sparta, which since 418 had been merely smouldering but might flare up again at any time. If Alcibiades had been ostracised, the most important advocate of the Sicilian expedition of 415 would have been removed from the scene, and those like Nicias who opposed that expedition might have been able to prevent its being despatched, or at least to limit its objectives and keep it to a moderate size; whereas if Nicias had been ostracised, the course followed by the generals who led the expedition could hardly have been so incompetent and ruinous to Athens. I should like to draw attention to an interesting feature of the ostracophoria of which Hyperbolus was the victim. If the story of collusion between Nicias and Alcibiades is true,99 as it surely must be, then that ostracophoria must have been voted against the will of the two most prominent and influential Athenian politicians of their day. *** 99

See n. 85 above.

204

Athenian Democratic Origins

We can now return to Aristotle’s explanation of ostracism and try to formulate a theory of the purpose of that institution which will take account both of what Aristotle says and of the circumstantial evidence we have now discovered from the historical occasions on which ostracism was actually used at Athens. ’ ’ Since Aristotle says that ostracism exists eniaxo &y . . . oiffl on en ’
Argei kai ’Au
ZnZsin [‘in certain places . . . such as Argos and Athens’],100 and we also hear of it, without any details, at Megara oB at Syracuse,102 and Miletus,101 and under the name of petalism
and since some of the other states may conceivably have used ostracism in Aristotle’s day, whereas Athens had ceased to employ it for nearly a hundred years, Aristotle’s conception of ostracism may have been formed at least partly on the basis of its functioning in cities other than Athens, where it may have taken rather different forms from the one kind we happen to know about. But surely it is likely that Aristotle had Athens chiefly in mind. At any rate, in what follows I shall be speaking only of ostracism in its Athenian form. As we have seen, what makes a man a ‘candidate for ostracism’, in Aristotle’s view (as in that of Thucydides, Plutarch and others), ‘ is that he is one who can be said to yper
exein dyn
amei,103 an expression we have seen reason to interpret as the possession of political power or influence so great as to endanger the safety of the constitution—and perhaps lead to a tyranny, but this is only a special case. I am going to suggest that Aristotle’s account will not do as it stands, but that we can accept it if we add to it one simple proviso. That Aristotle’s analysis of ostracism will not fit the historical facts so far as Athens is concerned should be immediately clear, for the following reasons: (a) First, there is the argument drawn from examination of the circumstances of each known ostracism. Probably not a single one of the men known to have been ostracised possessed at the time of his ostracism a superabundance of political power (d
ynamiB). Aristeides may have been such a man at the time of his famous assessment of the cities of the newly founded Delian League, but 100 101 102 103

Pol. V 3, 1302b 18–19; Schol. Ar. Eq. 855. Schol., as cited in the preceding note. Diod. XI 86.5–87.2; and see below. See pp. 195–6 above.

The Law of Ostracism: Its Date and Purpose

205

not in 483–2. Themistocles may certainly have been dangerously powerful about the time of Aristeides’ ostracism and until just after 480, but before he himself was ostracised in the late 470s he had obviously been discredited and undoubtedly had far less influence. In the years following 477 Cimon rapidly achieved a leading position in the State: by the time of his victory at the Eurymedon he was the most powerful man at Athens, and he remained so until the repudiation of his philo-Laconian policy in 462.104 When he was ostracised in the spring of 461 he had already ‘fallen from power’. In 443 it was not Thucydides son of Melesias but Pericles who was the man of power. (The ostracism of Hyperbolus has already been discussed.) Aristotle’s theory does not fit a single one of these cases. We can say nothing with confidence about the first four ostracisms, but is it not likely that Hipparchus and the others were less rather than more influential after Marathon than they had been in the last years of the sixth century? Those who believe that relatively minor figures like Alcibiades the Elder, Damon and Menon were ostracised will have even greater difficulty in accounting for Aristotle’s theory of the purpose of ostracism. The memory of how important and influential the victims of ostracism had once been may have influenced Aristotle unduly, and led him to assert ‘ yperox Z dyn
amevB as the cause of their ostracism. (b) Secondly, there is an argument of a general character: the greater the personal influence of a man with the demos, the less was he likely to be ostracised! One cannot conceive of Cleisthenes being ostracised in 507, Themistocles in 482–79, Cimon in 475–62 or Pericles in at any rate 442–32. If those who wished to ostracise such a man were able to procure the holding of an ostracophoria, the most likely victim would be not the man himself but his principal political opponent. To imagine that the Athenians forged a double-edged weapon of this sort, so absurdly unsuitable for its alleged purpose of getting rid of an over-mighty politician, is to credit them with far too low a degree of intelligence. *** The proviso I would add to Aristotle’s theory of ostracism is that the man who has the excess of political power must, in order to become a suitable ‘candidate for ostracism’, be one who is not just 104

See OPW 172–3, 179.

206

Athenian Democratic Origins

more politically influential than other individuals, but likely to make his will prevail against the majority of the citizens. The man who is ostracised will be the one who is leading the opposition to the will of the majority on some vital matter. As a pupil of mine once put it, in terms of modern British politics, ‘The man who gets ostracised is the Leader of the Opposition’. Aristotle assumes that the man possessing excessive d
ynamiB, and his faction, are likely to try to seize power, turning the constitution into a tyranny or an oligarchy; but, as we have seen,105 this is not central in Aristotle’s thought about ostracism and appears in only one passage in the Politics, though it occupies an important place in his statement of Cleisthenes’ motives in introducing ostracism, in the Ath. Pol. Since there is no reason to think Aristotle had any good sources of information on this subject, I suggest that we are justified in attributing to Cleisthenes himself the same ideas as those which can be clearly seen at work in the historical ostracisms about which we are best informed. As I see it, then, ostracism was intended to be used at a time when there was a deep division within the State on a matter of great importance, a division which if it continued might gravely weaken the city and might even lead to stasis or treachery—an even likelier eventuality in the late sixth or early fifth century than later, when the democratic process had taken firmer root. At such a time ostracism could usefully be employed. Its effect was to deprive for a time of its leader that faction which had the less popular support—a thoroughly democratic procedure! And severe measures were not taken against the victim of ostracism: his citizenship and his property were not taken from him (not even the income from his property), and after the expiration of ten years he could return, and even assume high office again, as Xanthippus, Aristeides and Cimon did. Furthermore, his children seem not to have been affected in any way, even during his absence. All this contrasts strongly with the position of the man sentenced to exile (wyg
Z), a punishment very often inflicted by Greek states on those leaders of the losing side in a stasis who survived: a sentence of exile normally involved perpetual loss of civic rights and confiscation of property, and it would frequently extend to the children of the banished man.106 105 106

See pp. 195–6 above. See e.g. Schol. Ar. Vesp. 947 (Idomeneus, FGrH III B 338 F 1).

The Law of Ostracism: Its Date and Purpose

207

Ostracism, on the other hand, was confined to securing the one essential means of reducing tension and aborting stasis: the departure for a time of the ‘Leader of the Opposition’ (if I may for once use that very anachronistic term). With luck, one ostracism might settle the issue for the time being, as those of Aristeides, Themistocles, Cimon and Thucydides evidently did. But if the division among the citizens still remained deep, and too nearly equal, the place of the departed leader might soon be taken by another prominent man and another recourse to ostracism might then follow. From the point of view of a fifth-century Athenian democrat there could be only two major potential defects in the system of ostracism. First, unless the Athenians exercised the greatest restraint in voting that an ostracophoria be held, it might be abused, for what we should call ‘party-political’ purposes: that is to say, it might be employed when there was no deep division within the city on any really essential matter, but rather a personal rivalry between two or more prominent men for political leadership. But, as we shall see presently, there was a rule that if fewer than 6,000 votes were cast the ostracophoria had no effect, and this would provide some insurance against the misuse of ostracism in this way—I imagine that this was precisely what it was intended to do. And secondly, if the two leading figures of the community in a fairly well-balanced situation took fright and struck a bargain, and successfully urged their followers to vote against some third person, ostracism would be prevented from securing its proper end. However, when this happened once, in c. 417, the Athenians learnt their lesson and apparently never again held another ostracophoria. I suggest that if we consider the purpose of the law of ostracism, as discussed above, in relation to the known date of its introduction, that purpose becomes even more understandable, for Cleisthenes himself had just been through the very situation which I have suggested the law of ostracism was designed to prevent: he had been the leader of one side (the popular side) in a stasis which might have been avoided altogether had ostracism been available already.107 Cleisthenes evidently had the majority 107

This essential point, which too many recent writers have missed, was seen by Wilamowitz, AuA II 87 n. 28, and esp. F. Schachermeyr, ‘Zur Chronologie der kleisthenischen Reformen’, in Klio 25 (1932) 334–47, at p. 347.

208

Athenian Democratic Origins

of Athenians on his side in 508/7,108 probably the overwhelming majority apart from a handful of aristocrats. Isagoras, it is true, was elected archon for 508/7; but there is no reason to think that elections to the archonship down to 487, and thereafter to the strategia, were ordinarily made according to ‘party-political’ alignments. We must take care to allow for the very personal nature of claims to high office at Athens and all other ancient states. As in the case of the consular and other elections at Rome, men must have been elected to magistracies in sixth-century Athens (and fifth/ fourth-century Athens as well) mainly because of a combination of wealth, birth, natural ability as leaders, and that training in the art of politics which could best be obtained by membership of one of the great ‘political families’, not—save in very exceptional circumstances—because of any particular policy the candidate undertook to follow or any measures he promised to introduce. A candidate for election would never (or virtually never) canvass on the basis of a programme of intended measures. The election of a given man, therefore, need not necessarily mean that he and his faction are politically dominant. Too many groundless inferences about the course of Athenian politics in the late fifth century have been founded upon the notion that the elections to the principal magistracy, then the strategia, can be taken as a reliable indication of the strength of parties or groups. (The same error has produced much false theorising about Roman politics in the last two centuries B.C.) A particular ‘party-political’ complexion might be relevant at political crises, but as a rule it would probably weigh much less than the other characteristics mentioned above. Isagoras was one of the most prominent aristocrats of his day—perhaps, as the sources suggest,109 the most prominent, at any rate among those who were not compromised in the eyes of their fellows by having held office under the Peisistratids, as men like Cleisthenes and Miltiades perhaps may have been.110 Isagoras was therefore virtually certain to be elected archon when he became a candidate, even 108

See the first sentence of Hdts V 70.1. See esp. Hdts V 66; 72.1,2; 74.1. 110 See M/L 6 ¼ Fornara 23. 5–6. It is true that Arist., Ath. Pol. 20.1 refers to ’ t vn & tyr
annvn [‘friend of the tyrants’]; but I agree with Sandys (ad him as w
iloB vn loc.) that ‘the opponent of Cleisthenes, the foe of the tyrants, must necessarily be their friend’. I would add that anyone who is not prepared to accept this argument should look at Arist., Anal. Post. I 34, 89b 10–14, where Arist., giving characteristic ’ inoia), points to the man who examples of the possession of quick-wittedness (agx
109

The Law of Ostracism: Its Date and Purpose

209

if he were known to be a man whose political views were very different from those of the mass of Athenians. Such a man, however, might at the very same time be an obvious ‘candidate for ostracism’! The fact that Isagoras was elected archon in 508, therefore, is no evidence that he would not have been a likely victim of ostracism, if that institution had existed in 509/8 and the Athenians had realised that he was the principal opponent of the reforms which so many of them favoured, and was prepared to create a stasis and even call in the Spartans rather than submit to the introduction of Cleisthenes’ measures. The law of ostracism would have been a very natural reaction on the part of Cleisthenes and the members of his circle to the crisis which the Athenians had just surmounted. And Cleisthenes, of all Athenians from the time of Peisistratus to the heyday of Cimon, had least to fear from his own law, since as the author (or at any rate the sponsor)111 of the new democratic constitution he could obviously expect to enjoy great prestige for some time; and moreover he seems to have been an elderly man,112 with no long political future ahead of him. Ostracism, then, was designed to cope with just the sort of situation to which Cleisthenes himself had been exposed: a deep conflict within the State, likely to lead to stasis or the calling in of a foreign power; but such a situation probably did not arise again until after Marathon. We can now see that ostracism was not nearly such a foolish institution as it is generally supposed to be nowadays. Provided only that the Athenians could refrain from holding an ostracophoria except in the circumstances for which it was originally intended, as I have described them, ostracism might do no real harm, and sometimes much good. Indeed, its effects in the right circumstances would have been regarded by all Greek democrats as in principle beneficial, however much they might disapprove of the result of a particular ostracophoria. Even if one thinks the Athenians made a wrong choice among the various ‘candidates for ostracism’ on one or more occasions, and even if one admits that the ostracism of Hyperbolus—a unique case—was bad in principle, it is hard not to admit (and several critics of ostracism, sees two people in conversation and divines ‘that the friendship of these people sprang from a common enmity’! 111

Cf. p. 135 above.

112

See p. 190 & n. 50 above.

210

Athenian Democratic Origins

such as Hignett,113 have virtually admitted) that in the 480s it probably helped materially to strengthen Athens for her great struggle with Persia. I have perhaps not sufficiently stressed two features of ostracism of which we can now see the full significance: the preliminary vote ’ in the sixth prytany, by show of hands, epixeiroton
ia (and therefore probably without debate),114 and the quorum of 6,000 without which the ostracophoria could have no effect.115 At the prelimin’ ary epixeiroton
ia the Athenians of Cleisthenes’ day and later would know very well on what principles they should vote to hold an ostracophoria; and if there was no debate it would be difficult for unscrupulous politicians to whip up feeling in favour of an unjustified ostracism, except by a great deal of canvassing beforehand—an activity to which Athenian politics did not lend itself very easily, since most matters were decided by debate in the Assembly, and it was on these debates and not (as far as we know) on activity in advance of them that the more important politicians would normally rely for influencing their potential supporters.116 The quorum of 6,000 was a figure probably far above the normal attendance at the Assembly in the late sixth and early fifth centuries.117 The reason for fixing a quorum as high as this was surely that ostracism was intended to be used only when there was deep and divided feeling on some important political issue. If an ostracophoria was voted when feelings were not really running very high, responsible citizens could show their disapproval by abstaining: if a substantial number of Athenians did that, the quorum would not be reached and there would be no result. (If a deep 113

HAC 186. Arist., Ath. Pol. 43.5; Philoch., FGrH III B 328 F 30. 115 Against the impossible view, based on an obviously defective lexicographical condensation of Philoch. (F 30), that the figure of 6,000 is that of the number of votes against one individual, it will be sufficient to refer to Jac. i 315–16; Hignett, HAC 165–6. The decisive arguments are the double counting of votes in Plut., ’
Arist. 7.5–6, and the analogy with votes of indemnity (adeia). 116 I do not mean to imply that there was not endless discussion of political issues at Athens outside the Assembly itself—of course there was. But in the absence of party organisations in anything like the modern style (see p. 198 above) it would be much more difficult than it is today for a group of politicians not merely to decide on a common policy but also to influence their potential supporters to vote for it, except through their speeches in the Council and Assembly. 117 Cf. Thuc. VIII 72.1. Admittedly this refers specifically to war conditions, but by the late fifth century the citizen population of Athens must have been very much larger than in the late sixth. 114

The Law of Ostracism: Its Date and Purpose

211

division in the city really existed, of course, it might be dangerous for the supporters of an obvious ‘candidate for ostracism’ merely to absent themselves, for it would be safer in such cases to vote against the leader of the opposite faction.) If one looks at the two safeguards I have just been discussing with the right interpretation of the purpose of ostracism in mind, one can see that they were ingenious and likely to be quite effective. Again, the good sense of the late sixth-century Athenians shows itself conspicuously. For my own part, I am inclined to doubt whether the repeated use of ostracism in successive years, after the manner of the 480s, can have been contemplated originally. I cannot help wondering whether five ostracisms in five or six years can really have been justified according to the original plan, in the sense that unless the five men were silenced the State would continue to be deeply divided. But, as I have already argued, we really know almost nothing about Athenian politics in the first two decades of the fifth century. It is quite possible that the menace of Persian conquest which began to threaten Athens seriously soon after 499, and even more insistently after 490, distorted the situation to an unparalleled and unforeseen degree. Perhaps on the one hand those who followed the policy we associate particularly with Themistocles felt that any tactics whatsoever were justified in ridding the city of those prominent men they considered too ‘soft’ in their attitude to Persia or too misguided in the means they would employ for resisting Persia; and perhaps on the other hand there were men prepared to go to any lengths rather than allow the Athenians to be ruined utterly, as the Eretrians had been in 490, by following a reckless and untried policy. Whether there was a widespread fear of a recurrence of tyranny in the last decade of the sixth century or the first two of the fifth is a question upon which very different opinions have been expressed. I do not see how we can possibly settle this question with any confidence, but my own subjective impression is that there was no serious danger of tyranny after 510, from a Peisistratid faction or from any similar group.118 As Hignett puts it, ‘There was no sentimental glamour in the memory of the exiled tyrant, and there was now no political or economic discontent which he 118 See the sensible remarks of Gomme, in AJP 65 (1944), at p. 328, repr. in Gomme’s MEGHL 25–26.

212

Athenian Democratic Origins

could exploit in his own interest.’119 Athens had outgrown her need for a tyrant, and only aristocratic misrule, such as Solon had tried with only partial success to overcome by legislation and Peisistratus had put an end to by force, could have created a situation in which any substantial number of Athenians might have been willing to endure a tyrant once more. ‘The people regarded tyranny only as an expedient. They used it as a battering-ram with which to demolish the citadel of the oligarchs, and when their end had been achieved they hastily abandoned the weapon which wounded their hands.’120 Tyranny except as a product of acute class strife was a rare phenomenon in Greece. And in any event, ostracism was an extremely unsuitable expedient for checking a would-be tyrant. Greek aristocrats never willingly endured tyranny: the tyrant usually climbed to power by exploiting, against an oppressive or too exclusive aristocracy, the grievances of the classes below them. But if such a man had so much popular support that there was a real danger of his making himself a tyrant, and ostracism were employed, the most likely victim would be not the prospective tyrant himself but his most effective rival. Either the man was not dangerous enough to need ostracising, or he was not likely to be ostracised. The only real danger of tyranny at Athens after 510, I would say, came from the man who might seek to impose it by an armed coup or after a military defeat, with the help of an outside power, especially of course Sparta or Persia; and even in such circumstances oligarchy was a far more likely eventuality.121 I would not claim to be able to demonstrate all this to anyone who doubted my premises; but even if it could be shown (as it certainly cannot) that there were a substantial number of Athenians after 510 who desired a return to tyranny, the nature of ostracism, as I have described it, would forbid us to suppose that it was specially designed to check this particular danger. *** 119 HAC 179. I feel less in sympathy with the paragraph which follows. All that Kagan says about a supposed ‘tyrannist party’, in Hesp. 30 (1961) 397 ff., is pure speculation, and to my mind exceedingly improbable. 120 Glotz, GC 116. This general statement about tyranny is true of at least some early Greek examples, and especially so of Athens. 121 Cf. what happened in 508/7, 404/3 and 322.

The Law of Ostracism: Its Date and Purpose

213

Ostracism existed also—when and for how long we do not know— at Argos, and at Megara and Miletus; and under the name of ‘petalism’ it was introduced for a short period in the mid-fifth century at Syracuse.122 We have no details at all about the Argive, Megarian or Milesian forms of ostracism, and our only surviving account of the working of petalism at Syracuse comes from Diodorus, who is vague and superficial. According to Diodorus, petalism was introduced after several unsuccessful attempts at tyranny, by Tyndarides and others,123 and, just as ostracism at Athens was directed against to &y doko &yntoB m
alista d
ynasuai tyranne&in tvn & politvn & [‘that one of the citizens who seemed most able to establish a tyranny’], so Syracusan petalism was intended to strike at t on dynat
vtaton tvn & politvn & [‘the most powerful of the citizens’]: both sides hoped tapein
vsein t a ’ ta&iB patr
isi [‘to reduce wron
Zmata tvn & ple&iston i’ sxy
ontvn en the ambitions of the most powerful men in their countries’].124 The best citizens now refused to take part in politics at all, a host of demagogues and sycophants arose, and the city fell ei’ B synexe&iB kai meg
alaB . . . tarax
aB [‘into continuous serious disturbances’], with the result that the law was soon repealed.125 There is little to be learnt here. *** Aristotle observed, as we have seen,126 that Greek states employed ostracism stasiastikvB & [‘in a party-political manner’] instead of using it properly. What exactly does he mean? Taking the word st
asiB in its broadest sense, without any necessary implication of violence, we might feel that Aristotle would have been justified in applying it, for example, to the ostracism of Hyperbolus. I know of no other Athenian ostracisms to which Aristotle’s strictures could properly be applied, but of course there may have been such, and in other cities (Syracuse for instance) the institution may have been much abused for ‘party-political’ purposes. The real vice of ostracism was the danger of its being used improperly, against which it was hardly possible to make effective provision. Mere irresponsibility would always be a factor, but in Athens I fancy it is not likely to have been a major factor. Doubtless individual Athenians did 122 124

123 See p. 204 and nn. 100–2 above. Diod. XI 86.4–5. 125 126 Diod. XI 87.1–2. Diod. XI 87.3–5. P. 195 above.

214

Athenian Democratic Origins

cast their ostraca irresponsibly. The story in Plutarch, Arist. 7.7–8, of the man who wanted to ostracise Aristeides simply because he was tired of hearing him called ‘the Just’, though presumably apocryphal, may represent something that did happen at times. There is also the fact that the actual ostraca discovered in modern times include a few bearing the names of men who can hardly have been important enough to deserve this particular form of censure. And there is the ostracism of Hyperbolus. Irresponsible behaviour of this sort on the part of fifth-century Athenians will excite the contempt or the surprise of those only who have never taken an active part in modern politics. Political irresponsibility is widespread enough today to make us hesitate before accusing the Athenians of possessing an undue measure of that quality. It is often supposed127 that ostracism was originally invented for one purpose—usually the prevention of tyranny—and then used, from the first, or after a few years, for another: getting rid of ‘proPersian’ figures, or simply of personal rivals. There is no reason to adopt any such opinion. I have shown that all the ostracisms we know anything about can be satisfactorily explained on a single principle, which is that explained by Aristotle in the Politics, with the addition of one important qualification. (We must merely put a question-mark against the ostracisms of the 480s, about the circumstances of which we lack information.) Ostracism, as I have shown, was a highly unsuitable device for checking the rise of a would-be tyrant, and we must refuse to accept the explanation in these terms offered by Aristotle in the Ath. Pol.—perhaps generalising from a wrong interpretation of the first recorded ostracism, perhaps relying too easily on the superficial opinion of Androtion.128 In conclusion I would point out that in the late sixth and early fifth centuries (whatever we may think of their behaviour later) the Athenians showed the most remarkable political ability, behaved with extraordinary skill and good sense, survived the hostility of Sparta, Aegina and even Persia (playing undeniably the largest part in the repulse of the Persian invasions), and came to the very forefront of the Greek world. Many writers, ancient and modern, who have discussed the origin of ostracism have assigned motives 127 128

Most recently by Kagan, Hesp. 30 (1961) 396–401. Cf. pp. 182–8 above.

Archons and Strategoi

215

for it which would involve our crediting the Athenians with a very low degree of intelligence and political acumen. The theory I have proposed allows us to attribute to the inventors of ostracism a thoroughly sensible motive, such as we might expect from that brilliant generation of Athenians. Ostracism had latent dangers, it is true; but as soon as these dangers became apparent, with the ostracism of Hyperbolus, the Athenians drew back, and—oy’ g ar ‘ euZ [‘it was not for the sake of such toio
ytvn oy
‘ nek’ o
’ strax’ Zyr
people that ostraca were invented’]129—never again suffered the institution to be degraded by misuse.

(II) ARCHONS AND STRATEGOI I come now to the constitutional reform of 487, by which the ’ prokr
itvn [‘sortiarchons came to be appointed by kl
ZrvsiB ek tion from pre-selected candidates’]. I say ‘487’ in order to avoid choosing between 488/71 and 487/6.2 The latter is the date usually accepted, but there is a real doubt whether 487/6 is the year in which the reform was introduced or that in which the first archons chosen under the new system took office: Hignett3 goes too far in saying that to date the reform 488/7 is ‘in defiance of Ath. Pol. ’ yB de tfiv& yst
‘ erfiv etei ’
ep ’ i 22.5’. Aristotle’s actual words are eyu ’
’ ameysan toyB enn
’ ea a ’
rxontaB Teles
inoy arxontoB [487/6] eky
ktl. [‘straightaway in the following year, in the archonship of Telesinos (487/6) they appointed the nine archons by lot’]. We would certainly expect this to mean that the first choice of archons by lot took place in 487/6, so that the reform must have been enacted in that year, and the first year in which the archons chosen by lot actually took office will have been 486/5. However, in Ath. a t on ’Ewi
altoy u
anaton e’
gnvsan Pol. 26.2 we read, e
‘ ktfiv e’
tei met ’ zeygitvn ’ ea kai ek & prokr
inesuai toyB klZrvsom
enoyB tvn & enn

’ ontvn, kai prvtoB ’ ayt ’ vn arx
& Zrjen ej & MnZsiue
idZB [‘in the sixth year after the death of Ephialtes they resolved that preselection for those to be appointed by lot to the nine archonships should be open to the Zeugitai also, and Mnesitheides was the first 129

See n. 85 above. Beloch, GG II2 ii 139, and others. See Hignett, HAC 176 & n. 3 (giving Beloch’s date wrongly as 489/8); cf. 186 & n. 3. 3 HAC 186 n. 3. 1 2

216

Athenian Democratic Origins

of them to be archon’]. Now Ephialtes was assassinated in 462/1, and Mnesitheides, the first archon who actually held office under the new system, was archon in 457/6. According to Greek reckoning this was ‘the sixth year after the death of Ephialtes’. But what Aristotle says is that in this very year (the sixth after the death of Ephialtes) the Athenians passed the reform (e’
gnvsan). It cannot be true both that in 457/6 the reform was passed and that in the same year the first archons chosen under it took office. Either ’
v e’
tei) is a mistake for ‘the fifth year’ (p
emptfiv ‘the sixth year’ (ektfi e’
tei), or else (more probably, it seems to me) Aristotle is writing in a way that would seem inaccurate to us but would, I think, have ’
been natural to a Greek, and he means, despite egnvsan, that the reform became effective in the year 457/6. If the latter is right, then although the wording of Ath. Pol. 22.5 is different, Aristotle may well mean that the choice of archons by lot was first made for, rather than in, the archonship of Telesinus, 487/6. This reform has been misrepresented, in my opinion, by nearly all recent historians. Only Hignett has given, in one passage,4 substantially the right explanation; but even he is not consistent;5 he does not produce the really conclusive argument why the usual view is impossible, and his general picture is marred by his dating the law of ostracism to the same period. By far the most usual interpretation of the reform nowadays is that it was a ‘democratic’ measure.6 Beloch7 even speaks anachronistically of ‘eine Verfassungsreform im ultrademokratischen Sinne’ [‘a constitutional reform of hyper-democratic intent’], Busolt8 of a reform ‘im Sinne der Radikalen’ [‘in the spirit of the radicals’]. Jacoby,9 adducing arguments in favour of Beloch’s theory that the law of ostracism was introduced in 488/7, thinks it possible to reason that ‘the law about ostracism in 488/7 B.C. and that about the appointment of archons by lot in 487/6 B.C. sprang from the same democratic attitude of mind and are quite credible 4 HAC 187 n. 3, 188; contrast 173 ff. It is worth noticing that Wade-Gery in 1933, in an article reprinted in his EGH (at pp. 153–4), dropped a hint which points in the right direction. 5 See HAC 173 ff. (esp. 175–6), 230–1. 6 In addition to the writers mentioned below, see Walker, in CAH IV 156; Macan, Hdts IV–VI II 143; Hdts VII–IX II 202. 7 GG II2 ii 139. 8 GS II 887; cf. GG II2 638–9. 9 Jac., i 122–3.

Archons and Strategoi

217

for the first years of the ’eighties’; and he adds in a note,10 ‘This argument, which is perhaps the most conclusive, escaped Beloch.’ It is true, of course, that in the ‘Persian debate’ on the rival merits of democracy, oligarchy and one-man-rule which Herodotus11 incorporated in his work, sortition of magistrates is regarded as a characteristically democratic principle, to be mentioned in the same breath as accountability and the sovereignty of the Assembly;12 and that political theorists, in the fourth century anyway, tended to consider election by vote as aristocratic in principle, by lot as democratic: see, for instance, Plato, Rep. VII 557a; Aristotle, Rhet. I 8, 1365b 31–4; II 20, 1393b 4–8 (with Xen., Mem. I ii 9); Aristotle, Pol. IV 9, 1294b 7–9, 12; VI 2, 1317b 17–21; II 11, 1273a 18. But, as I shall show, to conclude that the reform of 487 was specially ‘democratic’ would be entirely fallacious—as, for quite different reasons, would the opposite conclusion, that the reform was not democratic but oligarchic. I think it can be proved that the reform was not of a ‘party-political’ nature at all: it was neither ‘democratic’ nor ‘oligarchic’, but was part (and by no means the most essential part) of a vitally necessary improvement in the efficiency of the organisation of the State. The notion that the reform was a ‘democratic’ one can easily be refuted. First, election by lot may in principle be more ‘democratic’ than election by vote, but only when the election is from the whole citizen body and not the upper classes alone, as it was in the early part of the fifth century, when only Pentakosiomedimnoi and Hippeis were eligible for the archonship. (It was not until 457/6 that even the Zeugitai were admitted.)13 When the whole citizen body, in which Zeugitai and Thetes together must have greatly outnumbered the two upper classes, elected the archons by vote, the lower classes could, if they wished, choose those members of the two top classes who were most friendly to the people (dZmotiko
i).14 If, on the other hand, the choice were made by lot, the men elected would be a much more representative sample 10 Jac., ii 117 n. 38, where he refers to Macan and Ehrenberg as having used this argument. (Cf. p. 182 & n. 12 above). 11 Hdts III 80–82. Hdts was quite sure that the debate was genuine (see III 80.1; VI 43.3); but this seems to me inconceivable. My own guess is that Herodotus found it in some written source he trusted, and that it comes from the early fifth century. 12 Hdts III 80.6. 13 Arist., Ath. Pol. 26.2. 14 For the meaning of this word, see my CAE, at pp. 21 ff., esp. 22–3.

218

Athenian Democratic Origins

of the superior classes, and therefore would probably be less ‘democratic’ in outlook. This would be only partly mitigated by the preliminary pre-selection (pr
okrisiB), and only to a very small extent if the prokritoi numbered as many as 500, the figure in the London papyrus of Ath. Pol. 22.5. Many commentators have distrusted this figure,15 and some would substitute 100, from Ath. Pol. 8.1 (describing the practice existing in Aristotle’s day); but in 22.5 Aristotle speaks of the prokritoi as being chosen by the demesmen, and if the elections at first took place by demes16 a large number of the order of 500 seems much more likely. This argument against the alleged ‘democratic’ nature of the reform of 487 can be supported from ancient authority: Isocrates, praising the ancestral democracy in a characteristically dishonest passage (VII 23), remarks that the Athenians of the time he is talking about thought election of magistrates by vote dZmotikvt
eran than the use of the lot, ‘since in election by lot chance would decide the matter and those who desired oligarchy would often obtain the magistracies; but by election of the most ’ ’ de tfiv& prokr
inein17 toyB epieikest
atoyB) the suitable men (en people would have the power to elect by vote those who were most devoted to the existing constitution’. Cf. Aristotle, Pol. IV 15, 1300a 9--b 5, a corrupt and difficult passage, where (in so far as 15 My own feeling is certainly that there could hardly have been enough eligible candidates each year from the two top property classes—whose numbers it is impossible to estimate, but who will have been even fewer than most people have supposed if I am right about the nature of the qualification of the Hippeis (see pp. 25–6, 47–8 above). However, we need not suppose that there would have to be a different 500 each year. Why should not most of the unsuccessful candidates in any year have tried again next year? [E. Badian, Antichthon 5 (1971) 17–19 calculates that there were perhaps 1500–1600 Pentakosiomedimnoi and Hippeis alive at any given time in the 480s, of whom at least half will have been excluded by unwillingness to serve and c. 230–240 by the fact of previous service (if iteration was forbidden). He supports Kenyon’s emendation ‘100’, as does Rhodes, CAP 273.] 16 Presumably with provisions for ‘proportional representation’ of demes analogous to those regulating the election of Councillors (see p. 165 above). This too seems very unlikely to me, for there would be quite a number of demes with few or no members of the two top classes. Incidentally, if Aristotle is right here, the stage referred to in Ath. Pol. 62.1, at which the sortition of archons (as well as some other ’ QZse
ifiv: ‘those in the magistrates) took place by tribes, while other sortitions (ai‘ en Theseum’) were carried out by demes, must (in Aristotle’s mind) be a rather later one, intermediate between that described in Ath. Pol. 22.5 and the situation in the late 4th century, when all sortitions were by tribes, except of councillors and wroyro
i, for which lots were still cast by demes (Ath. Pol. 62.1). 17 By which Isocrates must mean simply election: see pp. 101–3 above.

Archons and Strategoi

219

the text is intelligible) there seems to be a distinction between modes of appointment (katast
aseiB) which are dZmotika
i, with all choosing from all by election or lot or both (1300a 31–4), ’ and a situation later described as politik
on or aristokratik
on, ’ tinvn) where choice is sometimes from a limited number (ek & (1300a 34--b 5). It is absurd to claim, as so many writers have done, that the underlying intention of those who opened the archonship to the lot was to curb the power of the principal magistrates, who would be likely to be less distinguished men and enjoy less prestige if they were elected by lot. Advocating this view, Hignett18 claims, ‘The archonship at once declined in importance and ceased to attract the ablest men in the state. Sortition, like rotation, was a device to protect the sovereign demos against the acquisition by its potential rivals, the boule¯ and the magistrates, of an authority and an influence which might threaten its own sovereignty.’ This theory ignores an essential fact. Hignett himself points out elsewhere19 (thereby amply refuting his own statement just quoted) that ‘the reform had merely substituted a new and more insidious danger for the old’: the Strategos who could be re-elected year after year, as the archons could not. The Athenians could hardly have been so stupid as not to realise that a magistrate who can be constantly reelected is a far greater potential danger than one who can hold office for one year only! The arguments against conceiving the reform of 487 as a ‘democratic’ one are conclusive, and I cannot understand why they have not been obvious to recent writers. The whole question was argued at great length in the eighteenth and nineteenth centuries, although of course two vital pieces of information first revealed in 1891 on the publication of the London papyrus of the Athenaion Politeia were not then known: the dating to 487 of the law introducing sortition of archons, and the fact that Plutarch20 was wrong in making Aristeides responsible for the passing of a law that all Athenians should be eligible for the archonship. Most recent scholars have tended to ignore the pre-1891 controversy; and this is a pity, because several of the older scholars21 thoroughly 18

19 HAC 230–2, at p. 231. HAC 187. Arist. 22.1; contrast Arist., Ath. Pol. 26.2; cf. 7.4 fin. 21 Above all Grote (see below) also Scho¨mann (1854 and 1861), Emil Mu¨ller (1857), Duncker (1860), Lugebil (1868 and 1871) and Mu¨ller-Stru¨bing (1873). It 20

220

Athenian Democratic Origins

understood that the opening of the archonship to the lot would have been anything but a democratic reform so long as the archons were still chosen only from the wealthiest citizens. After making this very point, Grote22 concluded (wrongly, but with excellent judgement on the evidence then available to him) that ‘these three points—1. The opening of the post of archon to all citizens indiscriminately; 2. The choice of archons by lot; 3. The diminished range of the archon’s duties and responsibilities . . . are all connected together, and must have been simultaneous, or nearly simultaneous, in the time of introduction: the enactment of universal admissibility to office certainly not coming after the other two, and probably coming a little before them.’ Against this, Scho¨mann and Mu¨ller-Stru¨bing,23 each for different reasons, actually went the whole hog and concluded that the opening of the archonship to the lot was an anti-democratic measure. This theory scarcely needs refuting nowadays; it would be ridiculous to conceive the Athenians as voting for an oligarchic reform three years after Marathon. But what then are we to conclude? I suggest that the opening of the archonship to the lot was only part of a bigger reform having no ‘party-political’ significance, and will be sufficient to give references to the last two writers, who between them review the whole controversy: K. Lugebil, ‘Zur Gesch. der Staatsverfassung von Athen, II. Das Archontat u. die Strategie zur Zeit der Perserkriege und die hist. Bedeutung der Beamtenerloosung’, in Jahrb. f. klass. Philol., Supplementband V, Heft 4 (1871) 564–699 (a revised version of a previous publication in Russian in 1868), esp. 567–79, 650–84, 691–2 (the work is insufferably lengthy, but the table of contents on pp. 694–9 will enable the reader to find the material he is interested in); H. Mu¨ller-Stru¨bing, Aristophanes und die historische Kritik (1873), esp. 200–25, 236–48, 256–9. The eccentricities for which Mu¨ller-Stru¨bing is notorious do not entirely deprive this work of value. It too is anything but concise, and the table of contents (p. ix) is useful. 22

HG III 362, 364, ch. xxxi. Scho¨mann believed the introduction of sortition for the archonship was the work of Cleisthenes, whereas Mu¨ller-Stru¨bing attributed it to Aristeides. (Lugebil put it as late as Ephialtes.) Mu¨ller-Stru¨bing’s argument against Grote is that sortition could not be really democratic until pay was given to magistrates: until then, more rich men would apply, and of course the lot would favour the faction with the greater number of candidates (op. cit. 209–13). This, however, does not allow for the preliminary prokrisis (Ath. Pol. 22.5 was not available to M.-S.); and of course it would be an error to suppose that the poorer classes ever aimed at getting themselves elected in any numbers to high office—in so far as ‘party-political’ factors entered into elections at all (cf. p. 208 above), the poor would normally have been content if they could elect members of the upper classes who were reasonably dZmotiko
i and sympathetic to their needs. 23

Archons and Strategoi

221

that the really essential part of it was the promotion of the Strategoi, which entailed a certain consequent demotion of the archons, at any rate the polemarch and—in practice if not in constitutional theory—the eponymous archon. A mere change in the practice of the Assembly, in instructing the Strategoi rather than the polemarch to do this and that, might account for a good deal, but not everything. All we can be certain of is that the Strategoi now took over the military functions of the polemarch, who had remained down to 487 the commander-in-chief of the army.24 The Strategoi may well have received at the same time some of the powers we find them exercising in the late fifth and fourth centuries, notably that of summoning the Assembly;25 but here we have no information, and it is better to refrain from guessing. How far the archons other than the polemarch were legally ‘demoted’ we cannot tell: they may not have been deprived in theory of any of their powers and privileges, but of course the very fact that they were henceforward chosen by lot would make them persons of less consequence in practice. The eponymous archon in addition would inevitably lose a good deal of his influence, simply because of the appearance beside him of an important new board of magistrates, whose military functions would surely involve the exercise of a certain amount of political leadership from the first. It is possible that at this very time the eponymous archon was deprived of the presidency of the Assembly and Council, which he is generally (and perhaps rightly) believed to have enjoyed under the constitution of Cleisthenes; and it is also possible that at least the first step was taken towards reducing the judicial role of the archons (including the Thesmothetai) to the very modest one we find them exercising in the late fifth and fourth centuries. But on these points there is no clear evidence, and Wade-Gery26 may be right in supposing that the functions in question continued until the time of Ephialtes. It was the change in the method of choice of archons which got into the record—perhaps simply because of the accidents of survival. What really mattered, however, was not the demotion of the archons but the promotion of the Strategoi, who, unlike the archons, 24

Arist., Ath. Pol. 22.2. See below. Or at least co-operating with the prytaneis in doing so: see Thuc. II 59.3; IV 118.14. 26 EGH 171–9, 183–6, 188–9, 195–7. 25

222

Athenian Democratic Origins

could be re-elected year after year, so that the best men would always be available. The standard view nowadays27 is that as a result of the opening of the archonship to the lot the political importance of the archonship declined, while that of the generalship increased. But this will not do at all: it makes the Athenians behave in a quite uncharacteristically idiotic manner. They were in a very dangerous situation. There had long been unfriendly relations and intermittent hostilities with Aegina, a state which, as yet, the Athenians were far from being able to overwhelm. And now there was something infinitely worse: they were very likely to have to face within a few years a much more serious Persian attack than in 490. In these circumstances it would have been sheer lunacy for them to entrust their superior magistracies, above all that of their military commander-in-chief, to men elected by lot, even from prokritoi. (I cannot believe they ever did so, even in the sixth century: see pp. 92–9 above.) There can be no doubt, in my opinion, even if there is no positive evidence for the fact, that at least a law transferring to the Strategoi—or, conceivably, abolishing—the military powers of the polemarch was passed before or at the same time as the law changing the method of election of archons. Various other measures may have accompanied this change, giving the Strategoi greater powers; but probably fewer alterations of the constitution were required than we might think, since magistrates under the constitution of Cleisthenes, whether archons or Strategoi, may have possessed little initiative and may normally have been put in charge of particular activities and given specific instructions by the Assembly or (in these early days) the Council; and after the reform these instructions might be increasingly directed to the Strategoi rather than the archons. ‘Possibly the reformers, having decided on a new executive, abolished popular election for the archonship, in order to avert the danger of a conflict between two boards of executive officials both created by direct election.’28 Or perhaps the lot was introduced to prevent undesirable forms of competition for offices at which a much larger circle of Athenians might now 27

See most recently Hammond, HG 221–2. Hignett, HAC 187 n. 3. That the demotion of the archons and the promotion of the Strategoi was a ‘necessary concomitant’ of the opening of the archonship to the lot was seen by Macan, Hdts VII–IX II 202–3, who however was mistaken about the reasons for the reform, which were more practical and less doctrinaire than he supposed (cf. Hignett, HAC 187). 28

Archons and Strategoi

223

hope to aim, and which were still by no means unimportant and might be coveted by men who desired to profit from them.29 In this connection it is interesting to recall Aristotle’s remark30 that sortition of magistrates was introduced in place of election at ‘ &ynto toyB eriueyom
’ Heraea, o
‘ ti fi Zro enoyB [‘because they used to elect their own partisans’], and the statement in the Rhetorica ad ’ aB kai t aB mikr aB arx aB Alexandrum31 that democracies ought t ’ poll aB klZrvt aB poie&in: astas
iaston g ar to &yto [‘to appoint the majority of the magistracies, being minor, by lot, for in this way faction is avoided’]. The duties of the archons would henceforth be such as almost any member of one of the two top property classes could reasonably be expected to perform; and there was always the preliminary prokrisis, to exclude incompetents. The replacement of archons by Strategoi as the holders of the supreme political and military posts in the State was a reform which could only make for greater efficiency, and I do not see why anyone should have wanted to oppose it. Athens under the constitution of Cleisthenes had nothing like the Roman Senate or the Spartan Gerousia to advise and exercise supervision over her archons, whose term of office was limited to a single year, and who apparently could not be elected to any of the nine archonships after holding it or any other—I know of no positive proof of this, but I think it must be right, and there is certainly no indication to the contrary. The eponymous archon, whose office until 487 was the most important of all,32 could not be re-elected, ex hypothesi; but if any other archonship, especially the polemarchy, had been open to iteration, it would surely have become a still greater prize and tended to overshadow the office of the eponymous archon. It would have been theoretically possible, of course, to make any of the archonships except that of the eponymous iterative; but I suppose the Athenians felt that these magistracies, being very ancient, should not be tampered with. At any rate, in 501/0 they took the first step on the course which was to be completed in 487: ‘they began to elect the Strategoi kat a wyl
aB, one from each tribe’ (Arist., Ath. Pol. 22.2). Very different views have been taken about 29

[Ste. Croix refers to a lost or unwritten Endnote.] Pol. V 3, 1303a 13–16. 31 P. 21, ed. L. Spengel and G. Hammer ¼ Anaximenes, Ars. Rhet. II 14, ed. Fuhrmann. Cf. p. 94 above. 32 See, for the early sixth century, Arist., Ath. Pol. 13.2. I cannot see why this position should have altered much before 487. 30

224

Athenian Democratic Origins

this laconic statement: for instance, (1) that the Strategoi had existed before, from 508/7 (if not earlier), as colonels of the tribal regiments, elected by the tribes, and that what happened in 501 was the transfer of their election, still one from each tribe, to the whole Athenian people;33 (2) that the Strategoi were a new creation in 501, but that each was elected by as well as from one tribe;34 (3) that Aristotle is just wrong, and that the Strategoi, who must have been elected by as well as from their own tribes from 508/7 onwards, were not elected by the whole people until 487/6.35 Hignett36 says that Aristotle cannot have meant the Strategoi were a new creation in 501, ‘since he believed that they had existed in the time of Drakon’; but he adds in a footnote, ‘unless we assume that the ‘‘constitution of Drakon’’ is a last-minute addition to the Ath. Pol., unknown to the author when he was writing 22.2’ and this very reasonable admission robs the objection of any force. Hignett also declares that ‘the turn of the sentence indicates that it is a change in the method of appointment to which he [the author of the Ath. Pol.] is referring’. I cannot see the slightest ground for this: I believe Aristotle is referring to the first appointment of Strategoi.37 33 Thus, e.g., W. Schwahn, in RE Suppl. VI (1935) cols. 1071–3, and many others, including Busolt-Swoboda, GS II 881, who cannot believe Arist.’s statement that the polemarch had remained commander-in-chief and suppose that his continued supremacy was merely nominal. 34 Raubitschek, OO, esp. 223; Hammond, HG 190; Eliot, CDA 146–7 n. 18. Against this view see Kahrstedt, UMA 46 n. 4, referring to ‘die merkwu¨rdig weit verbreitete moderne Theorie, das Beamte, die fu¨r einzelne Phylen zusta¨ndig sind, Taxiarchen, Phylarchen, fru¨her die Strategen, von den Phylenversammlungen gewa¨hlt worden seien’, which is ‘in Athen unbekannt’. (His reference to Ath. Pol. 42.4 must be a mistake for 42.2.) Kahrstedt is right. A statement that the Athenians made an appointment kat a wyl
aB, either by lot (Ath. Pol. 22.5; 48.1) or (as here) by election, can only mean that the magistrates concerned were elected by the whole Assembly, each from one tribe. If the individual tribes had made the election, we should have some such phrase as in Ath. Pol. 31.1 (where the councillors, 40 from ’ prokr
itvn oyB ‘ an ’ e
‘ lvntai oi‘ wyl
etai [‘from preeach tribe, are to be chosen ek selected candidates chosen by the tribe-members’]) or 42.1 (oi‘ de dZm
otai ’
’ ayt ’ vn katZg
oroyB ai‘ ro &yntai p
ente andraB ej & [‘the deme-members select five of their own members as prosecutors’]), or Aeschin. III 30 (o y‘B ai‘ wyla
i kai ai‘ ’ eayt ‘ & tritt
yeB kai oi‘ dZmoi ej vn & ai‘ ro &yntai [‘whom the tribes and the trittyes and the demes select from among themselves’]). 35 Hignett, HAC 170, who finds it ‘tempting to assume this’. 36 HAC 169 & n. 3. 37 The words stratZg
oB, stratZge&in, stratZg
ia are used several times of men who commanded Athenian armies in the field before the fifth century: e.g. Hdts I 59.4; Arist., Ath. Pol. 14.1; 17.2; 22.3, all in relation to Peisistratus; and see the other

Archons and Strategoi

225

Were the Strategoi ever merely colonels of the tribal regiments, as most modern writers have supposed? I believe they were always, from the very first, general staff officers, with a sphere of competence that was not limited to the regiment of each general’s own particular tribe (although he would doubtless march at the head of that regiment into battle), but included the whole army. Two arguments are strongly in favour of this: the statement of Herodotus (V 69.2) that Cleisthenes ‘made ten phylarchs instead of four’ (implying that phylarchs continued to exercise the same military38 functions as before: the command of their tribal regiments), and the etymology of the word stratZg
oB. Tribal commanders might be called phylarchs (the name borne in the late fifth and fourth centuries by the commanders of the tribal cavalry squadrons), or taxiarchs (the commanders of the tribal infantry regiments); but a stratZg
oB is surely a man who leads, solely or jointly, an army or an expedition and not a mere segment of it. The title stratZg
oB is known from about the mid-seventh century,39 but as far as I am aware there is no evidence of its being used before the Hellenistic period in any other sense than commander of an army or an expedition,40 except in special circumstances in Crete, where the office of the startag
etaB who appears in an inscription oB (or st
artoB)—perhaps a of Roman times41 as head of a start
military division of the population42—may go back to early times, since the start
oB appears at least as early as the fifth century.43 As I see it, although the Athenians did not make the final change until 487, they first decided in 501 that they were dissatisfied with examples quoted by A. Hauvette-Besnault, Les Strate`ges athe´niens (Bibl. des e´coles fr. d’Ath. et de Rome, fasc. 41, 1885) 4–8. But in no case is there any reason to think that the stratZg
oB-words are being used in a technical sense: they simply describe the supreme commander of the army, who may have been the polemarch or a man specially appointed for the purpose by the Assembly or by the polemarch. 38 Phylarchs never appear at Athens in anything but a military capacity. The civil heads of the Cleisthenic tribes were Epimeletai; the old Ionic tribes had Phylobasileis. 39 Archilochus, fr. 114.1 (West). 40 Against Wilamowitz’s interpretation of Eur., Suppl. 713 (panti Kranaid vn & stratv fi & ), see Hauvette-Besnault, op. cit (in n. 37 above) 9 n. 4. 41 IC IV 80 (¼ SGDI 4985) 4–5. 42 See Hesych., s.v. st
artoi
ai‘ t
ajeiB to &y pl
ZuoyB [‘Startoi: the ranks of the people’]. For the quite different view of De Sanctis and Guarducci, see IC IV p. 185. 43 IC IV 72 (Leg. Gortyn.), V 5; cf. IC IV 80.7 (and perhaps 142.2); IC I xviii 11.1.

226

Athenian Democratic Origins

a system of command in which an annually changing polemarch, who could not be re-elected, had under him only the commanders of tribal regiments—who, I believe, continued to be called phylarchs44 until the creation of an organised State cavalry force in the 450s, at which time taxiarchs45 were created to command the tribal infantry regiments (t
ajeiB),46 and the older name ‘phylarchs’ was attached to the commanders of the tribal cavalry squadrons,47 membership of which had, of course, superior social prestige.48 At the time of Marathon the situation was still that the polemarch commanded the army, advised by a general staff of ten Strategoi, each elected by the whole people from his own tribe. Herodotus’ 44

The term is first used by Hdts V 69.2. Cf. n. 48 below. The phylarchs may have been elected by their tribes, as in later times, or perhaps chosen by the polemarch. 45 On whom see Busolt-Swoboda, GS II 891 & n. 1. The earliest surviving source to use the word taj
iarxoB is Aesch., fr. 182 (Radt), ap. Athen. I 11d, from the Palamedes; but in the most general sense and not with reference to Athens. Hdts (e.g. VIII 67.2; IX 42.1) applies the term to subordinate Persian commanders. Aristoph. (Ach. 569 etc.) and Thuc. (IV 4.1 etc.) are the first surviving writers to use taj
iarxoB in the technical Athenian sense. 46 On which see Busolt-Swoboda, GS II 881 & n. 2. 47 On these and the superior cavalry commanders, the hipparchs, ibid. 1128–9. 48 Cf. pp. 15–16 above. It has often been thought that taxiarchs existed much earlier, from 487/6 or soon after (thus Hignett, HAC 348, cf. 176), or from the years after the great increase of the fleet in 483/2 (Busolt-Swoboda, GS II 890–1), or from soon after the invasion of Xerxes (Wilamowitz, AuA II 88); but if so it is difficult to see what the function of the phylarchs was before the existence of formal cavalry units (on which see pp. 15–16 above). The statement of Hdts V 69.2 that Cleisthenes ‘made ten phylarchs instead of four’ must surely refer to more important officers than the mere commanders of cavalry squadrons of his own day; his statement is more easily understandable if the phylarchs, after as well as before the late sixth century, remained the military commanders of their tribes (i.e. of infantry and mounted infantry), even if the Strategos might personally lead his own tribe in battle. If so, of course, the Strategoi must have been more than mere tribal commanders. I cannot see that either the creation of the Strategoi in c. 501 or the elevation of the Strategoi in 487 need have made any difference to the system of command within the tribe, whereas the creation of a formal cavalry force would be an obvious opportunity for a reorganisation. The fact, surprising at first sight, that in the second half of the fifth century a man with the exalted title of phylarch commands only the cavalry of his tribe (militarily unimportant as they were) is best explained, to my mind, on the assumption that (as Hdts V 69.2 implies) the phylarch retained command of the whole military contingent of his tribe until the moment when the cavalry unit came into being. The separation of tribal cavalry and infantry commands was now necessary, and the phylarch took over the cavalry command as having greater prestige, although of course in actual warfare the infantry commander, the newly created taxiarch, had a position of far greater responsibility and military importance.

Archons and Strategoi

227

account of the Athenian command in the Marathon campaign49 is muddled and unsatisfactory, probably owing in part to his information being obtained from a source which over-stressed the role of Miltiades.50 Despite attempts to save his credit,51 Herodotus clearly believed that at this date the polemarch was already chosen by lot.52 He seems to have conceived the polemarch as having a mere casting-vote in the event of a division of opinion among the Strategoi;53 but he does not say clearly that the Strategoi were equally divided on the occasion in question,54 although he perhaps implies it. Now in 490 Athens may have been fortunate in her polemarch, but the system of non-iteration would mean that at any given moment a high proportion of the men best qualified to command the army would already have filled the polemarchy or one of the other archonships (especially the eponymous) and would consequently be ineligible for the post of commander-in-chief. The same considerations, of course, apply to the principal political office, that of the eponymous archon. The Athenians, therefore, made a most natural and sensible decision; they would raise the reelectable Strategoi to the supreme command in war (from which political pre-eminence was scarcely separable), and altogether deprive the polemarch of his military functions. The position of Themistocles in 480 is the first certain evidence of this change; but, as I have shown, it cannot have taken place later than 487, and it may even have been a year or so earlier. 49

Hdts VI 109–11. Some modern interpretations of Hdts have been very foolish. Lugebil, op. cit. (in n. 21 above) 581–667, seems to have been the first to prove in detail that Callimachus the polemarch was ‘Oberfehlshaber’ at Marathon; but he wrote at enormous and quite unnecessary length, and some of his arguments were fallacious. The best treatment is that of H. Berve, Miltiades (¼ Hermes, Einzelschr. 2, 1937) 75–91 (esp. 78–87): this seems to me conclusive. See also Hignett, HAC 170–3, esp. 171. 51 See e.g. J. L. Myres, Herodotus Father of History (1953), 208: what was decided by lot was merely the allocation of the nine successful candidates to the respective archonships. This theory, which was first produced by W. Oncken, will appeal only to those who are determined to put the most favourable interpretation on everything Hdts says. 52 Hdts VI 109.2. 53 See Hdts VI 109.2,4; 110.1. 54 See esp. Hdts VI 109.2: ‘the worse view began to prevail.’ 50

228

Athenian Democratic Origins

Again we have vindicated the good sense of the Athenians, who in 487 were not just indulging an unrestrained passion for democracy, as so many scholars have thought, but were taking a very necessary (and long overdue) step towards reforming their whole system of military command and political leadership. Themistocles will surely have supported the measure and may well have been its original promoter. He had been archon in 493/255 and will doubtless have wanted to exercise the chief power in an office to which he could be re-elected again and again. It would be absurd, however, to conceive the reform primarily as a selfish move on the part of Themistocles. It is all too common for modern writers dealing with Athenian history to speak of a measure as being carried by an individual, without asking themselves how he managed to persuade the Athenians of its desirability.56 It is quite likely that a large number of Athenians in the early 480s wanted Themistocles to command their fleet; but their understanding of the need for reinforcing the system of command is likely to have been deeper than that. Finally, I would express the hope that even those who accept (as I do not) Ath. Pol. 8.1 and believe that Solon introduced ’ prokr
itvn [‘sortition from pre-selected candidates’] kl
ZrvsiB ek will be willing to concede, as Wade-Gery evidently is,57 that election by lot, at any rate of the polemarch, would have been a very different matter in the 480s, when Athens was in quite a different position and a much more dangerous one, and being a general was likely to involve being an admiral as well—a post requiring a far greater range of skills than the command of a hoplite army.58 Whatever we may think the situation was earlier, we can still say that to introduce sortition in 487 for an office of great military ’ prokr
itvn, would have been or political importance, even ek unthinkable.

55 See R. J. Lenardon, ‘The Archonship of Themistocles’, in Historia 5 (1956) 401–19. 56 This often has the effect of obscuring the reasons for the passing of a measure: a particularly good example will be found in Hignett, HAC 284, where concentration upon the alleged ‘selfish folly’ of Cleophon prevents the raising of the really important question: what caused the Athenians to vote for Cleophon’s proposal. 57 EGH 114. 58 Cf. Arist., Pol. V 9, 1309b 4--8: military capacity is a rarer quality than moral ’ goodness (epie
ikeia).

Afterword

229

AFTERWORD Ostracism The evidence that forms the basis for the study of ostracism was transformed by the discovery of almost 9,000 ostraka in the German excavations in the Kerameikos between 1965 and 1968. Only a minority have thus far been published (some of the most interesting are discussed and illustrated in Brenne, ‘Ostrakophoria’), but a list of the totals for each candidate is given by Willemsen and Brenne; for historical discussion see Lewis, 603–6. The best re´sume´ of the earlier position was the wellillustrated article by Vanderpool. Substantial discoveries continue to be made in the Agora, to add to the 1,337 ostraka from there and thereabouts published by Lang: see most recently J. M. Camp in Hesperia 68 (1999) 268–74. Developments can be easily monitored through SEG: see especially SEG xli 16 and xlvi 78–103. Siewert will now be a basic work of reference (note recent bibliography on pp. 518–20). It contains all testimonia pre-dating 322 on ostracism, with detailed commentary; totals of actual ostraka are listed, and all those containing more than a simple name are discussed. Brenne, Ostrakismos, is a prosopography of ‘candidates’. No substantive argument of Ste. Croix has been invalidated by new evidence, but his statistics are now outdated. Some feel that the extent of the ‘scatter-vote’ revealed by the new material renders more credible the traditions dismissed by Ste. Croix in n. 86 about the ostracism of various minor figures; see for instance Rhodes, CAAP, 271, 342, and, for Damon, Wallace. Lewis, ‘The Kerameikos Ostraka’, has used the new evidence to make a very strong case that Megakles was ostracised a second time, as Lysias claims (14.39). On the ancient tradition about the origins of ostracism, Ste. Croix was an early adherent to a view which is now widely accepted: ‘since [the articles of Sumner and Dover] the consensus of scholarly opinion has accepted the obvious: Aristotle and Androtion were both saying the same thing and both ascribed the origin of ostracism to Kleisthenes’ (Harding, ATA, 96, with references). Most scholars therefore now accept the attribution to Cleisthenes, though Marr and Worswick reject the testimony of Aristotle and Androtion and date the introduction of the procedure to the early 480s. The report uncovered in a Vatican gnomologium by Keaney and Raubitschek, whereby ostracism originally took place in the boule, with a minimum of 200 votes required against the victim, has generally been rejected (Harding, ATA, 96–7). Stanton and Knight independently suggest that Cleisthenes invented ostracism as a weapon to use against Isagoras during the political

230

Athenian Democratic Origins

struggle of 508 (Hdt. 5.66–76); but Isagoras called in Cleomenes and wrought his own downfall. This theory is a special (Ste. Croix would doubtless have said too special) application of Ste. Croix’s view of ostracism as a safety-guard against stasis; it was indirectly influenced by Ste. Croix through Forrest, EGD, 201–2. Otherwise the theories in play are much the same as when Ste. Croix wrote. One cannot imagine him being very attracted by Vernant’s suggestion, 124–6, that ostracism had something in common with the ritual expulsion of scapegoats. Archons and Strategoi Buck offers a review of previous theories which shows Ste. Croix’s account to be rather selective; but his own new theory, that until 487 archons were chosen by the Areopagus from a list of pre-elected candidates, is plucked out of the air. Badian, like Ste. Croix, denies the ‘democratic’ interpretation of the reform of 487, but differs sharply over the importance of the Strategoi at this date. Badian accepts the old explanation ascribed to W. Onckena (mentioned and rejected by Ste. Croix p. 227 n. 51) of Herodotus’ description of Callimachus as ‘the polemarch appointed by lot’ at Marathon: a board of nine archons was chosen by election, but lot was then used to distribute the archonships among the nine. Badian then draws a political inference: a system whereby one sought electionship not to a particular archonship but merely to a board of nine, from which one might be assigned by lot to an archonship of minor importance, would devalue the archonship as a target of political ambition, and must have been so designed by its hypothetical author, Cleisthenes. The reform of 487 merely took the depoliticisation of the archonship a small stage further. But the aim cannot have been, Badian argues, to depress the archonship in order to elevate the generalship. ‘The reform . . . merely increased the vigilance with which the Cleisthenic constitution regarded authority. No one could have had the slightest premonition of the rise of the strategia.’ Ste. Croix would not, we think, have readily accepted this last claim in relation to the sharp-witted Athenians. Both Badian and Frost, 114–15, argue that the eponymous archonship was demonstrably already of diminished importance in the period 507–487; Rhodes, CAAP 273, judges this to be unknowable. Ste. Croix believes that the reform laconically described in Ath. Pol. 22.2 was the first institution of Strategoi as a regular, annually-elected board; also that they were ‘general staff officers’ from the beginning. a The theory is known to modern scholarship from a brief reference in Macan, Hdts IV–VI (1895), ii 157.

Afterword

231

Both propositions are accepted by most recent writers (see Rhodes, CAAP, 264), though only the latter by Hammond, 112–14. But such writers tend to see the institution of the new board as having immediately eroded the powers of the polemarch, whose ‘command’ was already at Marathon only titular,b whereas for Ste. Croix the polemarch still exercised effective command and had to be demoted by the reform of 487. The difference here is one of emphasis, since it is generally accepted on the authority of Herodotus (6.109.2, 111.1) that the polemarch still had some military functions in 490 that he later lost (precisely in 487 for Ste. Croix, though not for Badian). In n. 48 Ste. Croix suggests that the phylarchs attested for Cleisthenic Athens by Hdts V 69.2 were tribal military commanders, and only acquired their familiar function as tribal cavalry commanders when a state cavalry force was organised in the 450s. This suggestion neatly avoids difficulties which other explanations of the passage are faced by. Phylarchs can scarcely have been civil magistrates in Athens, for reasons given by Ste. Croix himself in n. 38.c Myres’s view (adopted by Gschnitzer, 1088–90), that Herodotus’ four phylarchs are the four old phylobasileis and his ten phylarchs the ten Strategoi becomes even less attractive than it always was if one acknowledges (see above) that the Strategoi had never been tribal commanders. As for the possibility that the phylarchs had always been tribal cavalry commanders, recent investigators have been willing to allow the existence of ‘true cavalry’ earlier than Ste. Croix does (see p. 72 above). But would such a force have been large enough to require organisation in distinct tribal units? Badian, E., ‘Archons and Strategoi’, Antichthon 5 (1971) 1–34 Brenne, S., ‘Ostraka and the Process of Ostrakophoria’, in W. D. E. Coulson and others, The Archaeology of Athens and Attica under the Democracy (1994), 13–24 —— , Ostrakismos und Prominenz in Athen (2001). Buck, R. J., ‘The Reforms of 487 B.C. in the Selection of Archons’, CP 60 (1965) 96–101 Frost, F. J., ‘Themistokles’ Place in Athenian Politics’, CSCA 1 (1968) 105–24 Gschnitzer, F., RE suppl. xi (1968), 1067–90, s.v. Phylarchos Hammond, N. G. L., ‘Strategoi and Hegemonia in Fifth-Century Athens’, CQ 19 (1969) 111–44 (repr. in his Studies in Greek History, 1973) Keaney, J. J., and Raubitschek, A. E., ‘A Late Byzantine Account of Ostracism’, AJP 93 (1972) 87–91 b Badian, 28–9; Rhodes, CAAP, 264–5; with some difference in detail Hammond, 114–23. c Lysias 12. 43, the supposed instance of non-military phylarchs adduced by G. R. Bugh, The Horsemen of Athens (1988), 5 n. 14, is very unpersuasive as a counter-case: these are surely cavalry commanders of familiar type.

232

Athenian Democratic Origins

Knight, D. W., Some Studies in Athenian Politics in the Fifth Century B.C (1970), 21–4 Lang, M. L., The Athenian Agora 25: Ostraka (1990) Lewis, D. M., ‘The Kerameikos Ostraka’, ZPE 14 (1974) 1–4 ¼ Selected Papers in Greek and Near Eastern History (1997), 110–13 —— postscript to the second edition of A. R. Burn, Persia and the Greeks (1984) Marr, J. L., and Worswick, N. D., ‘The Institution of Ostrakophoria at Athens’, Ostraka 3 (1994) 285–90 Myres, J. L., ‘Cleisthenes in Herodotus’, Me´langes Gustave Glotz (1932), II 657–66 Rausch, M., Isonomia in Athen: Vera¨nderungen des o¨ffentlichen Lebens vom Sturz der Tyrannis bis zur zweiten Perserabwehr (1999) (non vidimus) Siewert, P. (ed.), Ostrakismos-Testimonien (Historia Einzelschrift 155, 2002) Stanton, G. R., ‘The Introduction of Ostracism and Alcmeonid Propaganda’, JHS 90 (1970) 180–3 Vanderpool, ‘E., Ostracism at Athens’, in Lectures in Memory of Louise Taft Semple, II (1973), 217–70 Vernant, J. P., ‘Ambiguı¨te´ et renversement. Sur la structure e´nigmatique d’ ‘‘Oedipe-Roi’’ ’, in Vernant and P. Vidal-Naquet, Mythe et trage´die en gre`ce ancienne (1973), 101–31 Wallace, R. W., ‘Damon: A Music Philosopher Ostracized?’, forthcoming in P. Murray and P. Wilson, eds., Music and the Muses in Classical Athenian Culture Willemsen, F., and Brenne, S., ‘Verzeichnis der Kerameikos-Ostraka’, AM 106 (1991), 147–56

6 The Athenian Citizenship Laws

What survives is an informal working paper, the very provisional character of which Ste. Croix stresses in an accompanying note. Certain problems were postponed for discussion in eight ‘long notes’, which were apparently never brought close enough to completion to be given to a typist. We have marked with a star points where Ste. Croix referred to one of the projected notes.

Most of what has been written in modern times about the Athenian citizenship laws is of very poor quality. Even Jacoby’s judgement entirely deserts him.1 Gomme2 makes an essential contribution, but in my opinion fails to get to the heart of the problem. For an extreme but by no means unusual view it is worth looking at Adcock in CAH V 167–8: accepting the most common interpretation of the events of 445/4, as ‘the retrospective enforcement’ of Pericles’ citizenship law, he speaks of ‘this narrow policy’ as ‘a grievous error . . . , a great reproach on the state-craft of Pericles, a denial of Athens’ past, and a menace to Athens’ future’. A view to which many modern historians have subscribed is expressed by Walker in CAH V 102–3, that ‘the restriction of the franchise . . . meant the assertion of the principle of privilege in its most offensive form’, and although ‘the pretext for the law of Pericles is stated to have been the excessive number of the citizens, . . . the real motive was to enhance the value of the lucrative privileges attaching to the franchise, by limiting the number of those entitled to share in them.’ Cf. Hammond,3 who adds, ‘The discrimination against non-Athenian mothers was likely to apply for the most part to women of the allied states.’ A different view is expressed by Hignett: the Athenians were trying ‘to preserve the racial purity of the citizen-body’,4 ‘to preserve the purity of the citizen stock from contamination by undesirable alien elements’.5 The most absurd 1 2 4

Jac. i 462–82; ii 372–88, on Philoch. 328 F 119. 3 EGHL ch. iv, pp. 67–88. HG 301–2. 5 HAC 345, cf. 346–7. Ibid. 255.

234

Athenian Democratic Origins

theory of all, to my mind, is set out in the most detailed recent examination of the law of Pericles and its supposed consequences, that of Jacoby,6 who thinks the law was formulated as a potential weapon against Cimon, the most conspicuous mZtr
ojenoB [‘alien on his mother’s side’] of the day, should he prove politically troublesome. Cf. the earlier (1935) statement of Schwahn, in RE XVI ii, col. 2056: ‘Vielleicht was das Gesetz des Perikles in erster Linie eine Drohung gegen Kimon’ [‘Perhaps Pericles’ law was first and foremost a threat against Kimon’]. I begin by setting out the main facts.

A. N
ouoi and MZtr
ojenoi [‘Bastards’ and ‘Aliens on the mother’s side’] 1. The n
ouoB, the opposite of the gn
ZsioB (Ar. Birds 1650), was a man whose parents had not been ‘legally married’: see Ps.-Dem. XLIV (Leoch.) 49; Poll. III 21. In so far as the institution of marriage existed at Athens in any technical legal sense, a marriage ’ yZ (egg
’ yZsiB) or epidikas
’ must have been created either by egg
ia: ’ iklZroi, the latter, ‘adjudication’, applied only to heiresses, ep
’ yZ and the vast majority of formal marriages were created by egg
(a term of which neither ‘marriage’ nor ‘betrothal’ is an adequate translation), the giving of the bride by her legal representative, her k
yrioB (usually her father), and her acceptance by the bridegroom, for a union to be consummated either immediately or at a future date. A law cited by Ps.-Dem. XLVI (Steph. 2) 18 (cf. Hyper. III ‘ an ’ eggy
’ ’ i dika
ioiB d
(Athenog.) 16, col. 7) reads: Zn Zsfi Z ep amarta ’ pat ’ ’

einai Z Zr Z . . . ek ta
ytZB einai pa&idaB gnZs
ioyB [‘if a woman is betrothed/wedded lawfully by her father or . . . then her children are to be legitimate’]. The mZtr
ojenoB (apparently not a technical term) was a man whose mother was a non-Athenian. The two classes, n
ouoi and mZtr
ojenoi, were not at any time identical: some n
ouoi were of Athenian parentage on both sides (although their parents were not legally married) and were therefore not mZtr
ojenoi, and down to 451/0 (and apparently for a time again, in practice, towards the end of the fifth century) some mZtr
ojenoi ouoi, because their parents, although not both Athenwere not n
ians, had been legally married, like Agariste to Megacles (see B 3 6

As cited above, esp. i 478–80.

The Athenian Citizenship Laws

235

below). After 403/2 (see E 2 below) all mZtr
ojenoi were n
ouoi, although all n
ouoi were still not mZtr
ojenoi. 2. By a law attributed to Solon,7 if there were legitimate sons, ’ pa&ideB gn
Zsioi, the n
ouoB had no agxiste
ia [‘right of kinship’] and could not inherit at all. (Whether he could inherit some share in the absence of pa&ideB gn
Zsioi is doubtful.) For the re-enactment of the Solonian law in 403/2, see E 3 below. 3. In the fourth century, probably most phratries and gene¯, and perhaps all,* excluded n
ouoi altogether: they insisted that all en’ trants should be sons of a citizen father by a valid marriage, ej ’ ZB ’ ’ vB) & kai eggyZt & (or gametZB) & ast ZB gynaik
oB (or gegon vB oru & [‘from a married citizen women’, or ‘properly born’].8 From Ar., Birds 1668–70, we can presumably infer that in the late fifth century a n
ouoB could not (or would not normally) be admitted to the phratry. Omission by a father to register a child, even a daughter, in his phratry might allow doubts to be raised concerning its legitimacy: see Isae. III (Pyrrh.) 75–6. I know of no certain evidence before the fourth century, but a very puzzling text,* Craterus FGrH III B 342 F 4 (ap. Harp. s.v. naytod
ikai), suggests to me that in the first half of the fifth century at any rate admission could perhaps be obtained to a phratry, or to some phratries, if either parent was an Athenian: Craterus speaks of a prosecution ’ an . . . tiB ej ’ amwo& ’ (before the naytod
ikai) e in j
enoin gegon vB wratr
izfi Z [‘if anyone born of two aliens is a member of a phratry’]. 4. Those n
ouoi who were not mZtr
ojenoi were not disqualified from obtaining citizenship at any time. The contrary is often asserted but there is no evidence at all against the view I have stated and a good deal of evidence in favour of it.* ouoi or not, were not disqualified from 5. mZtr
ojenoi, whether n
obtaining citizenship until 451/0, and they were again admitted in practice for a time towards the end of the fifth century.

B. Citizenship before 451/0 B.C. There is scarcely any evidence for the years before 451/0, which I would divide into three periods: 7

Ar., Birds 1649–70, esp. 1660–3. See Isae. VIII (Ciron) 19–20; VII (Apollod.) 15–17 (cf. XII [Euphil.] 9); Ps. Dem. LIX (Neaer.) 60; Dem. LVII (Eubul.) 54 (cf. 41, 43, 69); SIG3 921(¼IG II2 1237). 109–11; Poll. III 21; cf. Andoc. I (Myst.) 127; Isae. VI (Philoct.) 25. 8

236

Athenian Democratic Origins

1. Before Solon. I suggest that until major political decisions came to be constitutionally taken by majority vote of the whole people, citizenship of the polis, Athens, might be something not very clearly defined and anyway rather unimportant: the talisman would be membership of a phratry, and if possible a genos in addition. But once the decision was taken to settle major political questions by actually counting votes, membership of the polis would become something that needed careful definition, because otherwise there would be nothing to prevent non-citizens from penetrating the Assembly. 2. From Solon to Cleisthenes. Just as from Cleisthenes’ time onwards it was membership of a deme which conferred citizenship of Athens, so it is generally believed that before Cleisthenes the unit which gave citizenship to its members was the phratry. In the absence of any real evidence, this seems to me by far the most probable hypothesis. I know of no evidence concerning the qualifications for membership of a phratry at this time: see however A 3 above. 3. From Cleisthenes to Pericles. Membership of a deme and nothing else now makes a man a citizen – unless of course he is given the citizenship by special decree of the Assembly (pol
itZB poiZt
oB or dZmopo
iZtoB). But in virtue of what qualification did he become a member of a deme? We cannot pretend we know all the rules. Evidently it was sufficient to have a citizen father. As far as I know, there is no evidence at all whether citizenship could be claimed through an Athenian mother alone: this was the rule in some democracies.9 It looks as if in the mid-fifth century admission to a phratry (or to at least some phratries) could perhaps be gained if either parent were a citizen (see A 3 above); and if such a rule was general before Cleisthenes, it might have been taken over by the demes and even preserved right down to 451/0. But my own guess would be that down to 451/0 citizenship depended on the father alone. It is sometimes said that the law of Pericles merely represented a return to an earlier state of affairs; but if this is so (and I do not myself believe it), it was a very much earlier state of affairs and not merely a situation which had come about after the Persian wars,10 9 10

Arist., Pol. 1278a 27–8, 32–3; 1319b 6–10; 1275b 22–3. As supposed, e.g., by Rogers, on Ar., Birds 1650.

The Athenian Citizenship Laws

237

for in the sixth and fifth centuries we hear of several prominent Athenians marrying foreign wives, by whom they had children against whom there are no recorded imputations of doubtful citizenship. I give the most conspicuous examples: ’ (a) Cylon had married (egegam
Zkei) the daughter of Theagenes, tyrant of Megara: Thuc. I 126.3; Paus. I 28.1. (b) At some time in the 570s (Hdts VI 126–31.1) the Alcmaeonid Megacles had married Agariste, the daughter of Cleisthenes of Sicyon, by whom he had a son Cleisthenes, ‘the lawgiver’, almost certainly archon in 525/4.11 According to Hdts VI 130.2, the marriage was announced by the bride’s father with the ’ ’ Zn ’Agar
istZn n
words, . . . Megakl
ei eggy v& pa&ida t Zn em omoisi to&isi ’AuZna
ivn [‘I betroth my daughter Agariste to Megacles according to the laws of the Athenians’]. (c) Themistocles is said to have been the son of Neocles by a foreign woman, described as a Thracian, Carian or Acarnanian (Plut., Them. 1.1; Nep., Them. 1.1–2; Athen. XIII 576cd), and yet was very probably archon in 493/2.12 Themistocles is also called a n
ouoB: see Plut., Them. 1.3. Cf. Plut., Mor. 753d; Ael., VH XII 43. (d) Miltiades, the victor of Marathon, married a Thracian princess, Hegesipyle daughter of Olorus, and by her became the father of the great Cimon (Hdts VI 39.2; Plut. Cim. 4.1). Stesimbrotus (FGrH II B 107 F 6; cf. Plut., Per. 29.2) alleged that the mother of Cimon’s twin sons Lakedaimonios and Eleios was a woman of Cleitor; but this is no doubt a fiction.13 (e) Thucydides son of Melesias very likely married Cimon’s sister, a daughter of the marriage of Miltiades and Hegesipyle, and another daughter was probably the maternal grandmother of Thucydides the historian: see the stemma by Cavaignac and WadeGery, in Wade-Gery, EGH 246. It is also said by Marcellinus, Vita Thuc. 19, that Thucydides the historian married a rich Thracian wife (cf. 17). And see Plut., Cim. 4.2–3; P. Oxy. XV 1800. fr. 2, 64 ff.14 M/L 6c ¼ Fornara 23c 3. See R. J. Lenardon in Historia 5 [1956] 401–19. See Diod. Perieg., FGrH III B 372 F 37, also ap. Plut., Cim. 16.1. 14 [J. K. Davies, Athenian Propertied Families (1971), 233, rejects the marriage of Thucydides to a Thracian heiress; it is unclear whether Thucydides son of Melesias married a sister of Cimon, or a daughter: ibid. 232.] 11 12 13

238

Athenian Democratic Origins

It is perhaps also worth mentioning that according to Aristotle & (Ath. Pol. 17.4) Peisistratus ‘married’ (e’
gZmen, gZmai) his Argive wife, Timonassa. Yet Aristotle himself (Ath. Pol. 17.3) distinguishes the two sons he had by his (earlier) Athenian wife as d
yo ’ tZB & gametZB & [‘two by his wife’] (Hippias and Hipparchus), in ek ’ tZB & ’Arge
iaB [‘two by the Argive woman’], contrast with the d
yo ek and moreover Thucydides (VI 55.1) calls Hippias and Hipparchus ’ the gn
Zsioi adelwo
i [‘legitimate brothers’], and Herodotus ’ (V 94.1) speaks of Timonassa’s son Hegesistratus as a n
ouon . . . ej ’Arge
iZB gynaik
oB [‘bastard . . . by his Argive wife’].15

C. The Citizenship Law of Pericles, 451/0 B.C. 1. This law provided that no one could hold the citizenship unless both his parents were citizens: Arist., Ath. Pol. 26.4; Plut. Per. 37.3; Ael., VH VI 10; Suid., s.v. dZmopo
iZtoB. The only ancient source which explicitly gives a reason for the law is Ath. Pol. 26.4: & it was di a t o plZuoB tvn & politvn & [‘due to the excessive number of citizens’]. This opinion has been accepted in recent times, as far as I can see, only by Gomme and Latte.16 I do not know why Jacoby17 should say, ‘we must probably admit that this was the first law not only in Athens but in all Hellas to demand that both parents must belong to the same city, if the State is to acknowledge the children as citizens’. Aristotle (Pol. 1278a 28; 1319b 6–12) certainly seems to conceive oligarchies as being less ‘liberal’ in their citizenship policy than democracies. 2. If we now combine Schol. Ar., Wasps 718 (citing Philoch., 328 F 119), with Plut., Per. 37.4, we obtain the following picture. In 445/4 the rebel Egyptian prince Psammetichus made Athens a present of 30,000 medimni of corn (Plut.: 40,000 med. of barley), which was distributed to 14,240 Athenians, 4,760 being now disqualified (par
eggrawoi). (Philochorus may not have meant ‘disqualified as citizens’: see Jacoby i 464. Plutarch of course thought they were not merely struck off the citizen registers but sold into slavery.) Whether the law of Pericles was put into operation retroactively, against mZtr
ojenoi born before 451/0, is disputed. It has been held that the large numbers said to have been 15

[On this problematic case see Rhodes, CAAP, 225–6 and D. Ogden, Greek Bastardy (1996), 45–6.] 16 17 EGHL 86–8; RE XVII i 1071. Jac. ii 387 n. 69.

The Athenian Citizenship Laws

239

disenfranchised in 445/4 were retrospective. I do not accept this assumption.* 3. Ar., Birds 1642–70. This passage, of great interest for the Athenian law of inheritance in the late fifth century, touches at only one point (and indirectly) upon the question of citizenship: in lines 1651–2 Peisthetaerus tells Heracles that he is a n
ouoB because he is the son of a j
enZ (Alcmene, a mortal). As we shall see presently (E 2 below), the law of Aristophon in 403/2 specifically declared that anyone whose mother was not a citizen was a n
ouoB. In the light of the passage from the Birds, we can recognise this as the re-enactment of an earlier law, which perhaps in the late fifth century had not been strictly observed. But it is in his capacity to inherit from his father Zeus that Heracles is a n
ouoB: the passage has really no bearing on qualification for citizenship.

D. The late fifth century There are two different reasons for thinking that the citizenship law of Pericles was not properly enforced in the later years of the fifth century: (a) the wording of the law of Nicomenes (see E 1 below), which, in re-enacting Pericles’ law, expressly exempts those born before 403/2,18 and (b) the fact that there are several examples (or possible examples) of fourth-century Athenians who are known to have had foreign mothers (in the sense that the mothers were not children of Athenian parents according to the law of Pericles), but were nevertheless allowed to retain the citizenship because they or their mothers* were born before 403/2. The most obvious examples are as follow: (a) Timotheus, very prominent in the first half of the fourth century, is said to have been the son of Conon by a Thracian woman (Athen. XIII 577ab). He was no doubt born before 403/2. (b) Demosthenes’ mother, Cleobule, was the daughter of Gylon, an Athenian who while in exile in the Crimea towards the end of the fifth century married a rich wife who was certainly not an Athenian: she may well have been a ‘Greek’ of Panticapaeum or one of the other cities in the Bosporan kingdom, although of course she is referred to slightingly by Aeschines (III 172) as a Scythian, whose daughter married Demosthenes’ father in defiance of the laws of Athens (cf. Aeschin. II 83, where Demosthenes is called a 18

See Dem. LVII [Eubul.] 30.

240

Athenian Democratic Origins

n
ouoB). Demosthenes was born in the late 380s. There is no positive evidence about the date of Cleobule’s birth, but it must presumably have been before 403/2, or Demosthenes’ citizenship would certainly have been attacked.* See also Isocr. VIII (Peace) 88. It is often said nowadays that Pericles procured a formal decree bestowing citizenship on his n
ouoB, Pericles, by his Milesian hetaira, Aspasia, after his legitimate sons, Xanthippus and Paralus, had died; but among the various sources which are relevant19 only Plutarch (Per. 37.5) mentions a decree regarding young Pericles, and significantly enough this relates to his registration ei’ B toyB wr
atoraB [‘into the phratry’]. Probably young Pericles was accepted as a citizen after the death of Xanthippus and Paralus, but a decree was still necessary to procure his acceptance by his father’s phratry, the admission of a n
ouoB to a phratry (or to some phratries) being probably already illegal, as it certainly was in the fourth century. The phratries may well have been less willing than the demes to relax their rules for admission in the late fifth century. It may not be amiss if I give here an extract from the article on Pericles in my edition of Lemprie`re. Pericles’ ‘vicious partiality for the celebrated courtesan Aspasia subjected him to the ridicule and the censure of his fellow-citizens; but if he triumphed over satire and malevolent remarks, the Athenians had occasion to execrate the memory of a man who by his example corrupted the purity and innocence of their morals, and who made licentiousness respectable, and the indulgence of every impure desire the qualification of the soldier as well as of the senator’!

E. The Law of Nicomenes, 403/2 B.C. 1. Schol. Aeschin. I 39 cites Eumelus, FGrH II A 77 F 2, for the text of Nicomenes’ law. The scholiast, and perhaps Eumelus, called it a decree (c
Zwisma), but it must surely have been a law (n
omoB). Its date is presumably 403/2. Except that it specifically ’ excepted toyB pr o Eykle
idoy [‘those before Eukleides’], i.e. those born before the year 403/2, the law was a repetition of that of Pericles: ‘those after the archonship of Eukleides’ had to have Plut., Per. 24; 38.5; Schol. Plat., Menex. 235e; Harp. and Suid. s.v. ’Aspas
ia; Suid., s.v. dZmopo
iZtoB; cf. Ael., VH VI 10, who speaks of Pericles’ n
ouoi as being prevented from gaining the citizenship by their father’s law. 19

The Athenian Citizenship Laws

241

’
mwv toyB gon
eaB asto
’ a yB [‘both their parents citizens’]. That ‘those born before Eukleides’ and ‘those after the archonship of Eukleides’ are those born before and after 403/2 respectively is proved by the references to this law in Isae. VIII (Ciron) 43; Dem. LVII (Eubul.) 30. Doubtless being born in the archonship of Eukleides counted as being born before it! 2. The law of Nicomenes has often20 been wrongly seen as an amendment of another law of the same date, that of Aristophon, mentioned by Carystius, ap. Athen. XIII 577b (who calls it a n
omoB). In fact the law of Aristophon has nothing whatever to do with the qualifications for citizenship. It provided that the man ‘ an ’ m ’ ast ’ ZB & g
enZtai) was to whose mother was not a citizen ( oB Z ej be a n
ouoB. But the man whose mother was a non-citizen was himself not a citizen, unless he (or his mother)* was born before 403/2, and in either event his being a n
ouoB made no difference to his claim to citizenship. The only result of this law was to make all mZtr
ojenoi into n
ouoi once more—as we have seen (C 3 above), the law of Aristophon must have been a re-enactment of an earlier law, which had perhaps been relaxed or merely disregarded in the last years of the fifth century. 3. Probably about the same time the ‘law of Solon’ denying ’ agxiste
ia [‘right of kinship’] to n
ouoi if there were legitimate sons (pa&ideB gn
Zsioi) (see A 2 above) was re-enacted,21 probably in a more stringent form.*

F. Two fourth-century Laws At some unknown date or dates before the delivery of Ps.-Dem. LIX (Neaer.) in the late 340s two further laws were passed which ought to be noticed here: 1. The first, cited in Ps.-Dem. LIX 16–17 (cf. 73, 124), forbade a citizen to synoike&in [‘co-habit’] with a non-citizen of the opposite sex. In either case the non-citizen was to be sold (as a slave), and the male citizen who synoikfi Z& [‘co-habits’] with a j
enZ [‘alien woman’] was to be fined 1,000 dr. Section 17 adds the words ’ e paidopoie&isuai [‘nor have children by her’]. oyd 20 [See e.g. A. R. W. Harrison, The Law of Athens, The Family and Property (1968), 26 n. 1.] 21 See Isae. VI [Philoct.] 47; Ps.-Dem. XLIII [Macart.] 51; cf. XLVI [Steph. 2] 18.

242

Athenian Democratic Origins

2. The second, in Ps.-Dem. LIX 52 (cf. 53, which uses the ’ technical term ZggyZk
enai [‘give in betrothal/marriage’]) threatened disfranchisement and confiscation of property against ‘ anyone who gave a j
enZ [‘alien woman’] as wife to an Athenian vB ‘ eayt fiv& pros
Zkoysan [‘as if she were related to himself’], i.e., presumably, representing himself as her k
yrioB [‘legal representative’].

G. The diacZw
iseiB [‘revisions of the citizen register’] under the decree of Demophilus, 346/5 B.C. The authorities are Harp., s.v. diac
ZwisiB, citing Androtion (FGrH III B 324 F 52) and Philochorus (ibid. 328 F 52); Liban., Hypoth. ad. Dem. LVII (Eubul.); Aeschin. I 77, with Schol. (naming Demophilus). Dem. LVII (Eubul.) was delivered for an appellant, Euxitheus of Halimous, who had been removed from the register of his deme under this decree and was appealing to a dicastery for reinstatement. Libanius speaks of the decree as a law (n
omoB) but I agree with Gomme22 and Jacoby23 that a decree (c
Zwisma) was surely sufficient. The decree ordered each deme to investigate the claims to citizenship of all its members. As a result, ‘many were rightly expelled from all the demes’.24 According to Libanius: (a) The investigation of those on the deme registers (lZjiarxik a grammate&ia) was directed towards ascertaining whether they were gn
Zsioi pol&itai [‘legitimate citizens’] or not. I do not believe the word gn
Zsioi was used in the decree: I think Libanius is glossing the text. The expression gn
ZsioB (as we have seen) is the opposite of n
ouoB, and although it could be used loosely, I do not think it would be so used in a decree ordering diacZw
iseiB [‘revisions of the citizen register’]. ’ asto ’ ’ ast ’ ZB & [‘from a male (b) All those not born ej &y kai ej citizen and a female citizen’] were to become metics, but it is perfectly clear from Dem. LVII 30 (and from Isae. VIII [Ciron]) that anyone born before 403/2 was not disqualified merely because his mother was a j
enZ [‘alien’]. 22 24

23 EGHL 68–9. Jac. i 161 & n. 33, on Androt. F 52. Dem. LVII 2, cf. 49.

The Athenian Citizenship Laws

243

(c) Those disfranchised could appeal to a dicastery for restoration, but if they were unsuccessful they were to be sold (as slaves). Some scholars accept this, and see a reflection of it in Dem. LVII 70; but I would reject it, with Gomme25 and Jacoby.26

H. The Qualification for Citizenship in the 320s Arist., Ath. Pol. 42.1, gives the qualification for citizenship: citi’ amwot
’ ’ vn zens are oi‘ ej ervn gegon
oteB ast & [‘those born of citizen parents on both sides’], and they are enrolled at the age of 18.27 In each case the demesmen decide by vote under oath, first if the ’ yuer
candidate appears to be of the legal age, and secondly ei’ ele
oB ’ esti kai g
egone kat a toyB n
omoyB [‘if he is a free man and was born according to the laws’]. [A long note was to have attacked the view that the applicant had also to be the offspring of a valid marriage.] I. The Purpose of the Citizenship Laws In considering the purpose of the citizenship laws, we must begin from four statements made by Aristotle. In Ath. Pol. 26.4, as we & have seen, he explains that Pericles’ law was passed di a t o plZuoB tvn & politvn & [‘due to the excessive number of citizens’]. In Pol. 1278a 26--34 he makes a series of statements which I would summarise as follows: (a) Some constitutions even admit certain foreigners to citizenship. For example: (b) In some democracies a man may be admitted whose mother only is a citizen. (I am sure Aristotle takes it for granted that any city which allows citizenship to be derived from a citizen mother alone will a fortiori allow it to be derived from a citizen father alone.) (c) In many states, similarly n
ouoi are admitted. (d) But such things are done because of a lack of legitimate ’
citizens (di’ endeian tvn & gnZs
ivn politvn): & it is because of ’ a shortage of men (di a g ar oliganurvp
ian, which must mean ‘shortage of citizens’) that they pass such laws;

25 27

26 EGHL 75–86. Jac. i 463–4. What we should call 17, according to R. Sealey, in CR 71 ¼ n.s. 7 (1957) 195–7.

244

Athenian Democratic Origins

(e) And when the number (of citizens) becomes larger ’ (eyporo &ynteB d Z o
’ xloy), they disqualify, by stages, first sons of a slave father or slave mother, then sons of citizen mothers only (toyB ’ o gynaikvn), ap & and finally they confine citizenship to those of citizen parents on both sides. In Pol. 1319b 6--12, again, Aristotle says that popular leaders anxious to make the demos strong are accustomed to accept the largest possible number of men as citizens, not only the legitimate (toyB gnZs
ioyB) but also bastards (toyB n
ouoyB) and those who have only one citizen parent, whether father or mother. Cf. Pol. 1275b 23--4, where Aristotle also mentions the possibility of deriving citizenship from either parent, and adds that on the other hand some states go further back, insisting for example on two or three or more citizen p
appoi [‘ancestors’]. The passage which follows, 1275b 25--30, giving the story about Gorgias’ ironical definition of Larisaeans as those who are made by larisopoio
i [‘makers of Larisaean pots/citizens’], shows the difficulty of avoiding an infinite regress in deciding who is a citizen. Presumably at Athens one did not usually go back behind a dead father, if he had actually been entered in his deme register, though Dem. LVII (Eubul.) 18–30 proves that it could happen. Possibly on the mother’s side one might often in doubtful cases be obliged to go back to her parents and show that they were citizens, although in practice it must often have been sufficient to show that she was a member of the phratry, which would create a presumption of citizenship, and that she was the recognised legal wife of the father. Now let us arrange the various stages contemplated by Aristotle in chronological order, beginning with the most liberal and ending with the most exclusive. ’ First we have a situation of oliganurvp
ia [‘shortage of citizens’]: here (A) Citizenship may be claimed if only one parent, father or mother, is a citizen, even if the other parent is a slave; and (B) It makes no difference if the parents are not legally married, so that the man concerned is a n
ouoB. A & B are partly overlapping, partly independent conditions: a man may be a n
ouoB if his parents are both citizens (but not legally married), and on the other hand he may have only one citizen parent and yet not be a n
ouoB, provided the law of his city recognises marriages between a citizen and a non-citizen—as it must at

The Athenian Citizenship Laws

245

this stage, because otherwise everyone who is not born of two citizen parents will automatically be a n
ouoB. In the second stage, when the citizen population has increased, the law still allows citizenship to be claimed if one parent is a citizen, but not if the other parent is a slave. In the third stage it is not enough to have a citizen mother, even if the father is free: citizenship can now be claimed through the father only. A slave mother, of course, is still a disqualification but a free non-citizen mother is not. The fourth and final stage is that in which both parents must be citizens. Athens reached the fourth stage with the law of Pericles in 451/0. As far as I know (see B 3 above), there is no evidence which of our other three stages Athens was in before 451/0, or during the later years of the fifth century, when Pericles’ law was not being strictly enforced. I imagine that during the latter period, at any rate, a citizen father was essential. I have one other observation to make before I give my views about the purpose of the citizenship laws. We must not concentrate entirely on the legal situation but must realise how likely it was that there would be attempts to evade the law and obtain the registration of children who were not legally qualified. Certain passages in the sources* suggest that such attempts may have been quite common and that the registration of unqualified persons may often have been procured by bribery,28 or by negligence, or by the natural tendency of the demesmen to accept the father’s29 assurance that the candidate was duly qualified.30 I can now discuss the purpose of the citizenship laws of Pericles and Nicomenes. I shall speak throughout for convenience of the law of Pericles. My suggestions are as follows: 1. Pericles’ law was not introduced ‘to preserve the purity of the citizen stock from contamination by undesirable alien elements’ (see p. 233 above). I can find no trace in any source of any idea that 28

See e.g. Dem. LVIII [Eubul.] 25, 59; Isae. XII [Euphil.] 2; Hyper IV [Euxen.] & citing Hyper. [fr. LIX 5 Sauppe] 3, col. 2, with Harp. and Suid., s.v. ’AgasiklZB, and Deinarch. [fr. XVI Sauppe] presumably also Anaxandrides fr. 4 K/A, ap. Athen. VI 263bc (Sunium). 29 Or mother’s: see Arist., Rhet. II 23, 1398a 33--b 3; Ps.-Dem. XXXIX [Boeot. 1] 25–36. 30 See esp. Isae. VI 18–26; Ps.-Dem. XXXIX–XL [Boeot. 1 & 2].

246

Athenian Democratic Origins

the racial ‘purity’ of the citizens of a particular city needed to be preserved, or might be ‘contaminated’ by alien elements. If Greeks did attribute special value to pure breeding, as some aristocrats did,31 such ideas were attached to individual families or gene¯, or conceivably to phratries, but not to the polis as a whole. Citizenship was a jealously guarded privilege in most Greek states, but any exclusiveness which might be shown in regard to the admission of new citizens was not due to a desire to preserve the purity of the ‘blood’ or the ‘racial stock’. It is true that we hear of an attempt by the Athenian aristocrats, on the expulsion of the Peisistratids (Arist., Ath. Pol. 13.5), to disfranchise some citizens on the ground that they were tfiv& g
enei m Z kauaro
i [‘impure in descent’]; but of course at that time, before the demes became the units through which citizenship was obtained, it must have been the phratries which kept the citizen registers (see B 2 above), and they might well feel, or at any rate the aristocrats who must have controlled most of them might conveniently pretend to feel, that the lineage was important. Demesmen made no claim to kinship, and I know of no scrap of evidence that they regarded non-Athenian stock as ‘impure’. Incidentally, citizenship was sold by some Greek cities, especially in the Hellenistic and Roman periods, when citizenship was just as much a privilege as ever. See also the protest of Jacoby32 against the notion that the introduction of alien blood would be likely to have a deleterious effect: he ends, ‘The village idiot frequently is a product of inbreeding.’ 2. The motive which is specifically attributed to the law of Pericles by the only source (Arist., Ath. Pol. 26.4) which supplies a motive is given in the words, ‘because of the large number of & citizens’ (di a t o plZuoB tvn & politvn): & we might legitimately interpret this as the prevention of a further increase in the number of citizens. Although this explanation has often been attacked in modern times,33 I believe it is right, and it fits in well with the general discussion in Aristotle’s Politics34 of the qualifications for citizenship imposed by Greek states. But we must be quite sure we have conceived the situation correctly. 3. Far too many scholars have assumed that either the intention or the effect of Pericles’ law was to prevent marriages between 31 33 34

32 See esp. Theogn. 183–92, 535–8 etc. Jac. ii 384 n. 59. e.g. by Busolt, GG III i 338 n. 3; Jac. ii 384 n. 61. Ar. Pol. 1278b 26–34; 1319b 6–12, quoted on p. 244 above.

The Athenian Citizenship Laws

247

Athenian citizens and foreign women. On this side I have three things to say: (a) Statistics are not available, but I cannot think that many Athenians had been marrying j
enai [‘alien women’]. We must not be misled by the handful of examples we have of wealthy Athenians taking foreign wives, doubtless because of the exceptionally large dowries they brought. A certain number of citizens of the lower classes may well have married metic women, although I know of no evidence of this. If, before the law of Pericles, citizenship could be obtained through the mother alone, then no doubt ’ male metics would have been very glad to marry Athenian asta
i [‘citizen women’], and thus put their children in a position to obtain Athenian citizenship; but again I do not know of any evidence. (b) The second point is much more important but is almost always entirely overlooked. It is true that Pericles’ law may have been very successful in discouraging marriages between Athenians and foreign women, for an Athenian who knew he could not transmit citizenship to his children by a foreign wife would no ’ Z [‘citizen doubt be most unlikely to marry anyone but an ast
woman’]. But merely to prevent Athenians from marrying foreign women would have no significant effect whatever upon the citizen birth-rate; the Athenian who would earlier have married a foreign woman would now simply marry an Athenian wife instead! As Gomme realised, only if there were ‘citizens who refused to marry at all because they could not marry foreigners’35 could Pericles’ law have produced any reduction at all in the citizen population—and the number of such citizens would surely have been very small indeed. (c) If before 451/0 Athens admitted to citizenship the children of ’ Z [‘foreign man by a citizen woman’], then the a j
enoB by an ast
effect of Pericles’ law, by discouraging such marriages, will have been somewhat greater. But it does not seem to me likely that such children were in fact entitled to citizenship, and I cannot believe that such unions were at all common—the only one I remember coming across is the marriage of Archedice, daughter of the tyrant Hippias, to Aeantides, son of Hippoclus, tyrant of Lampsacus (Thuc. VI 59.3), a dynastic marriage of a special character. 35

EGHL 87 n. 1.

248

Athenian Democratic Origins

Doubtless, as I have said above, metics living in Athens might have been particularly glad to have Athenian wives, if their children by them could become citizens; but we have no means of discovering how common such unions were. The law of Pericles, then, may or may not have been aimed at discouraging marriages between citizens and non-citizens. It would no doubt have had some discouraging effect, but this is not very important, as there is no reason to think such marriages were at all common. In so far as the law was aimed at checking any further increase in the number of citizens, however, any effect it may have had upon marriages between citizens and non-citizens will have been almost entirely irrelevant, unless citizenship could be claimed before 451/0 through the mother alone. It is also a fallacy to suppose that allowing Athenian citizens to marry aliens instead of Athenian women in the later years of the fifth century can really have been a way of making up for the losses of citizens during the Peloponnesian war, for there was certainly no shortage at that time of citizen women, relative to male citizens—in fact, the opposite is usually assumed, and is probably true. 4. The key to the situation is surely that it was not marriage with foreign women which this legislation was intended to discourage, so much as what we may perhaps call extra-marital unions, likely to result in issue additional to the children of lawful marriage. Now Athens was a slave society. In any such society, sexual relations between slaveowners and female slaves (not merely their own slaves) or freedwomen tend to be very common indeed. I submit that the great majority of ‘extra-marital unions’ at Athens would have been with slaves or freedwomen, and that the majority of mZtr
ojenoi would have been the issue of such union. If admitted to citizenship, they would of course have increased the number of Athenian citizens, because they would all (except for those fathered by Athenian bachelors—probably a rare class) have been additional to the children of lawful marriage. 5. Even if Athens before 451/0 was one of those cities mentioned by Aristotle (Pol. 1278a 32–3) in which the children of slaves were disqualified from receiving citizenship, it would be easy for a man who was pleased at the idea of having a son by a slave girl to free the girl before the child was born, or to free her immediately afterwards and declare that he had done so before the birth. But the absence of any special emphasis on the servile condition of Neaera

The Athenian Citizenship Laws

249

and Phaeno in Ps.-Dem. LIX (Neaer.), for instance, or of Alce in Isae. VI (Philoct.) 19–26, suggests to me that the children of slave women may not have been subject to any special disability.36 6. The most obvious objection to the theory I have put forward is that if the law of Pericles had really been directed against the issue of extra-marital unions, it could simply have excluded n
ouoi from citizenship. I believe there may have been very good reasons against casting the law in any such form. 7. This brings us to the Athenian law of marriage, a subject almost all aspects of which are obscure. My conclusions are as follows: (a) It might often be very difficult to establish the existence of a ’ yZ. See in particular Isae. proper ‘marriage’ (see A 1 above) by egg
’ yZ. III (Pyrr.) on whether Phile was married to Xenocles by egg
There were no technical words in Attic Greek corresponding to our ‘marriage’, ‘marry’, ‘husband’, ‘wife’. It is very significant that in one Demosthenic passage (Ps.-Dem. LIX [Neaer.] 122) which seems to be describing regular and permanent cohabitation the term used is synoike&in [‘cohabit’]. (b) In (a) above I have suggested that the fact of there having ’ yZ in many an individual case might be very been a marriage by egg
difficult to prove. I now come to a more radical proposition. I suggest (i) that ‘being legally married’ was a concept which had virtually no significance except in connection with the inheritance of property,37 and therefore (ii) that most Athenians belonging to the lower classes, who owned no valuable property the destination of which on their deaths might be important, would not have known or cared whether they were ‘legally married’ or not. I do not mean that most lower-class Athenians were not married by ’ yZ, just like their betters. I do not doubt they were: what I am egg
saying is that the gnZsi
otZB [‘legitimacy’] of the children of the ’ yZ, was so marriage, which was the one essential produce of the egg
unimportant except in the context of property ownership that among the lower classes there would be nothing to keep its memory alive once the marriage was in being. 8. We can now see why it would have been a mistake to attempt to exclude n
ouoi from the citizenship: it might have been very 36 37

And see G. R. Morrow, Plato’s Law of Slavery (1939). See e.g. Dem. XXXVI [Phorm.] 32.

250

Athenian Democratic Origins

hard, perhaps impossible, to say of many low-class Athenians whether they were n
ouoi or gn
Zsioi. And admission to the deme register, by which citizenship was conferred, took place only at the age of 18 (?17: see n. 27 above), whereas admission to the koin on grammate&ion [‘common register’] of the phratry, for which gnZsi
otZB [‘legitimacy’] was apparently a general and perhaps a universal requirement,* seems usually to have been effected soon ’ yZ would be comparatively recent and after birth, when the egg
most of the witnesses before whom it took place would still have been available. 9. The main aim of the law of Pericles, then, was to check the rapid increase in the citizen population, by making it impossible for citizens legally to introduce to their demes as citizens their sons by women other than their legal wives. It is possible that an additional motive may have been the desire that daughters of citizens should find citizen husbands: cf. Ps.-Dem. LIX [Neaer.] 112–13. 10. In fifth-century Athens, before 451/0 and again shortly before 403/2, the great majority of Athenian citizens must have married citizen wives, by whom they had legitimate children. But many, especially among the richer citizens, would also have had children in other ways, apart from temporary unions with common prostitutes, whose children they would not wish to recognise, or with citizen women not their wives, to whose offspring citizenship was never legally denied.* We often hear of rich Athenians having regular sexual relations with women other than their wives: (a) With slave girls of their own, who were sometimes kept as regular pallaka
i [‘concubines’], whether continuing as slaves or later receiving their freedom.38 Such unions might often produce children, and if the man fell in love with the girl he might even be induced to take over her children by an earlier lover, and try to have them made citizens, representing them as his own, as alleged (with what truth it is very hard to say) in Isae. VI 21–6; Ps.-Dem. LIX 13, 38, 118, cf. 119. 38

For some examples, see Antiph. I 14, cf. 20 (these two passages show the girl was certainly a slave); Athen. XIII 590cd, with Ps.-Plut., Mor. 849D (where Hypereides buys Phile as a slave and later frees her); Isae. VI (Philoct.) 19–21 (where Alce was no doubt purchased and freed by Euctemon); Ps-Dem LIX (Neaer.) 18–49, esp. 21–3 (Metaneira, a slave of whom Lysias (a metic) was so fond that he had her initiated at Eleusis), 31–5, with 45–8 (Neaera, purchased by Phrynion), 37–49 (Neaera is purchased and freed by Stephanus).

Afterword

251

(b) With hetairai who were either slaves belonging to others or else freed or free women (even sometimes Athenians: see Wyse, SI 318). Many prominent Athenians, including Isocrates (Athen. XIII 592d), Demosthenes (592e), Demades (591f), Aristophon, and Demetrius of Phalerum (577c), are said to have had children by such women, and it can even be asserted that they had them made citizens (591f, Demades)—illegally, of course. There may be no truth in most of these stories, but at least they tally with the information we derive from the orators39 and are likely to convey the truth about a considerable section of the Athenian upper class, if not the actual individuals named. Pericles himself had at least one child by Aspasia, a Milesian hetaira who became his pallak
Z [‘concubine’]: we might never have heard of young Pericles had not his father’s two legitimate children died. Since attempts to procure the grant of citizenship for children of such unions were apparently common even in the fourth century, although such grants were then contrary to law and presumably could be procured as a rule only by false oaths, it is easy to believe that they were very frequent indeed before the law of Pericles, when they were perfectly legal. 11. The law cited in Ps.-Dem. LIX (Neaer.) 16–17 (see F above), prohibiting Athenian citizens from ‘setting up house’ (ibid. 22) with foreigners, was a further attempt (perhaps many years after 403/2) to discourage Athenians from having children at all by foreigners. Such children were not merely n
ouoi: they could not legally become citizens in any event; but there was always the chance the parents might succeed in having them enrolled on the register of their deme, and it was thought better, therefore, to threaten penalties against such unions themselves.

AFTERWORD Much has been written on matters remotely or directly relevant, and this Afterword cannot be comprehensive. Humphreys, 93–4, briefly suggests that the Periclean law was directed against the upper classes and their penchant for marrying non-Athenian women. Ruschenbusch sees it as a 39

See above, n. 38: add Isae. IV [Nicostr.] 10.

252

Athenian Democratic Origins

defence against an influx of ‘aliens on the father’s side’, patroxenoi. Ste. Croix had not wholly discounted the possibility that patroxenoi could secure phratry membership at Athens before 451 (A 3; I 3 c), but he is surely right that such a category (almost invisible in our sources) cannot have been significant at Athens; see too the criticisms of Ogden, 69. Patterson argues that the Periclean law was the first state regulation of the criteria for citizenship, a matter hitherto left to individual demes or phratries to settle ‘on a traditional and familial basis’. Its main aim was not to change patterns of marital choice (though it introduced a new severity against metroxenoi) but to bring the whole area under effective control for the first time; the excesses of demes in the preceding twentyfive years in granting citizenship to many ill-qualified persons had made it necessary. The thesis that citizenship was unregulated before Pericles is not easy to accept; Craterus F 4 appears to offer counter-evidence, as does fr. 35a of Philochorus, which must now be moved back, on the basis of a neglected book ascription,a from book 4 to book 3 and so perhaps to the context of Solonian legislation. Two recent writers take up an old suggestion of Gomme (JHS 50 (1930) 106) that ‘the law of 451 was an attempt to restore what was regarded as normal by the many; it was in accordance with average sentiment’. Boegehold suggests that it was through judgements in the newly-instituted popular courts that the popular prejudice was revealed. ‘The Athenian people by their judicial decisions in particular cases showed that they were ready for a rather narrow definition of citizenship . . . Perikles did not anticipate some pressing need: he could instead have been confirming a manifest consensus.’ Ogden, in the thorough study which is now the point de repe`re on this theme, tries to track back a prejudice against metroxenoi and a tendency to deny their legitimacy to the time of the Peisistratids; Pericles’ law was an unrevolutionary expression of ‘autochthonous pride’. Osborne’s main concern is with the effect on iconography of the new premium set by the law on visible possession of a citizen mother. But he also emphasises that in the conditions of the empire ordinary Athenians had more opportunities and incentives to take foreign brides than ever before. Ve´rilhac and Vial, 50–60, offer no new interpretation. In contrast to all these writers, Walters suggests that Pericles’ law was intended to prevent Athenians from securing citizenship for their children by slave-women; the simpler expedient of denying citizenship to children of slaves would not have worked, because slaves can always be manumittedb (as is envisaged in just such a context in Menand. Epitrep. 538–40). This is in effect Ste. Croix’s theory, except that Ste. Croix a b

See C. Theodoridis, ZPE 138 (2002), 40–2, on Marcianus gr. 433 fol. 56v . Ogden’s critique, 69, appears to ignore this point.

Afterword

253

speaks more generally of ‘extra-marital unions’, not merely those with slave-women. The hypothesis has the great advantage that it allows Pericles’ measure to be indeed concerned, as Aristotle claims, with restricting the size of the citizen body. But it stands or falls with the following two inter-connected propositions: (1) bastards could be Athenian citizens, and (2) access to citizenship was controlled exclusively by demes, not also by phratries. (1) is necessary because children born of extra-marital unions were bastards; if all bastards were ipso facto excluded from citizenship, there was no need to exclude a particular class of bastards, those born of noncitizen mothers, by special legislation in 451. (2) is necessary because it is generally agreed that bastards were excluded from phratries; if phratry membership was a condition of citizenship then (1) is untrue. Both propositions are thoroughly controversialc—the first is rejected, for instance, in Ve´rilhac and Vial’s valuable study of Greek marriage, which covers many topics discussed by Ste. Croixd—, and the loss of Ste. Croix’s own statement of the case is much to be regretted. Boegehold, A. L., ‘Perikles’ Citizenship Law of 451/0 B.C.’, in Boegehold and A. C. Scafuro, eds., Athenian Identity and Civic Ideology (1994), 57–66 Humphreys, S. C., ‘The Nothoi of Kynosarges’, JHS 94 (1974) 88–93 Lambert, S. D., The Phratries of Attica (1993) Ogden, D., Greek Bastardy (1996) Osborne, R., ‘Law, the Democratic Citizen and the Representation of Women in Classical Athens’, Past and Present 155 (May 1997) 3–33 Patterson, C., Pericles’ Citizenship Law of 451–0 BC (1981) Ruschenbusch, E., Athenische Innenpolitik im 5. Jahrhundert v. Chr. (1979) Ve´rilhac, A. M., and Vial, C., Le Mariage grec (BCH suppl. 32, Athens and Paris, 1998) Walters, K. R., ‘Pericles’ Citizenship Law’, Classical Antiquity 2 (1983) 314–36 c On (1) see the many references in Ogden, 151, who adds to the arguments against it; on (2) see Lambert, ch. 1, who denies it (for the other view see his 26 n.6). d They write ‘le´gitimite´ et citoyennete´ sont indissociables’ (55).

7 The Athenaion Politeia and Early Athenian History

So important is Aristotle’s Athenaion Politeia for the political and constitutional history of Athens in the seventh, sixth and fifth centuries that anyone who deals with that subject must at least make clear what his attitude to the Ath. Pol. is; and this is what I now propose to do. Some historians today—a minority, I believe—have great confidence in the Ath. Pol., which they have no hesitation in regarding as the work of Aristotle himself, embodying the results of further research (including examination of state archives and other documents) conducted after the completion of the Politics, and therefore having some claim to be preferred to that work in case of disagreement. At the other extreme there are those who deeply distrust the Ath. Pol. (especially where it conflicts with the Politics), mainly on the ground that it cannot be based upon reliable sources; and some of these scholars even decline to speak of the Ath. Pol. as the work of Aristotle himself, attributing it rather to his school— such are the views, for example, of Hignett, whose History of the Athenian Constitution (1952) is now a ‘standard work’. Between the two extremes a variety of attitudes can exist. The publication in 1891 of the London papyrus of the Ath. Pol. immediately gave rise to a large number of books and articles in which the problems raised by the work were attacked from different points of view; but after the first three or four years most writers settled down to reproducing the conclusions of one or other of their predecessors,1 without saying anything new. Only within the last fifteen years have the foundations been laid for a 1 The works of the first phase are discussed by Busolt, GG II2 14–61, esp. 15–35; and at greater length by V. von Schoeffer in Bursian’s Jahresb. 75 (1893) 1–54; 83 (1895) 181–264. The most influential works have been those of Wilamowitz, AuA, and—unfortunately—Keil, SVAVA.

The Athenaion Politeia and Early Athenian History

255

fresh advance, by researches into the composition of the historical works of Aristotle, culminating in the admirable monograph by Raymond Weil, Aristote et l’histoire,2 and still more by Jacoby’s magnificent works on the Atthides.3 It is a great pity that Jacoby was obliged by the scope of his subject to avoid a systematic discussion of the Ath. Pol. and its problems, and that Weil did not feel inclined to deal with the sources of the Ath. Pol. and therefore, like most students of Aristotle, paid very little attention to the Atthidographers and made virtually no use of Jacoby’s great commentary upon their fragments, which refers over and over again to Aristotle’s Ath. Pol. and Politics.4 Recent works on the Ath. Pol.5 seem to me to have contributed little. The best commentary remains that of Sandys,6 the second edition of which appeared over fifty years ago. We still need a critical work on the Ath. Pol. which will take advantage of recent advances in knowledge, and refrain from the guesswork which has been the curse of so much recent writing in this field. An examination of the way in which ancient authors, especially lexicographers and scholiasts, made use of the various Atthidographers as well as of the Ath. Pol. might provide some useful clues. I can do little more here than indicate my own views on some of the chief problems. One of my main objects is to challenge certain assumptions about the composition and sources of the Ath. Pol. which are commonly made nowadays but which seem to me to lack foundation. I shall be very selective in citing modern works and make no pretence to give complete documentation. If I refer far more often to Jacoby than to any other modern scholar, it is because he towers above all others in this field. 2 Sub-titled Essai sur la ‘Politique’ (1960). This book contains elaborate and accurate documentation, and (pp. 421–9) a large (though not complete) bibliography. 3 A (1949); and Jac. i and ii (1954) (see the list of abbreviations, p. 441 below), the commentary on the fragments of the Atthidogaphers, published earlier in FGrH III B (1950) 323a–334. 4 See previous note. In his ‘Note pre´liminaire’ (p. 8), describing his own practice, Weil can even say, ‘Les fragments des historiens sont tire´s des F.H.G. de Mu¨ller (Didot) plutoˆt que des Fr. Gr. Hist. de Jacoby qui sont excellents mais difficiles a` utiliser’! 5 e.g. the latest English translation, with notes and well over 100 pp. of commentary, by von Fritz and Kapp (ACA); also Day and Chambers, AHAD. With the latter work I am entirely out of sympathy: see e.g. pp. 284, 298. 6 ACA2 . [But see Afterword, pp. 325–6.]

256

Athenian Democratic Origins

(I) THE ATHENAION POLITEIA AND THE POLITICS Aristotle can hardly have written, in addition to all his other known works, even those Politeiai which may with some probability be ascribed to him, some fifty to seventy or more, let alone the whole number of 158 with which he was credited—rightly or wrongly— in antiquity.7 To my mind, there is some force in the argument that Aristotle would have written the Athenian Politeia himself first, as a pattern to be followed in other cases—a task he could easily have carried out, since he spent a large part of his adult life at Athens, including most of his last thirteen years, 335/4 to 322. I find, however, that this argument does not seem a strong one to some people, who retort that this was the Politeia Aristotle could most easily commit to pupils. The fact is that we do not know beyond reasonable doubt (a) how far Aristotle himself conducted the research upon which the Ath. Pol. was based, or (b) whether he actually wrote the entire work himself, or entrusted it wholly or partly to a pupil, or (c), if the latter, how far he revised it himself. But since the Ath. Pol. was certainly produced in Aristotle’s lifetime and from his school, and no ancient writer who mentions it doubts the ascription to him, and since the work is thoroughly ‘Aristotelian’ in outlook and contains nothing Aristotle himself could not have written, we must accept it as his and speak of its author as Aristotle, and not invent a separate ‘writer of the Ath. Pol.’. In face of the unanimous ancient evidence of Aristotle’s authorship, we should require some positive evidence to the contrary before denying the work to him; and such evidence does not exist. Some of us may ‘feel’ (as I do myself) that the intellectual level of several chapters of the historical portion of the Ath. Pol. is lower than that of the great bulk of the Politics; but this can hardly be demonstrated. In any case, most of the sillier8 or more trivial9 parts of the Ath. Pol. consist of anecdotes related with more detail than is usually allowed to appear in parallel passages in the Politics. For Hignett,10 ‘the supreme difficulty’ in recognising the Ath. Pol. 7 For the Aristotelian Politeiai, see Weil, AH 97 ff., 255 ff., with the works there referred to; also Sandys, ACA2 xxiii–xxviii. 8 e.g. 22.7; 25.3–4; 26.1; 27.4–5; 28.4. 9 10 e.g. 14.4; 15.4; 18.2–6. HAC 27–30, at p. 29.

The Athenaion Politeia and the Politics

257

as the work of Aristotle is ‘the immeasurable superiority of the Politics . . . in breadth of treatment and soundness of judgement’. But, as Day and Chambers have recently remarked,11 ‘Breadth of treatment inheres in generalizations of the sort found in the Politics; its asserted absence in the AP could scarcely threaten the authenticity of such a historical treatise. And, as Hignett concedes, Aristotle’s soundness of judgment has not excluded factual errors from the Politics.’ Again, as Weil has said,12 ‘L’e´le`ve me´diocre et consciencieux a` qui C. Hignett attribue la Constitution n’aurait-il pas e´te´, d’abord, chercher des re´fe´rences et des renseignements dans la Politique, au lieu de la contredire sans jamais la citer?’ [‘Would not the mediocre and conscientious student to whom C. Hignett attributes the Ath. Pol. first of all have sought references and information from the Politics, instead of contradicting it without ever citing it?’]. For my part, I agree with Mathieu’s verdict, in his Introduction to the Bude´ text of the Ath. Pol.:13 ‘En effet il nous faut, pour toutes les oeuvres d’Aristote, ne pas oublier que le travail en commun a e´te´ la re`gle a` peu pre`s absolue dans l’e´cole pe´ripate´ticienne14 . . . Mais ce travail en commun suppose naturellement une surveillance et une revision attentive de la part d’Aristote; pour les oeuvres les plus importantes (et la Constitution d’Athe`nes est de celles-la`), sans doute les disciples n’ont-ils fait que le travail de recherches, et la coordination et la re´daction sont a` peu pre`s certainement le fait d’Aristote seul’ [‘Indeed for all Aristotle’s works, we must not forget that collaborative work was the almost universal rule in the Peripatetic School . . . But this collaborative work naturally presumes supervision and a careful revision on the part of Aristotle; and for the most important works (among which is the Ath. Pol.), no doubt the pupils only carried out the research, and the coordination and redaction were almost certainly the work of Aristotle alone’]. I may add that stylistic considerations15 are indecisive, especially because the Ath. Pol. is a work of a different kind from virtually all the others in the surviving Aristotelian 11

12 AHAD 3. AH 105. Aristote, Constitution d’Athe`nes4, ed. G. Mathieu & B. Haussoullier (1952), pp. I–II. 14 Mathieu refers here to the fundamental work of Usener, OWA. 15 Some of them are examined by Sandys, ACA2 lvii–lxiv; and see G. Kaibel, Stil u. Text der Pol. Ath. des Aristoteles (1893). 13

258

Athenian Democratic Origins

corpus. Sandys16 goes so far as to call it ‘the sole representative of the more popular class of writings attributed to Aristotle’. Simplicius17 makes the interesting observation that Aristotle’s ’ lack of clarity was not due to asuene
ia l
ogoy [‘linguistic inability’]: ‘ en ’ to&iB Mete
when he wanted to be very clear he could be, vB vroiB ’ &y Polite
iaiB [‘as in the kai to&iB Topiko&iB kai ta&iB gnZs
iaiB ayto Meteorologica and the Topics and his genuine Politeiai’]. *** We naturally want to determine the chronological relationship of the Ath. Pol. to the Politics. Here I think the positions adopted by many scholars are seriously at fault, and we must go very cautiously and take nothing for granted. It is true that if we are inclined to date a given passage before some particular time, a certain amount of the context is likely to be involved; but this need never be true of a whole book of the Politics or the whole historical section of the Ath. Pol. And on the other hand, even if here and there it does seem probable that some particular passage was written after a given time, we shall usually find that only a sentence or so is necessarily datable, and never much more; and if so, then again we must not go on to draw conclusions about the date of composition of the remaining portion of the work. We will first consider the evidence for the dating of the Politics, and then of the Ath. Pol. The question of the date of composition of the Politics is complicated by the fact that the whole work can easily be broken up into five or six separate sections, or methodoi,18 which some believe to have been written at widely different times. I do not think it is yet possible to produce a solution supported by arguments cogent enough to ensure general acceptance. At present many would adopt some form of Jaeger’s theory,19 the essence of which is that Books II–III and VII–VIII belong to what is often called Aristotle’s ‘middle period’, the years of his absence from Athens 16 ACA2 lx. Few will wish to follow Josef Zu¨rcher, Aristoteles’ Werk u. Geist (1952) 17, in supposing that we have ‘ganz echte, reine, unverfa¨lschte Aristotelica’ only in the Ath. Pol. (‘ein Jugendwerk des Aristoteles’) and the ‘exoteric fragments’, the rest of our Aristotelian corpus having been rewritten by Theophrastus. 17 In Categ., Prooem., p. 7. 12–18 (ed. C. Kalbfleisch, 1907). 18 Anyway Books I; II; III; IV–VI; VII–VIII. Some, like Ernest Barker (PA xxxvii ff.), would separate VI from IV–V. 19 Aristotle2 (1948) ch. x, xiii.

The Athenaion Politeia and the Politics

259

(348/7 to 335/4),20 whereas IV–VI (the ‘empirical books’) are late, and come from the last period (after 335/4), during which Aristotle was teaching at the Lyceum, Book I being the last of all. Weil, after conducting a most searching examination of the whole question of the composition of the Politics and the Politeiai, extending to more than 150 pages,21 has endorsed Jaeger’s scheme with minor modifications.22 However, Barker23 and Ross,24 who at first were inclined to agree in the main with Jaeger, later changed their minds, and each of them has declared his belief that the whole of the Politics was composed during the last thirteen years of Aristotle’s life. By speaking of the compilation rather than the composition of the Politics, I intend as far as possible to avoid facing the chronological problems of composition, which seem to me at present largely insoluble. A philosopher may ‘feel’ that his interpretation of Aristotle’s development necessitates a particular set of dates; and he may be right—or wrong. The historian has no warrant for going too far beyond the evidence, and in this case the evidence is inconclusive. By using the term ‘compilation’ I allow for the possibility (without actually assuming it) that parts of the work may have been written earlier, and merely incorporated, if with some revisions, in the Politics as we now have it. In any event, the compilation will have taken place during Aristotle’s last period, 335/4–322. The Politics cannot be proved to refer to anything later than the assassination of Philip II of Macedon in 336.25 Even the passage in question, which comes from Book V, could easily be an insertion into material written earlier; but that need not delay us, since few will wish to date Book V before c. 335.26 Many attempts have been made either to date certain passages (and even books) after 20

I do not wish to imply that he cannot have visited Athens during these years. AH 181–323, with 104–16, 160–2 etc. 22 See esp. AH 210, 254, 323, 337. 23 See his Politics of Aristotle (1906) xxxvii–xlvi, esp. xlii n. 1. 24 DAT 68–72, esp. 70–2; contrast the earlier Aristotle5 19 n. 1. 25 Pol. V 1311b 1–3. 26 At one time it could be thought that the Politics as a whole was written in or before 335, i.e. before Aristotle returned to Athens: see e.g. Keil, SVAVA 122 ff., esp. 124 (‘Die Politik ist nicht in Athen vollendet, sondern in Kleinasien und Makedonien wesentlich wol auf Grund der Materialien, welche Aristoteles bis zum Jahre 347 in Athen gesammelt hatte’). I doubt if anyone would maintain this today. 21

260

Athenian Democratic Origins

particular dates later than 336 because of supposed references to later historical situations or events, or on the other hand to provide termini ante quos for certain passages (and Books) because of Aristotle’s failure to mention later situations or events. I do not think that any single example of either of these procedures would be confidently accepted by the majority of historians. Some may find some of these speculations fairly plausible, as indeed I do myself, without wishing to base any firm conclusions upon them concerning the dates at which different parts of the Politics were written. A few years ago Ross brought forward some points already made by Newman and Barker, as evidence that five particular passages were written in the last years of Aristotle’s life; but my own feeling is that there is no substance whatever in any of these points, and I am confident that many if not most historians would agree.27 Even Weil, who is less prone to jump to unjustified conclusions than most, is in my opinion too sanguine in the theories he propounds about supposed chronological strata in the Politics:28 he sometimes indulges a desire for a tidy solution at the expense of probability. Everyone has his own pet theories, but no theory commands general assent. My conclusion is, then, that we cannot date any part of the Politics with certainty much before c. 335, although in fact some parts may be distinctly older (and probably are), and that no precise datings can be established for particular passages within the years 335–322. I shall have something to say about the relative dates of the Politics and the Ath. Pol. when I have examined the evidence for the date of the latter work, to which I now turn. In the form in which we now have it, the Ath. Pol. must come from between 328 and 322 at the extremes. The evidence is as follows: (a) Ath. Pol. 61.1 speaks of a single Strategos in charge of the symmories; but as late as 333/2 the Strategoi were still acting together as a board in this sphere.29 27 DAT, at p. 71. I notice that Weil (AH 181–210) was not impressed by any of the suggestions in question, which I deal with in Endnote i (p. 323 below). 28 I deal with three examples in Endnote ii (pp. 323–4 below). 29 IG II2 1623 A. b. 154–5. The earliest certain evidence we happen to possess for a single Strategos in charge of the symmories is from 325/4: IG II2 1629 a. 208–10.

The Athenaion Politeia and the Politics

261

(b) 61.1 also speaks of two Strategoi in charge of the Peiraeus, but in 333/2 there seems to have been only one Strategos acting in that capacity.30 (c) 46.1 mentions quadriremes, which first came into service in the Athenian fleet in the late 330s.31 (d) 54.7 mentions the archon-year 329/8 (Cephisophon). (e) In the year 322 (the year of Aristotle’s death) the Athenian democracy, which is described in the Ath. Pol. as a going concern, was destroyed by Antipater and an oligarchy installed,32 and Samos, which in Ath. Pol. 62.2 is still controlled by Athens, was taken away from her.33 Since Aristotle left Athens after the death of Alexander the Great in 323, never to return, the year 323 can presumably be regarded as a terminus ante quem. And we can probably reduce the span of five or six years within which the Ath. Pol. must have been finished, and date it in its present form to 326–5, or at any rate 326/5–323. Meritt has argued plausibly, on the basis of epigraphic evidence,34 that Ath. Pol. 54.4 was not written before 326/5. And if the work had been written, or subjected to any general revision, in or after 325/4, we might perhaps have expected 46.1 to contain some reference to quinqueremes, as well as triremes and quadriremes, for quinqueremes appear in the Athenian ‘navy lists’ for 325/4 onwards.35 (I have spoken of a ‘general revision’ deliberately. An author may easily make isolated corrections or additions to parts of a work of this nature, without altering statements in other parts which have become out of date but do not happen to catch his eye.) It has often been supposed, for example by Nissen, 30

See B. D. Meritt, in Hesp. 9 (1940) no. 8, col. II 9–10 (p. 62, cf. 65) [¼ O. W. Reinmuth, The Ephebic Inscriptions of the Fourth Century B.C. (1971), no. 9]. 31 See Endnote i ii (p. 324 below). 32 See Diod. XVIII 18.4–5; Plut,, Phoc. 27.5; 28.7. 33 See Diod. XVIII 18.9; Diog. Laert. X 1. 34 B. D. Meritt, in AJP 61 (1940) 78. This brief note evidently remained unknown to Weil. 35 See IG II2 1629 (325/4) d. 783 ff., at line 811. Wilam. (AuA I 211 n. 43), speaking of the inscription in question as being dated 325, wanted to have the quinqueremes built in 326 and to date the decree ordering them to 327. But the inscription refers to the situation at the end of 325/4, and there is no reason why the quinqueremes should not have been ordered and built in the same archon-year. I should add that I do not consider the argument from the absence of any reference to quinqueremes to be a strong one: see Busolt, GG II2 17 n. 1 (on pp. 17–18).

262

Athenian Democratic Origins

H. Weil and R. Weil,36 that the reference to the Ammonias in Ath. Pol. 61.7 must be later than 324; but this is entirely groundless,37 and no chronological inference can be based upon the passage. I would therefore date the completion, or final revision, of the Ath. Pol. to 326/5, or perhaps just afterwards; and it cannot well be later than 323, the year in which Alexander died and Aristotle left Athens for the last time. On the other hand, this certainly does not justify us in speaking of the Ath. Pol. as having been written in the 320s, as so many writers have done,38 or in assuming that all historical statements, or any given historical statement, in the Ath. Pol. will be later than corresponding statements in the Politics. The arguments are as follows: (a) The Ath. Pol. divides into two quite different parts, the historical (to chapter 41) and the systematic (chapters 42 ff.), and it is only in the second that we find any indications of the date of composition. Even the second part of the work, however, could have been mainly written several years earlier, as we shall see in a moment, and have received later additions and corrections. All the more is the date of composition of the first part an entirely open question.39 (b) There are only four passages in the Ath. Pol. which indicate a date in the 320s, and these do not even carry their contexts with them. (i) In 61.1 the references to the two specialised Strategoi40 could easily have been altered after their functions and numbers changed. (ii) In 46.1 the quadriremes do not appear in the first ’ ‘ boyl part of the sentence (epimele& itai de [sc. Z Z] kai tvn & 36 See Weil, AH 104–6, who is led to believe, quite unnecessarily, in a ‘contradiction’ in the evidence about the date of the Ath. Pol. Cf. G. Mathieu, who in his Introduction to the Bude´ text of the Ath. Pol. at first (p. iii) merely mentions the views of H. Weil and Nissen with a ‘peut-eˆtre’, but later (p. xiii) says that the Ath. Pol. ‘date de 329 au plus toˆt, plus probablement de 324 ou 323’. 37 See Jac. ii 235 nn. 8–9. The name of the trireme is explained by Lex. Patm. (K. ‘ ap
’ agoysa t Latte and H. Erbse, Lexica Graeca Minora, 160): Z aB uys
iaB ’
Ammvni Di
i [‘the ship which carried offerings to Zeus Ammon’]. Cf. Schol. Dem. Meid. (p. 637.6 Ddf.); Harpocr., s.v. ’Ammvn
iB. The naming of the trireme has nothing whatever to do with the deification of Alexander the Great. Ammon had a cult at Athens at least as early as the 360s (see IG II2 1428. 73–4, as restored on p. 808 of the Corpus, Vol. II–III, Pars ii, Fasc. 2; also the note on p. 809) and a temple apparently c. 334 (see SIG3 281 [¼ IG II2 378]. 14–15, & n. 3; with IG II2 1496 A. IV 95–6). And see Sandys, ACA2 245 n. 38 e.g. Jaeger (see below). 39 40 Cf. Weil, AH 115 n. 114. See pp. 260–1 above.

The Athenaion Politeia and the Politics

263

pepoiZm
envn tri
Zrvn kai tvn & skeyvn & kai tvn & nevso
ikvn [‘the boule administers the triremes which have been built and the equipment and the ship-sheds’]), but only in the second clause ’ tetr
‘ ’ o‘ (kai poie&itai kain aB [de] tri
ZreiB Z ZreiB, opot
eraB an & Zsfi Z, kai ske
yZ ta
y taiB kai nevso
ikoyB [‘and it dZmoB xeiroton
builds new triremes or quadriremes, whichever the people may vote for, and equipment and ship-sheds for them’]). It seems to me an irresistible inference, either that this sentence was revised to allow the mention of the quadriremes, or that it was first written as it stands, in the actual year in which the construction of quadriremes was first undertaken: only thus can we account for the contrast between ‘the triremes which have been built’ and ‘new triremes or quadriremes’. That is to say, the sentence in question was written or revised not later than 330 and probably a year or two earlier.41 This sentence has often been taken to provide a terminus post quem for the writing of the Ath. Pol., but in fact it either indicates a particular date in the late 330s, or (if it represents a revision of an earlier text) gives us a terminus ante quem at that time for the original version.42 (iii) In Ath. Pol. 54.7 Aristotle first lists the five penteteric festivals and then says [n &yn] de pr
oskeitai [kai ‘H]wa
is[tia] ep ’ i KZwisowvntoB ’
& arxontoB [‘and [now] the Hephaes[tia also] has been added, in the archonship of Kephisophon’]. These words too look to me like a later addition to a text written before the inception of the Hephaestia in 329/8, the more so because this is the only archon-date in the whole of the systematic part of the Ath. Pol. Had Aristotle been writing about these festivals after 329/8, would he have been likely to single out the ’ i toyB Hephaestia as he has done here?43 (iv) In 54.4 the words ep n
omoyB [‘in charge of the laws’] may very well have been simply ’ i t substituted for an earlier ep a cZw
ismata [‘in charge of the decrees’].

41

See p. 261 and n. 31 above. I do not think that Tovar and Weil (see Weil, AH 111–12) have proved that the present text is a revision, although my ‘feeling’ is that they are right in this, especially since the d
e of poie&itai de kain aB tri
ZreiB does seem to be a grammatical anomaly, and may perhaps be a sign of hasty revision—if it is not a mere copyist’s error. I agree with Weil that we cannot hope to reconstruct the original version. 43 [The new penteteric festival mentioned by Ath. Pol. was probably the Amphiaraia, not the Hephaestia (Parker, AR 149 n. 109); Ste. Croix’s point is unaffected.] 42

264

Athenian Democratic Origins

Clearly, the four passages we have just examined do nothing to establish the date at which the Ath. Pol. as a whole was written, and no other purely internal evidence is any more decisive. It may be true to say, with Jaeger,44 that the Ath. Pol. was ‘not published before 329/8’, but it is certainly going much too far to add, as he does,45 that it ‘was written between 329/8 and 327/6’, an opinion on which Jaeger proceeds to base the further conclusion that ‘the work on the other constitutions . . . cannot therefore have been done before Aristotle’s last years, if, indeed, it was completed at all during his life’.46 When Hignett47—without giving any reasons—declares his ‘firm belief that the Athenaion Politeia was completed at a date not long before that of Aristotle’s death and therefore later than the completion of the Politics’, he apparently means that the work was written not long before 323/2, for he goes on to characterise as ‘indecisive’ an ‘attempt to find a terminus ante quem for the composition of the Athenaion Politeia earlier than 323/322’. Mathieu,48 placing ‘la re´daction’ of the Ath. Pol. between 329/8 and 322, says, ‘Nous avons donc affaire a` l’un des derniers ouvrages compose´s par Aristote’ [‘So we are dealing with one of the last works composed by Aristotle’]. Many similar statements might be quoted.49 When we cease to concentrate entirely on the internal evidence of the Ath. Pol., we begin to see the whole question in a new light. I shall maintain that the ‘research’— in the broadest sense50—upon which the Ath. Pol. was based, and probably the original version of the work, can hardly be dated much after 335/4, when Aristotle returned to Athens and established himself as an independent teacher at the Lyceum, and indeed that the research may well have started earlier, even perhaps (although there is no clear evidence for this) before 348/7, while Aristotle was still a member of the Academy. There are several pieces of evidence which stand out. 44

45 Aristotle2 (1948) 327. Ibid. n. 1. Ibid. 327–8. And see Endnote i v (pp. 324–5 below). 47 48 HAC 390. Op. cit. (in n. 13 above) p. iii. 49 Thus Sandys (ACA2 xlix–l) speaks of the Ath. Pol. as being ‘not ready for public perusal until six or eight years’ after 334, and seems to conceive it as being written only then. See also Gilbert, CASA xxi–ii. 50 Perusing the works of the historians and Atthidographers, and such activities as questioning members of the leading political families (cf. pp. 304–5 below), would of course play at least as large a part as ‘documentary research’ in our sense. 46

The Athenaion Politeia and the Politics

265

The most decisive is the last page or two of the Nicomachean Ethics,51 affording an introduction to the Politics, the next work in the Aristotelian corpus. It furnishes conclusive proof from Aristotle’s own pen that at least some of his Politeiai were already in existence—as indeed we ought to have expected—before the Politics was compiled.52 Aristotle contemptuously rejects the claim of ‘the sophists’—and he is clearly referring to Isocrates XV (Antidosis) 79–83 in particular, as has long been realised53— that politics, and especially that branch of politics which is lawmaking (nomoues
ia), is on a par with and even inferior to rhetoric. Isocrates has said it would be a simple matter toyB n
omyB . . . toyB ’
lloiB eydokimo ’ par a to&iB a &yntaB . . . synagage&in [‘to collect the laws which are thought well of among other states’].54 Aristotle, repudiating this notion, echoes Isocrates’ very words in referring ‘ fiadion ei
nai t & to those who think it r
o nomouetZsai synagag
onti ’ toyB eydokimo &yntaB tvn & n
omvn [‘to be easy to legislate by collecting those laws which have a good reputation’].55 As well as intellectual ability and a theoretical training we require practical ’ experience, empeir
ia, if we are to understand political matters ‘ politik
(Z Z) and become capable of law-making (nomouetiko
i).56 On the analogy of medical writings, both general and specific, which are useful for the experienced doctor, Aristotle believes that collections of laws and constitutions (kai tvn & n
omvn kai tvn & politeivn & ai‘ synagvga
i) are likely to be useful to the trained student of law-making (nomoues
ia).57 He will begin, he says, by reviewing earlier work on the subject, and then conduct an examination of certain problems he proceeds to specify, on the basis of ’ tvn the constitutions he has collected: ei
ta ek & synZgm
envn & politeivn & uevrZsai ktl.58 We may compare the interesting passage in the Rhetoric59 in which Aristotle puts special emphasis on law-making (nomoues
ia) among the subjects the political orator (o‘ symboyle
yvn)60 needs to know. For this activity he recommends the orator to understand what form of constitution is expedient by ’ tvn considering what has happened in the past (ek & parelZlyu
otvn 51 52 53 54 56 58

EN X 9, 1180b 28 ff., esp. 1181b 12 to the end. For the meaning of ‘compiled’, see p. 259 above. See e.g. H. Bonitz, Index Aristotelicus (1870), s.v., ’Isokr
atZB. 55 Isocr. XV 83. EN X 9, 1181a15–17. 57 1181a 9–13, with 1180b 28–9. 1181b 2–12. 59 60 1181b 15–20. I 4, 1360a 18–37. 1359a 30 ff.

266

Athenian Democratic Origins

uevro &ynta), and to get to know the constitutions of other states, ’ ’
& gZB & per
iodoi61 . . . (ente &yuen g ar labe&in estin making use of ai‘ tZB ’ vn toyB tvn & eun & n
omoyB) [‘the descriptive geographies . . . (for from these one may learn the laws of the various tribes)’]. He also says that for political oratory it is useful to know the works of those who have written about historical events (ai‘ tvn & peri t aB pr
ajeiB graw
ontvn i‘ stor
iai).62 The Epicurean philosopher Philodemus (of the last century B.C.), in a passage in his Rhetoric,63 the importance of which in this connection was first brought out by Usener,64 spoke of Aristotle as to
yB te n
omoyB syn
agvn a
‘ ma tvi & mauZte&i aB tosa
y taB polite
iaB kai t a [Theophrastus, of course]65 kai t peri tvn & t
opvn dikai
vmata kai t a pr oB toyB kairo
yB [‘collecting the laws along with his student (Theophrastus) and the various constitutions and the territorial claims and the matters regarding crises’]. Philodemus’s references to the Politeiai and the Dikaio¯mata Poleo¯n66 of Aristotle, and the Nomoi and Politika ta pros kairous (‘Politics regarding Crises’) published after Aristotle’s death by Theophrastus, are explicit and unmistakable. As Herbert Bloch has said in an important article,67 after quoting this passage, ‘Hence Aristotle not only composed the Politeiai together with his pupils, especially with Theophrastus, but he also suggested and actively began work on the Nomoi and on the Politik a t a pr oB toyB kairo
yB, both finished and published after the master’s death by Theophrastus.’ On the basis of this evidence we are bound to conclude that what we should now call the research upon which the Aristotelian Politeiai were based, and the Politeiai themselves, had been substantially completed before the Politics was compiled. I agree, of 61 Arist. certainly consulted these himself: see Pol. II 1262a 19; Meteor. I 350a 16; II 362b 12. 62 See on the whole passage of which this forms part Weil, AH 91–3; Bloch, SHLFC 365–6 (who brings out the connexion with Pol. V 1309b 14 ff.). 63 Col. 53. 7–18 (ed. S. Sudhaus, II 57) [¼ W. W. Fortenbaugh et al. (eds.) Theophrastus of Eresus: Sources (1992), 594]. 64 OWA, at p. 98. 65 Cf. Polyb. XII 11.5, with 23.8, for the attribution of the Locrian Politeia to ‘Aristotle and Theophrastus’. 66 On which see Weil, AH 127–30. 67 SHLFC, at p. 361. Of this article, Parts II (‘Notes on the Atthis of Androtion’) and III (‘Theophrastus’ Nomoi and Aristotle’) are of special interest for the subject of this essay and others in this book.

The Athenaion Politeia and the Politics

267

course, with Weil’s recent statements68 to the effect that ‘il ne faut ’ tvn pas prendre au pied de la lettre le parfait de l’expression ek & synZgm
envn politeivn. & Il n’exclut certes pas remaniements et additions . . . Toutes les constitutions n’e´taient assure´ment pas encore termine´es’ [‘one must not take the perfect tense in the expression ‘‘on the basis of the constitutions I have collected’’ absolutely literally . . . Certainly not all the constitutions were yet complete’]. Even the Ath. Pol. may have received late additions— more particularly the Ath. Pol., indeed, since Aristotle was living at Athens and would be more likely to come across new historical material there, or become aware of constitutional changes immediately they took place. However, I cannot believe that anyone will accept Weil’s statement that the Politics is ‘contemporaine de plusieurs de ces travaux [the Politeiai], mais ante´rieure a` la plupart’ [contemporary with a number of these works, but earlier than the majority’],69 which seems to me to rest upon the flimsiest possible foundation—mainly the quite gratuitous assumption that a vague or general reference in the Politics to some historical event in a particular city must show that Aristotle’s pupils were in a position to obtain further information from some other source, and therefore (as a rule) that the relevant Politeia was complete, or nearly so,70 whereas if the reference in the Politics is fairly precise or detailed, then the Politeia of the city concerned could not yet have existed, or could not have been finished, because if it had been Aristotle would not have needed to give particulars.71 Weil even feels able in some cases to say that the Politeia in question must actually have been ‘in course of composition’.72 Such a method of classification is entirely subjective, not to say capricious. I should add that Weil does not sufficiently allow for the fact, 68

AH 79, 308. AH 309 (my italics). e.g. Chalcis (pp. 305–6), Cyme (285–6), Cyrene (286–7), Elis (276–7), Eretria (274); perhaps also the two Apollonias (267), Chios (306) and Cnidos (282), for none of which is there independent evidence of the existence of an Aristotelian Politeia. In several other cases Weil conceives the Politeia in question as either substantially complete or at any rate in course of composition: e.g. Argos (267–9, 298), Colophon (282–3), Megara (291–2), Sicyon (297–8) and Sybaris (298–9). 71 e.g. Aphytis (p. 272), Leucas (288), the Malians (290), Massalia (290–1) and Taras (303). Cf. also Aegina (264), Erythrae (274) and Rhegium (295). 72 e.g. Ambracia (pp. 265–6), Corinth (283–5), Delphi (273), Rhodes (295–6), Syracuse (299–303) and Thurii (280–1); perhaps also Epidamnus (261, 266) and Magnesia on the Maeander (289–90). 69 70

268

Athenian Democratic Origins

demonstrated by Bloch73 for Athens in particular, that Aristotle and his pupils must have collected a far greater body of material than was actually displayed in the various Politeiai, much of it being set forth only in the Nomoi (and the Politika ta pros kairous) ultimately published by Theophrastus. And if the bulk of the Politeiai, or indeed any of them, were substantially completed before the Politics was compiled, then surely the Ath. Pol. (apart from late additions) must be appreciably earlier in date than the Politics. It is true that there is no explicit evidence for the priority in time of the Ath. Pol. over the other Politeiai; but the constitution of Athens was (with that of Sparta) the most famous of all Greek constitutions, and after 335/4, at any rate, it would necessarily have attracted Aristotle’s attention in preference to any other. I would therefore assign a date as near as possible to 334 for the main investigations on which the Ath. Pol. was based, and for the writing of that work, which would surely follow closely upon the research. Revisions were made later, almost certainly down to 326/5. Is there any reason to go still further back, to the years before Aristotle finally established himself at the Lyceum? Jacoby—to judge by one of those obiter dicta buried in the notes at the end of his Atthis,74 which only too easily escape notice—evidently believed that Aristotle’s collection of constitutions, or at any rate of laws, was at least partly compiled before Plato wrote the Laws (presumably in the late 350s), for he remarks, ‘It is absolutely necessary to treat the question whether Plato in his N
omoi made an early use of the material collected under the supervision of Aristotle. Personally I have no doubt that he did.’ Unfortunately he then shies away from this problem with the excuse that ‘these questions lead to the problem of the origin of the Politics, upon which I must not touch here’. I do not think we can be at all certain that Aristotle had made any systematic attempt to collect constitutions before 335/4, but it seems to me not at all unlikely that he had already done, or had encouraged others to do, some of the research on the basis of which the Politeiai were written; and we certainly cannot exclude even the period of his first sojourn at Athens, when he spent nearly twenty years, 367/6 to 348/7, study-

73

SHLFC 367–76.

74

A 386 n. 51.

The Athenaion Politeia and the Politics

269

ing under Plato in the Academy. The main considerations are as follows: 1. Weil’s very thorough and careful study has shown that Aristotle was deeply interested in i‘ stor
ia, in its original sense of mainly empirical research in the broadest sense, from quite an early state in his development.75 The well-known theory of Jaeger, that Aristotle began by being a devoted Platonist but drew steadily further away from Platonism, and indeed from pure philosophy in general, towards natural science, and became increasingly interested in empirical rather than theoretical investigations, cannot be accepted in anything like the extreme form in which Jaeger stated it. I agree entirely with Ross76 that while ‘Aristotle moved from a Platonic, other-worldly view for which the physical world was of little interest, and super-sensual reality was all that mattered, to one for which the problems of the physical world mattered a great deal, the movement of his mind proceeded neither so far nor so fast as Jaeger describes.’ I would add that there is some evidence of a most un-Platonic interest in empirical research from quite an early period in Aristotle’s life. Ever since D’Arcy Thompson’s brief Prefatory Note to the ‘Oxford Translation’ of Aristotle’s Historia Animalium, published in 1910, it has been known that Aristotle must have been pursuing detailed researches in natural history during his stay at Mytilene c. 345–3, some eight to ten years before he finally settled at Athens.77 It is one of the many merits of Weil’s recent book78 that it brings out the interdependent character of Aristotle’s researches. ‘La documentation ‘‘historique’’ d’Aristote n’est en re´alite´ qu’un aspect, qu’une partie, d’une documentation 75 Weil, AH, esp. 87 ff. I cannot resist quoting here an unintentionally comical remark by Zu¨rcher, op. cit. (in n. 16 above) 258. Aristotle, he thinks, must have conceived the idea of collecting constitutions at an early stage in his development— ‘es passt ja sogar besser in die vorphilosophische Periode eines Menschen’! In view of this remark, it is perhaps not surprising that Zu¨rcher should cite Ath. Pol. 54 as providing a date ‘323/22’. (He can only be referring to 54.7, which in fact mentions the archon-year 329/8.) Perhaps Zu¨rcher’s own ‘pre-philosophical period’ could have been prolonged with advantage. [On Aristotle’s evaluation of history see Ste. Croix’s ‘Aristotle on History and Poetry (Poetics 9, 1451a 36--b 11)’ in B. Levick (ed.) The Ancient Historian and his Materials (Essays in honour of C. E. Stevens) (1975), 45–58, repr. in A. Oxsenberg Rorty (ed.), Essays on Aristotle’s Poetics (1992), 23–32.] 76 DAT 75. 77 See also von Fritz and Kapp, ACA 32 ff. (esp. 40–1, 52–3), and the works cited by Weil, AH 81 n. 183. 78 AH 311 ff., cf. 87 ff., 102–4.

270

Athenian Democratic Origins

plus vaste’ [‘Aristotle’s ‘‘historical’’ documentation is really only an aspect, a part, of a larger documentation’].79 Herodotus, for example, is cited not only in the Ath. Pol. but in the Poetics, the Rhetoric, the Eudemian Ethics, the Historia Animalium and the De Generatione Animalium,80 and is also referred to very often without mention of his name.81 Aristotle surely got to know his work well at an early stage. Much more of Aristotle’s historical ‘documentation’ must also have been assembled at a relatively early date, probably while he was still a member of the Academy82—the material embodied in the Nomima (barbarika) for instance, a work which is undatable but perhaps not likely to come from the last period of Aristotle’s life.83 2. Aristotle’s participation in research which had necessarily to be conducted to a large extent among documents and archives is particularly significant; and again there is no reason to think it all belongs to the very last years of his life. His work on the list of Pythian victors, which he compiled in collaboration with his nephew (or cousin) Callisthenes,84 must at least have begun before 334, when Callisthenes went off to Asia in the company of Alexander the Great, and was very probably carried out before Aristotle returned to Athens.85 The parallel list of Olympic victors cannot be dated, but Weil86 has shown that it may conceivably have been undertaken before the Pythian list. The Didaskaliai and the Nikai Dionysiakai87 could equally well have been compiled before 347 or after 335. All these works involved documentary research, as remote from ‘philosophy’ in the modern sense as any kind of historical activity can well be.

79

80 Weil, AH 311. The references are given by Weil, AH 313 n. 20. See Weil, AH 312–16. 82 Weil, AH 211–54, advances some interesting suggestions based on Aristotle’s treatment of Carthage and other ‘barbarian’ states, also Sparta and Crete. 83 See Weil, AH 116–21. 84 See Weil, AH 133–7. The Delphian inscription mentioning the work of Aristotle and Callisthenes is Tod II 187 ¼ SIG3 275 ¼ FD III i 400. The decree in question will have been voted before the Delphians heard of the death (or the disgrace?) of Callisthenes, i.e. hardly later than 326. The actual publication on stone of the Pythionikai (see SIG3 252 ¼ FD III v 58, lines 42–3) must now be dated to late 327: see D. M. Lewis in CR 72 ¼ n.s. 8 (1958) 108. 85 This conclusion is unaffected by the recent lowering of the date of the actual inscription which survives (see previous note). 86 87 AH 131–3. See Weil, AH 137–9. 81

The Athenaion Politeia and the Politics

271

3. Although I admit there is no clear proof that Aristotle was working on his collection of laws and constitutions well before c. 335, I think that Isocrates XV (Antidosis) 83, especially in view of Aristotle’s reply at the end of the Nicomachean Ethics,88 contains a very probable indication that he had in fact begun the collection as early as the late 350s.89 As I have already mentioned,90 Aristotle’s & rebuke to the sophists who suppose it rfi‘ a
dion ei
nai t o nomouetZsai ’ synagag
onti toyB eydokimo &yntaB tvn & n
omvn [‘to be easy to legislate by collecting those laws which have a good reputation’] is a clear allusion to Isocrates’ allegation in the Antidosis that it would ’
’ be easy toyB n
omoyB . . . toyB par a to&iB alloiB eydokimo &yntaB . . . synagage&in [‘to collect the laws which are thought well of among other states’].91 Now we never hear that anyone had ever considered making a collection of laws or constitutions before Aristotle, and no one will believe that the undertaking was first suggested to him by Isocrates’ words. Why should Isocrates contemplate such a hitherto unknown activity at all, unless he knew that Aristotle had actually begun to work on it, or at least was proposing to do so? And why should Aristotle go out of his way to make an unmistakable reply to Isocrates’ words, unless they would generally be recognised as a reference to his own work?92 Aristotle often distinguishes between polite
iai and n
omoi, but at other times he can use the word n
omoi in the broadest sense, to include the whole framework of the constitution as well as the

88

See p. 265 above. The date of the Antidosis must be shortly after the peace that ended the Social War in 355. The most probable date, to my mind, is 353–2. The passages I am referring to were taken by Bloch (SHLFC 362–3) to show that ‘at that time [surely the date of the Antidosis] both collections were already being commenced by Aristotle and his pupils’. Von Fritz and Kapp (ACA 45 n. 38) speak of Aristotle’s ‘collections of constitutions’ as ‘an undertaking which, of course, could not be known to Isocrates’. Why not? Doubtless the collections themselves could not, but the fact that they were being made might well be, and that is all that we need. 90 P. 265 above. 91 As H. Bonitz said (Index Aristotelicus, s.v., Isokr
atZB), ‘Ex orationibus Isocratis (ac praecipue quidem vel unice ex sex orationibus: [inter alias] 15 ’Antid.) saepe Ar[istoteles] vel ipsa verba affert vel omnino sententias comparat.’ 92 I must say, I cannot see why Keil (SVAVA 124) should want to date the passage in the EN soon after the composition of the Antid., which doubtless achieved a wide circulation for many years together and might still be thought worth replying to twenty or more years later. 89

272

Athenian Democratic Origins

regulations intended to operate within it,93 and I find it impossible to believe that he can ever have intended to collect only or even mainly n
omoi in the narrower sense, ignoring those which determined the ‘constitution’ proper. Surely his aim from the first will have been to inform himself about the structure of many city states and observe the results upon life and behaviour in each case. And if Aristotle did begin this ultimately very fruitful activity during his first stay in Athens, before Plato’s death, he will surely have worked from the first on the laws and constitution of Athens. If so, the original composition of the Ath. Pol. may have taken place at any time or times during not merely the last thirteen but the last thirty years or so of Aristotle’s life, though of course it did not reach the form in which we have it until shortly before his death— at least as far as the later (the ‘systematic’) section is concerned: revisions of this do not absolutely guarantee that the earlier (the historical) section will have been similarly revised, although no doubt it is likely to have been.94 Since the composition of both Ath. Pol. and Politics may have extended over a long period, it is impossible to provide a relative chronology for the two works, although there is every probability that an original version of the Ath. Pol. (how different from the present one, we do not know) will have existed from well before the compilation of the Politics. All we can say for certain of the Ath. Pol. is that portions at any rate of the second part of it (the systematic part, chapters 42 ff.) were revised down to the 320s. As we have seen, however, it is wrong to conceive the Ath. Pol. as having necessarily been written in the 320s. Partly, perhaps, because they proceed on this erroneous assumption, there are those who like to think that the Ath. Pol. represents the fruits of research, partly documentary, carried on after the completion of the Politics, and that where the two works conflict the Ath. Pol. is therefore likely to be the more reliable. No such argument has the slightest force: in every case of conflict we must look at all the evidence and not allow ourselves to be influenced by any imagined chronological relationship between different statements in the Politics and the Ath. Pol. In any event, it would not help much even if we could determine which of two contradictory

93

See pp. 310 ff. below.

94

Cf. p. 258 above.

The Athenaion Politeia and the Politics

273

statements was the later, because there is some reason to think that Aristotle’s ‘second thoughts’ sometimes effected the reverse of an improvement in his historical judgements. *** I would recognise only three major examples of conflict between the Ath. Pol. and the Politics:95 (A) the ‘constitution of Draco’ in Ath. Pol. 4, compared with Pol. II 1274b 15–16, where Draco is explicitly said to have provided n
omoi for an existing polite
ia (B) the description of the ‘revolution of the Four Hundred’ in 412/11, in Ath. Pol. 29–33, compared with the very brief but entirely different characterisation in Pol. V 1304b 10–15; and (C) the contradiction between Ath. Pol. 8.1–2 and Pol. II 1273b 39--74a 2, 1274a 15–17, regarding the method of election of magistrates in Solon’s constitution. I shall begin by taking the first two of these examples together: in both cases no one will doubt that the Politics is to be preferred to the Ath. Pol. (A) The ‘constitution of Draco’ is to us a patent forgery, an invention of the late fifth century or (more probably) the fourth.96 It is almost universally accepted nowadays, with the consequential additions which it entailed,97 as a late addition to the Ath. Pol. by Aristotle himself. It may, of course, have been added after his death by a member of his school (Demetrius of Phaleron, for instance). I myself would very much like to accept this hypothesis; but there is no evidence for it, and if we did adopt it we should then have to admit that the interpolator may have been at work elsewhere as well. (B) It has not been sufficiently noticed that the only statement Aristotle makes in the Politics about the nature of the revolution of 95

I cannot take seriously such minor matters as (a) the conflict between Pol. V 1315b 29–34 and Ath. Pol. 17.1; 19.6, on the chronology of the Peisistratids, on which see Jacoby, A 188–96, and esp. 373–4 n. 107 (for a bibliography, see 376 n. 122); (b) the mention of Themistocles as having assisted Ephialtes in the reform of the Areopagus, in Ath. Pol. 25.3–4, compared with Pol. II 1274a 7–8 (on which see Keil, SVAVA 120 n. 1), or (c) the fact that Charicles and Phrynichus are mentioned in Pol. V 1305b 25–8 but not in the Ath. Pol. I regard all these divergences as trivial. 96 See esp. Fuks, AC 84–101, where full references are given to earlier work. Add Weil, AH 106 ff. 97 & pr At any rate (a) the words tZB o Dr
akontoB [‘the one before Drako’] in 3.1; fi ZrZto Z kai pr
oteron [‘as they had been divided before’] in (b) the words kau
aper di
7.3; and (c) the sentence met a de ta
y tZn . . . pr vton & [‘and after this . . . first’] in 41.2. Although unable to produce any arguments that would satisfy a sceptic I have often wondered whether these insertions, with Ch. 4, may not be peculiar to one particular family of MSS, represented for us by the London papyrus.

274

Athenian Democratic Origins

the Four Hundred98 is very much at variance with his own description of that revolution in Ath. Pol. 29–33, but on the other hand agrees very well with the picture of that revolution given by Thucydides, which is undoubtedly much nearer to the truth. In the Politics, doubtless under the influence of Thucydides, Aristotle gives the revolution in question as a typical example of a revolution ’ atZ) and maintained by force (metabol
Z) procured by trickery (ap
(b
ia); but neither of these qualities is displayed by the Four Hundred in the Ath. Pol.99 In Ath. Pol. 33.1 the naval disaster off Eretria and the revolt of most of Euboea is represented as the sole reason for the fall of the Four Hundred, and there is no trace of the deep general discontent with their regime which we find in Thucydides, nor is there any mention of their treacherous dealings with Sparta.100 Indeed, the only hint of criticism appears in 33.2: the Four Hundred acted entirely on their own authority and did not refer anything to the Five Thousand. Yet there are distinct traces of Thucydides’ account visible in Aristotle’s very wording.101 Moreover, as I have shown elsewhere,102 the two constitutions in Ath. Pol. 30–1103 can never have been in actual operation, whatever the reality which lies behind them. They must surely come either from the Apology of Antiphon or the Atthis of 98 Pol. V 1304b 12–15. The reference to the Four Hundred in 1305b 27 is merely incidental; but note that again the Ath. Pol. has nothing to say about disagreements within the Four Hundred until we reach the movement engineered by Aristocrates and Theramenes (33.2). 99 Note esp. 29.1; 32.2–3; 33.1. 100 Note here the second half of Ath. Pol. 32.3. 101 Not so much the correspondence between Thuc. VIII 97.1 and Ath. Pol.33.2, which has often been quoted but might easily have come through an intermediary source, as the reference to Euboea in Ath. Pol. 33.1, where the verbal similarity to Thuc. VIII 96.2 is so close as to suggest that Arist. used Thuc. directly—as of course we should expect. 102 CFT. 103 I would accept Ath. Pol. 29.2–3, on the other hand, as a probably accurate report of the meeting of the Assembly mentioned more briefly by Thuc. VIII 67.1, and Ath. Pol. 29.4–5 as a substantially true account, so far as it goes, of the Assembly at Colonus: Thuc. VIII 67.2–69.1 gives a very different and much more realistic account of the same proceedings—as Jacoby remarks, in an excellent note (Jac. ii 137–8 n. 2), ‘he saw through the whole farce’. The essential truth of the statement, which is not to be found in Thuc., that supreme power was now voted, for the duration of the war, to ‘those Athenians most able to serve the state with their persons and their property, not being less than 5,000’, is guaranteed (even apart from Ps.-Lys. XX 13, cf. 16) by Thuc.’s statement that the Four Hundred were to summon the Five Thousand when they pleased—why would the Five Thousand be convened, if not to act as the Assembly to the Council of 400?

The Athenaion Politeia and the Politics

275

Androtion, or from both these works: each of these men would have good reason to misrepresent the revolution of the Four Hundred and pretend it was a much less fraudulent, violent and unconstitutional affair than Thucydides shows it to have been.104 In our first two examples, then, we find good reason to prefer the Politics to the Ath. Pol. I think there can be no doubt that the ‘constitution of Draco’ is a late addition to the Ath. Pol., and that it was inserted after the compilation of the Politics. If this indeed represents a change of mind by Aristotle himself, did he similarly change his mind for the worse about the Four Hundred? I see no way of deciding for certain whether the critical reference to the Four Hundred in the Politics was written before or after the present chapters 29–33 of the Ath. Pol. I should like to draw attention to the fact that the whole passage, Pol. 1304b 7–18, may just as well have been an earlier note incorporated at the time of compilation as a later addition. Lines 5–7 have a concluding formula, kau
oloy men ’
peri p
oyn asaB polite
iaB ai‘ arxa i kai ai’ t
iai tvn & st
asevn kai ’
tvn & metabolvn & to &yton exoysi t on tr
opon [‘so in general, the origins and causes of civil wars and revolutions in all constitutions take this form’], which at one stage was surely designed to round off the chapter; but our passage then follows, with another con
which Aristotle cluding formula (also introduced with the men oyn ‘ vB
peri p
asaB so often uses in such cases): apl & men oyn ’ tvn t aB polite
iaB ek & ei’ rZm
envn symb
ebZke g
inesuai t aB metabol
aB [‘so in a word, in all constitutions revolutions arise from the causes mentioned above’]. I would suggest that lines 7–18 may have been a note written before Aristotle read Androtion 104 Antiphon’s line, to judge by the surviving fragments, was to pretend that he was not an oligarch. It would then be necessary for him to demonstrate this by asserting that he and his colleagues were about to bring a more broad-based constitution into existence. Hence, perhaps, the disorganised state of Ath. Pol. 30–1, which may have been cobbled together from a series of extracts in Antiphon. That Arist. knew this famous speech of Antiphon, which Thuc. (VIII 68.2) so much admired, is shown by EE III 1232b 7. We do not know whether Androtion dealt with these events in detail (cf. FGrH 324 F 43), but if he did, he would presumably have wanted to defend the memory of his father Andron (a ‘Theramenist’ member of the Four Hundred), by trying to show that the Four Hundred were not all extreme oligarchs and that they had taken steps to introduce proper constitutional machinery. I suppose that Androt. is likely to be an immediate source of Arist. here, even if they both used Antiphon. I have made no reference here to M. O. B. Caspari [Cary], in JHS 72 (1952) 56–61, or Mabel Lang, in AJP 69 (1948) 272–89, because I think they have entirely failed to understand the nature of the source problem behind Ath. Pol. 29–33.

276

Athenian Democratic Origins

(or Antiphon) and not altered when it was inserted at the end of Pol. V 4, even if by that time Ath. Pol. 29–33 had been written. Alternatively, and (as it seems to me) more probably, Ath. Pol. 29– 33, in its present form, is later than the compilation of the Politics: it may well replace an earlier version based on Thucydides. In either event, Aristotle changed his opinion for the worse. There is one other possibility: that Aristotle wrote Ath. Pol. 29–33 first, at quite an early date, and then, under the influence of Thucydides, realised the truth and wrote the Politics passage. I find this harder to believe, however, because we have evidence of more than one alteration in the Ath. Pol. (if only in the systematic portion) as late as the 320s, but no clear evidence of any such late rewriting of the Politics; and if Aristotle had changed his mind about the Four Hundred early enough to incorporate his verdict in Pol. V, why did he leave the Ath. Pol. uncorrected? Moreover, he must have known Thucydides well from his youth up. (C) I have already shown105 that the statement in the Ath. Pol. ’ prokr
itvn [‘sortition from that Solon introduced kl
ZrvsiB ek pre-selected candidates’] for the election of magistrates, including archons, is exceedingly unlikely to be right, although it is conceivable that some magistrates, the treasurers in particular, may have been chosen in this way, or even by pure sortition (kl
ZrvsiB). Again, the Ath. Pol. proves inferior to the Politics when the two works conflict; and again we cannot be certain which statement is later, although presumably it is the one in the Ath. Pol. It is quite impossible to tell whether Aristotle arrived at his erroneous view, as many scholars have supposed, by false inference from the survivals of his own day which he describes in Ath. Pol. 8.1, or whether he was deceived by some literary source or a forged version of an alleged Solonian provision, concocted perhaps in that propagandist spirit which in the fourth century loved to credit Solon with the invention of the institutions of the developed democracy. I do not claim to be able to account satisfactorily for the conflicts between Politics and Ath. Pol. The most likely explanation, to my mind, is that in his later years Aristotle became rather ‘a sucker for documents’, accepting too easily what looked to him like documentary material. The peculiarity shared by all three of the statements 105

Pp. 92–103 above.

The Sources of the Athenaion Politeia

277

in the Ath. Pol. which most notably contradict the Politics is that they either come directly from documents or go back ultimately to a source of a documentary character, whether genuine or not. Having noticed these contradictions between our two works, we may now glance at a few cases in which they support each other: (1) Pol. VI 1319b 19–27, with Ath. Pol. 21.2–4 (note especially ’ ’ anameixu vsi= & aname& ijai), and Pol. III 1275b 35–7, with Ath. Pol. 21.4 (line 28 OCT); (2) Pol. II 1304a 20–1, with Ath. Pol. 23.1–2; (3) Pol. II 1274a 18–21, with Ath. Pol. 7.3–4 (is it fanciful to detect in the Politics passage a certain uneasiness on the subject of the ‘hippad’ class?); and perhaps (4) Pol. V 1315b 3–4, with Ath. Pol. 16.9. In all the passages quoted from the Politics we seem to find topics which appear in the Ath. Pol. already present in Aristotle’s mind. I do not doubt that other parallels could be quoted: I have simply listed a few of the more obvious examples.

(II) THE SOURCES OF THE ATHENAION POLITEIA I turn now to the question of the sources of Aristotle’s account of Athenian constitutional history, in the Politics and more particularly the Ath. Pol. Although individual problems in this field have been discussed over and over again in recent years, it seems to me that there is still something to be said in general terms, more particularly because too many writers during the last generation or two have left most of their premises unexamined, and it is above all our basic assumptions that I wish to analyse. I make no pretence to be giving a complete account: I shall concentrate on those points which seem to me most important. I am tempted to take as my text what Wilamowitz said about his treatment of the second part of the Ath. Pol.:1 ‘Ich kann keine sonderung seiner vorlagen vornehmen, und ich hu¨te mich wol mit dem namen Androtion oder anderen zu spielen’ [‘I can make no separation of its construction, and I am wary of playing with the name of Androtion or others’]. Quellenkritik was eagerly applied to the Ath. Pol. by several leading scholars immediately after the publication of the London papyrus in 1891. By the time the second volume of Busolt’s influential Griechische Geschichte appeared in its second edition,2 four 1

AuA I 257.

2

See Busolt, GG II2 32–54; cf. GS I 92–6.

278

Athenian Democratic Origins

years later, opinion had already crystallised in the form which remains standard today: that for the historical section of the Ath. Pol. (ch. 1–41) Aristotle derived most of his material—virtually all of it, indeed—from the literary tradition, above all Androtion. There have been a few dissenting voices. For example, Gilbert in 1893 was prepared to credit Aristotle with a good deal of examination of ‘original documents’,3 in particular a collection of laws, including those of Solon and Cleisthenes. Quite recently WadeGery4 has spoken of ‘that scientific examination of Solon’s work (whether by Theophrastos or another) whose fruits we find in Aristotle’s Constitution of Athens’, and has confidently assumed that when Aristotle wrote Ath. Pol. 20–1 he drew his information ‘from surviving contemporary documents: especially the actual Laws of Kleisthenes’.5 Not long before his death Jacoby expressed himself as ‘confident that Aristotle read the Solonian laws himself’:6 it appears from another passage7 that he conceived Aristotle as doing so ‘either in the original or in the Ms. of Theophrastos’ N
omoi’. (Nevertheless, as we shall see, Jacoby, unlike Wade-Gery, drew a vital distinction between the laws (n
omoi) of Solon, which were available to Aristotle, and his constitution (polite
ia), which had to be mainly inferred from the n
omoi.)8 The general opinion at the present day is that the historical part of the Ath. Pol. is founded almost exclusively upon the literary tradition available at the time Aristotle composed it—which is almost universally taken to be the mid-320s, although (as I have shown)9 the bulk of the work may well have been written some years earlier. Some historians are prepared to allow an occasional or even a considerable use of archives and other documentary sources, but my impression is that these historians are now a small minority. My own belief is that, for reasons I shall explain, we must allow for a much greater desire on Aristotle’s part to use documentary material than most recent historians, even Jacoby, have contemplated; but it remains a problem whether Aristotle had available to him much documentary material from before c. 403, or any such material, other than ‘the laws of Solon’, before the time at which the constitution of Cleisthenes came into operation. For the 3 4 6 8

CASA xxiii, cf. xxxiii–ix (¼ HGS I2 XII, cf. XXII–XXX). 5 EGH 197 n. 2. EGH 135: the italics are Wade-Gery’s. 7 Jac. i 147. Jac. ii 134 n. 8, on Androt. F 36. 9 See pp. 310–17. Pp. 260–73 above.

The Sources of the Athenaion Politeia

279

time being, however, I shall set aside the question of documentary sources and consider the possible literary ones. *** The only literary source Aristotle actually names in the Ath. Pol. is Herodotus,10 whom he has used at more than one point, and most obviously in ch. 20, describing the events of 508/7, leading up to the implementation of Cleisthenes’ constitution. The way in which Aristotle rephrases his source here is worth careful study, as the best example we possess of such a procedure in the Ath. Pol.11 Aristotle also used Thucydides12 and probably Xenophon,13 as we should expect. It is actually Thucydides (VI 58.2) who is the author of the received account (o‘ leg
omenoB l
ogoB) described as ’ alZu
’ false (oyk ZB) in Ath. Pol. 18.4. I have already14 said something about the way Aristotle sets aside Thucydides’ evidence about the nature of the revolution of the Four Hundred in 411. We may suspect that Aristotle perhaps also used Theopompus, at any rate the excursus on the Athenian ‘demagogues’15 which formed the last part of Book X of his Philippika.16 Conceivably 10 For the extensive use of Herodotus by Aristotle in his other works, see p. 270 above. 11 The parallel passages in Hdts (V 66–73) are conveniently given by ‘ & Sandys, ACA2 80–1 nn. Note esp. (a) esso
ymenoB de o‘ Kleisu
enZB t on dZmon prosetair
izetai [‘since he was being worsted, Cleisthenes took the people into ‘ vmenoB de ta&iB etaire
‘ partnership’] (Hdts), with Ztt
iaiB o‘ Kl: prosZg
ageto t on & dZmon [‘since he had been worsted in political alliances, Cleisthenes brought the ‘ ’ eetai Kleom
enea gen
omenon evyt v fi& people over to his side’] (A.P.); (b) epikal
’ je&inon [‘he called in Cleomenes, who was his xenos’] (Hdts), with epikales
amenoB ’
‘ t on Kleom
enZn, onta eayt v fi & j
enon [‘calling in Cleomenes, who was his xenos’] ’ & boylZB & [‘the boule resisted’] (Hdts), with tZB & de (A.P.); (c) antistaue
isZB de tZB ’ ‘ & antist
boylZB asZB [‘the boule resisted’] (A.P.); (d) Kleisu
enea kai t a eptak
osia ’ istia t ‘ o Kleom
eneoB metapemc
ep
a divxu
enta yp amenoi [‘recalling Cleisthenes and the 700 families which had been exiled by Cleomenes’] (Hdts), with ’
Kleisu
enZn kai toyB alloyB wyg
adaB metep
emcanto [‘they recalled Cleisthenes and the other exiles’] (A.P.). 12 Cf. p. 274 and n. 101 above. 13 See Endnote V (p. 325 below). 14 Pp. 273–4 above. 15 There is an excellent examination of the nature of the Greek ‘demagogue’ by Finley, AD. 16 FGrH II B i 115 F 85–100. According to Wade-Gery (EGH 236 & n. 1), Theopompus’ On the Athenian Demagogues (see Theopomp. F 100) was published separately as a pamphlet ‘in the late fifties’ of the fourth century. This seems possible. The date cannot be earlier, because of F 99–100, and may be several years later (cf. Jacoby, FGrH II D p. 358). I agree with the opinion that the abuse of Cleon in Ath. Pol. 28.3, for example, may well come from Theopompus, though

280

Athenian Democratic Origins

he also used Ephorus; but I know of no clear indication that he did,17 and I see no reason why he should have done. ‘Ephoros was a bookworm, who composed his great work almost exclusively from other books’;18 and his sources for Athenian history must all have been easily available to Aristotle himself.19 If we may judge by the fact that there are few surviving fragments of Ephorus dealing with Athens, later writers apparently found little in his history on this subject that was of any real interest. There remain the Atthidographers, behind whom no one in this post-Jacoby era will wish to create literary phantoms like Seeck’s ‘Anonymus of 392’ or Wilamowitz’s ‘Anonymus of c. 380’.20 I take it that Jacoby’s remarkable book, Atthis, will be universally admitted to have proved (in his own words)21 ‘that the ‘‘Atthis’’, i.e. the history of Athens, does not derive from an old and semi-official chronicle kept by the priestly board of Exegetai’—as Wilamowitz and his followers had maintained22—‘but was created in the lifetime of Thukydides by a learned man, the foreigner Hellanikos of Lesbos. He, while relying on living memory and to a certain extent on documents (among which the archons’ list seems by far the most important), had in the main to use his imagination in order to fit a mass of isolated, often contradictory, and mostly undatable evidence into the framework of a continuous history.’ There can indeed be no doubt that the tradition utilised by the historians and Atthidographers was above all oral tradition,23 not written down in works intended for publication until there may be merely a common source (Wade-Gery, EGH 263 & n. 3, believes it was Critias). Day and Chambers (AHAD 11, 79) suggest that Theopompus is the source of the accusations against Solon mentioned in Ath. Pol. 6.2. Since the names in the corresponding passage in Plutarch (Sol. 15.7) show that the slanders were invented much earlier, Theopompus cannot have been their inventor; but of course he may have repeated them. 17 The use of Ephorus by Aristotle, even Androtion, is of course chronologically possible: the history of Ephorus was perhaps published in separate parts from c. 356 onwards (see Jac. ii 55 n. 34; cf. II C pp. 24–5; G. L. Barber, The Historian Ephorus (1935), 7–14), and all the books down to the mid-fifth century at any rate may have been published by the early 340s—see Wade-Gery, EGH 205, whose view is regarded as at least possible by Jac. ii 55 n. 34; 97 n. 110b. 18 19 Jac. i 10. Cf. Jac. ii 55 n. 34. 20 On these imaginary figures, see Jacoby, A 3–8, 210, 225, 230–1 n. 25, etc. 21 Preface, p. v. 22 See Wilam. AuA, esp. I ch. 8, pp. 260 ff. Wilamowitz believed the chronicle was long kept in a ‘pre-literary’ form but was published as a literary work about 380. 23 See Jacoby, A 165 ff., esp. 169–70.

The Sources of the Athenaion Politeia

281

the late fifth century, but preserved mainly in the memory of the ’
ndreB [‘learned men’]24—above all, the ei’ d
oteB, the l
ogioi a members of the principal political families. As far as I know, there is no evidence that any Greek before Aristotle had ever done any systematic research into archives or other documents: this is a fact of some significance for our purposes, as will appear later. The most we can say of Aristotle’s predecessors is that a few of them seem to have done some ‘research into archives’ of a very elementary kind, publishing local lists—whether of officials (secular or sacred) or of victors in the Games—or using them as pegs on which to hang works of univer’ Zr sal local history. The earliest of these writers are Hellanicus, an & ‘
HraB poly
istvr [‘a much-enquiring man’]25 (whose ‘I
ereiai tZB ’
’Argei [‘Priestesses of Hera at Argos’]26 was certainly some ai‘ en kind of universal chronicle,27 in contrast with his later Atthis, which was a local history), and Hippias, who may have compiled ’ the first chronicle of Olympic victors ( ’Olympiakvn & anagraw
Z)28 29 30 about 400, although whether he really did so, and whether his work was anything more than a mere list of names with the briefest additions,31 are questions which in the present state of our knowledge should probably be left unanswered. 24

See Jacoby, A 215 ff., esp. 216; cf. 169–70, 204–5. FGrH I 4 T 13. 26 FGrH I 4 F 74–84. 27 See Jacoby, GH 282–4 (¼RE VIII 143–8); A 357–9 n. 26; Jac. i 1 ff., esp. 2–5. 28 FGrH I 6 F 2, ap. Plut., Numa 1.6. 29 See Jacoby, FGrH I p. 477; cf. A 281 n. 48: Hippias was ‘later than Hellanikos (as it appears, considerably so).’ 30 The only surviving reference to this work is by a writer who evidently did not know it at first hand and did not think much of it, or at any rate the earlier part of it: ffl t ’ ’ e Plutarch (see n. 26 above) speaks of the ’Olympionika
i as vn Zn anagraw Zn oc
’ ’ ’ ‘ vmenon anagka
’ wasin ‘Ipp
ian ekdo &ynai t on ’Hle&ion, ap’ oyden oB orm
ioy pr oB p
istin [‘they say that Hippias of Elis published the list (of Olympic victors) at a late date, starting from no basis so authoritative as to inspire confidence’]. Cf. Jacoby, A 88: ‘No list of victors began to be kept before the seventh century; there are serious doubts about . . . the list of the ’Olympionika
i and its era 776/5 B.C..’ Contrast, however, Jacoby’s comment on Hippias F 2, in FGrH I p. 477. 31 Jacoby had difficulty in making up his mind whether Hippias’ ’Olympionika
i were a local or a universal history: in his commentary in FGrH I p. 477 (of 1923) he says they are ‘eine lokale publikation [a reference to Dion. Hal., De Thuc. 5 follows] . . . die allgemeine chronographische abzweckung der ‘I
ereiai hatten sie kaum’. In his commentary on the Atthidographers (Jac. i 381), written in 1940–5 although not published until 1954 (see the Preface to Jac. i), he says the work ‘is perhaps more likely to have been a local book than a universal chronicle’. However, in his Atthis, published in 1949 but written about the same time as the commentary 25

282

Athenian Democratic Origins

Hellanicus published his Atthis at the very end of the fifth century. Doubtless Aristotle read him, but what use he made of him we have no means of telling. After Hellanicus, only three Atthidographers are of any real importance for our present purposes: Cleidemus, Androtion and Philochorus. (Like everyone else, I shall refer to these men’s works as if each of them bore the title ‘Atthis’, although in no case is the actual title known with certainty.)32 I need say nothing at this stage of Philochorus, since he wrote in the third century, after Aristotle’s death, and for our purpose is interesting mainly in the way he made use of Androtion and Aristotle.33 Cleidemus, an Athenian and an Exegete, is very probably the earliest of the native Athenian Atthidographers,34 but his date is very uncertain. Jacoby, proceeding from the assumption—which seems to me almost certainly wrong—that the hundred symmories referred to in Cleidemus F 8 must be the trierarchic and not the eisphora symmories,35 believed that Cleidemus published his Atthis (or perhaps Protogonia)36 ‘between 354 and 340 B.C., probably nearer the former date, i.e. c. 350’.37 Although I am quite prepared to accept a mid-fourth-century date as highly probable, I do not think we can be certain, except of a terminus post quem in 378/7, the date of the introduction of the eisphora symmories.38 Not many fragments survive, and they contain little that is of any interest for the political and constitutional history of Athens. The & gynaik
phrase paraibato
ysZB tZB oB [‘with the woman standing beside him’] in Aristotle’s version of the Phye story (Ath. Pol. 14.4), where the participle is ‘a noteworthy Ionism’,39 has sometimes been thought to come directly from t Zn paraibat
Zsasan ‘ vi ayt & gyna&ika [‘the woman standing beside him] in Cleidemus on the Atthidographers, he says that Hippias’ book ‘certainly was not a local chronicle’ (A 297 n. 6)! I do not myself see how we can feel confident that the work was much more than a mere list of victors, although of course it may have been. 32

See Jacoby, A 83–6. Cf. pp. 291–3 below. The only evidence seems to be Paus. X 15.5 ¼ Cleidemus, FGrH III B 323 T 1. According to Jacoby (Jac. i 58), ‘the assertion that he [Cleidemus] is the ‘‘oldest’’ Atthidographer must not be doubted, even it is only an inference [by Pausanias].’ 35 [Ste. Croix refers here to a Long Note which was never written or has been lost.] 36 37 See Jac. i 58–9. A 1, cf. 74–6; Jac. i 58. 38 39 See my DTAEFC, esp. 56–62. Sandys, ACA2 60 n. 33 34

The Sources of the Athenaion Politeia

283

F 15.40 However, Aristotle, who can hardly have failed to read Cleidemus, ignores or rejects the two interesting facts about Phye (if such they were) mentioned by the Atthidographer: that she ‘ Svkr
was F
yZ Z atoyB [‘Phye daughter of Socrates’], i.e. the daughter of a citizen,41 and that Peisistratus married her to his son Hipparchus; and this of course gives an opportunity for bringing in Androtion as an intermediate source.42 Elsewhere, in giving a story (generally believed to come from Androtion) about the & peri Salam&ina way in which the Areopagus had become tZB naymax
iaB ai’ t
ia [‘responsible for the sea-battle off Salamis’], Aristotle (Ath. Pol. 23.1) ignores the entirely different story in Cleidemus F 21, of which Themistocles is the hero. We are left with Androtion, who is universally credited with being an important source, indeed the main source, of the Ath. Pol. Androtion43 was almost certainly the son of Andron,44 who seems to have been a political associate of Theramenes in 411, and in that year moved the resolution for the prosecution of Antiphon.45 Androtion himself was a pupil of Isocrates, and played a prominent though not a leading part in politics from at least the early 370s onwards.46 Demosthenes (XXII 47) could refer to him, scornfully ’ ffl it is true, as o‘ kal oB kagau oB oytoB [‘that excellent gentleman’]. Jacoby47 dates the publication of his Atthis to the years immediately following 344/3, or let us say c. 340, and this date seems to be generally accepted. According to Plutarch,48 Androtion wrote in exile at Megara. He was certainly active in politics at Athens as late as 347/6, when he proposed a decree in honour of the sons of Leucon;49 and some historians, including Jacoby,50 believe that 40

41 On which see Jac. i 70–2, with 60–1. See Jac. i 71. See Busolt, GG II2 33 n. 1, on p. 34. 43 See for all these details the Testimonia in FGrH III B 324 (with F 72) and Jac. i 87 ff. [and Harding, ATA]. 44 45 Kirchner, PA 921. Ps.-Plut., Mor. 833D–4B. 46 In addition to the speech against him by Demosthenes (XXII), we hear of another by Lysias (Harp., s.v. ’Arxid
ameioB p
olemoB, and s.v. Ser
aggion), which is lost. If the Androtion concerned is indeed the Atthidographer (as seems very likely), and if this was a genuine speech by Lysias, then probably Androtion was already prominent by the 380s. 47 A 74. 48 Mor. 605C (¼ De Exilio 14), accepted by Jac. i 92, 103. 49 Tod II 167 ¼ Harding 82. The prescript shows that the decree dates from the late spring of 346. 50 Jac. i 90–3, cf. 531–3. See also Glotz-Cohen, HG III 319–22; Bloch, SHLFC 343–4 (Jac. ii 87 n. 58 gives a wrong reference). 42

284

Athenian Democratic Origins

he not merely recounted in his Atthis the debate in the Athenian Assembly in 344/3 over the reply to be given to a Persian embassy, but actually spoke during that debate, and perhaps was even responsible for the motion then adopted.51 The exile of Androtion, whether self-imposed or inflicted as a penalty, has sometimes been taken to be the direct consequence of his role in the debate in question;52 but this is far from certain. The Atthis of Androtion can therefore be dated after 347/6, probably after 344/3, and since Androtion seems to have been an elderly man by this time, it will presumably have been written not much later. Of course Aristotle will have read Androtion and made use of him. But how far is Androtion the principal source of the Ath. Pol.? How far did Aristotle rely upon him, borrow from him, take his material without verification from him? It has long been the standard view that Androtion is Aristotle’s major source for the Ath. Pol., his ‘Hauptquelle’. So widespread is this opinion that I need not multiply quotations from modern authors, apart from Jacoby, whose judgement on such a matter has great authority. According to the latest work on the Ath. Pol., Day and Chambers’s Aristotle’s History of Athenian Democracy (1962),53 Aristotle ‘used Cleidemus and depended most heavily on Androtion, who is probably his usual source for the factual narrative on which the AP is constructed. This view of his sources is universally accepted.’ No proof is given. Hignett54 can speak of ‘the large borrowings made by the author [of the Ath. Pol.] from the Atthides, especially Androtion’s’, as if these ‘large borrowings’ were an undoubted fact. Again, no detailed evidence is given, at this point or elsewhere. Jacoby, whose opinion on such a question is far weightier than that of any other scholar, living or dead, makes a series of confident statements about Aristotle’s dependence upon Androtion. Endorsing the view of Wilamowitz55 that we must take it for granted, even in the absence of proof, that Aristotle would have used Androtion, 51 Everything depends upon Diels’s restoration of the Didymus papyrus, ’Androt
ivn oB ‘ kai t[
ot’ ei
pe [‘Androtion, who t[hen proposed the motion’]: see Androt. 324 T 13 and F 53 ¼ Philoch. 328 F 157 ¼ Anaximenes 72 F 28. [For scepticism about Diels’s restoration see Harding, ATA 180, citing Cawkwell.] 52 See e.g. Glotz-Cohen, HG III 321–2; Jac. i 92–3. 53 P. 7, with a useful brief bibliography in n. 25. 54 55 HAC 30, cf. 9, 12 n. 3. AuA I 276–7.

The Sources of the Athenaion Politeia

285

Jacoby adds,56 ‘This fact, to which I shall come back repeatedly in the commentary on Androtion, is probably correct and seems to be generally agreed upon.’ (The ‘probably’ is significant.) Jacoby also says that for the historical section of his Ath. Pol. ‘Aristotle could take . . . not only facts but often the angle and interpretation of the facts as well from Androtion, who is almost universally agreed to be his main source, his handbook as it were for the history of Athens’;57 that Androtion is ‘the ordinary source of Aristotle in the ’Aup. for the facts of the history of the constitution’;58 that he is Aristotle’s ‘main Atthidographic source’ (Jacoby uses this expression at least four times);59 that Androtion was Aristotle’s ‘chief source in the survey of the history, or rather the constitution, of Athens: he probably took from A[ndrotion] all his dates and most of the historical facts’;60 that ‘the philosopher based his work on A[ndrotion]’s Atthis’;61 that ‘even when eliminating everything that is doubtful we may confidently maintain that A[ndrotion] stands as a source for Aristotle beside, or even before, Herodotos and Thukydides. . . . The Atthis of A[ndrotion] is the book which gave Aristotle the general frame-work, and which was used in the historical introduction for the details of Attic history and Attic institutions mostly, even if not alone.’62 This is certainly the prevailing view nowadays.63 Yet Jacoby can also admit, in the middle of a two-and-a-half page note on his Introduction to Androtion, that ‘The general 56

A 234 n. 36. The whole of this long note is worth careful study. A 213. Jacoby does add, ‘But there are places where Aristotle, on the basis of other accounts, corrects not only Androtion’s view of facts but the facts themselves . . . , and in these cases the philosopher proves the better historian’. 58 59 Jac. i. 160. Jac. i 145, 146; ii 55 n. 3; 133 n. 17. 60 61 Jac. i 103–4. Jac. ii 99 n. 124. 62 Jac. ii 101; cf. 91–2 n. 86; i 231. For Androtion as the alleged source of individual passages in the Ath. Pol., see Jac. i 99 (with ii 95 n. 104; 107 n. 27), 460, 467 (with ii 379–80 n. 27; 384 n. 61), 566–7; ii 81–2 n. 29, 357 n. 3 with 359 n. 9, 381 n. 33. 63 Even Ledl (SAAV 14–17, at pp. 16–17), while objecting to what I am calling ‘Keil’s theory’ (see pp. 296 ff. below), that Androtion is the source of various passages common to Plut., Sol. and the Ath. Pol., evidently did not wish to challenge the theory that Androtion was an important source of the Ath. Pol. See also Bloch, SHLFC 349 n. 3 (‘That Androtion’s Atthis was one of the main sources of Aristotle for his Athenaion Politeia is a fact which should no longer be doubted’), a statement which seems to be the only basis for the assertion of von Fritz and Kapp (ACA 13 n. 11) that Bloch ‘has proved on the basis of incontestable evidence that Aristotle made use of the Atthis of Androtion in his treatise on the Athenian constitution’. 57

286

Athenian Democratic Origins

assumption of the importance of A[ndrotion] for Aristotle has actually never been proved, and a systematic examination of the connexion between the two writers has never been made’!64 In fact, as I shall show in detail presently, we know very little indeed of what Androtion, or any other Atthidographer, wrote about the constitutional history of Athens in the sixth and fifth centuries— the subject with which almost the whole of the surviving part of the historical section of the Ath. Pol. is concerned. We must remember that the Atthidographers were writing general histories of Athens, not simply, or even mainly, constitutional histories;65 and in the absence of evidence we have no right to assume either that they necessarily dealt at all fully with constitutional matters or that they went to any great pains to ascertain the details of constitutional history. I would urge that until someone has carried out that ‘systematic examination of the connexion’ between Androtion and Aristotle, the absence of which was noted by Jacoby, we must cease to repeat like parrots that Androtion is the ‘Hauptquelle’ of the Ath. Pol. Meanwhile I offer some observations, mainly (I am afraid) of a disappointingly negative character, devoted to bringing out what we know about the relation between Aristotle and the Atthidographers, and what the main problems are.

(III) ARISTOTLE AND THE ATTHIDOGRAPHERS In what follows I shall consider first, as a test case, what the three main early (i.e. pre-Aristotelian) Atthidographers are known to have said about the two great Athenian lawgivers of the sixth century, Solon and Cleisthenes. Secondly, I shall examine the explicit evidence for supposing that Aristotle relied heavily upon Androtion (the representative early Atthidographer) as an authority for his historical material in the Ath. Pol. We shall find that under these two heads the evidence is very slight indeed. Thirdly, I shall try to show what in my opinion is the real basis of the theory of Aristotle’s close dependence upon Androtion: a set of infer64 Jac. ii 99–102 n. 127, at p. 100. Nevertheless in i 146 Jacoby, admitting that ‘we must not count F 35 among the ‘agreements’ between A[ndrotion] and Aristotle, which are to prove that the former was the main Atthidographic source of the latter’, adds ‘This fact is established otherwise’—and gives a reference to this very note! 65 See Jacoby, A, esp. ch. II.

Aristotle and the Atthidographers

287

ences—which I for one am not prepared to accept as even probably true—from certain correspondences between the Ath. Pol. and Plutarch’s Life of Solon. Fourthly, postponing to the next section of this essay the question how far official documents were available to Aristotle, I shall show there are grounds for supposing that Aristotle is much more likely to have searched for and used any available documentary material, including official archives, than any of his predecessors—especially, as it happens, Androtion. For this reason, the question whether Aristotle did or did not take much of his material without verification from Androtion (or another Atthidographer) is a significant one: it involves more than the mere replacement of one work of literature by another as the origin of much of our information about early Athenian constitutional history. 1. I begin, then, by trying to establish what the early Atthidographers said, beyond reasonable doubt, about the legislation of Solon and Cleisthenes. We have only a few insignificant scraps of evidence. (a) There is no evidence at all from the fragments of Hellanicus to tell us whether he so much as mentioned Solon, although of course he must have done.1 Jacoby speaks of ‘the fact that according to the fragments, in which Solon’s name is lacking (perhaps accidentally), the constitution of Kleisthenes was dealt with in detail’;2 but the sole foundation for this alleged fact, as far as I can see, is Hellanicus F 6, in which we are given the name and ancestry of Hippothoo¨n, the eponymous hero of one of the Cleisthenic tribes. From this all we can legitimately infer with reasonable certainty is that Hellanicus gave an explanation of the names of the ten tribes, or at least of those which were not too familiar to need explaining; but whether he did this as part of a general account of Cleisthenes’ reforms, we cannot tell. It is probable that he did; but if we may judge by the nature of the surviving fragments of his Atthis (perhaps not at all a safe guide), Hellanicus may have been more interested in the eponymous 1 See Jacoby, A 154; Jac. i 20; cf. 29–30, where Jacoby thinks it possible that ‘H[ellanicus] knew no particulars and no details about’ Solon. See also Jac. ii 9 n. 104: ‘There is nothing about Solon and little about the tyrants (F 9?). That may be merely accidental, but I am not sure that it is.’ 2 Jac. i 20 [my italics]; cf. 29–30. Contrast the rather more cautious statement in Jac. i 28.

288

Athenian Democratic Origins

heroes of the Athenian tribes than in any details of Cleisthenes’ constitution. (b) Of Cleidemus too we have no evidence what he said about Solon.3 Hignett’s inference4 that ‘the democratic version of Solon’s political achivement . . . must surely have been reproduced at some length by Kleidemos’ is founded on what seem to me a series of doubtful assumptions. Eberhard Ruschenbusch5 has claimed that Cleidemus must be responsible for the conception of Solon as the founder of radical democracy; but his whole position is misconceived,6 and his fixing upon Cleidemus as the inventor of the view in question is arbitrary—indeed, on Jacoby’s chronology, which Ruschenbusch himself accepts, Cleidemus could not possibly have been responsible.7 For Cleisthenes we have nothing certain from Cleidemus but F 8, a statement about the creation by Cleisthenes of fifty naucraries which some scholars have refused to accept.8 Jacoby’s assumption9 that F 7 may also be ‘assumed to belong to the account of Kleisthenes’ organisation of the state’ is gratuitous; and the fragment is of trivial importance anyway. (c) The situation is hardly different with Androtion. The total number of certainly genuine fragments surviving from his Atthis is 68, of which no less than 24 (roughly a third) are from purely topographical notes by Stephanus of Byzantium. Of the remaining 44 fragments, at most six (F 3/4, 5, 6, 34, 36) deal with the constitutional history of Athens prior to the mid-fifth century. Only one fragment (F 34) unmistakably alludes to Solon, and one (F 5) to 3 Against taking Cleid. F 14 as evidence that Cleidemus ‘appealed to the Kyrbeis of Solon’, see Jac. i 70. Jac. i 65 (the last sentence on F 7–8) seems to imply that Cleidemus deliberately ignored Solon. 4 5 6 HAC 7–8. PP 420 and n. 78. See pp. 313–17 below. 7 As we have seen (p. 00 above), Jacoby’s earliest date for Cleidemus is 354—and of course that is the earliest possible year, if Cleid. F 8 refers to the trierarchic and not the eisphora symmories (cf. p. 282 above), for Dem. XIV, which refers to 20 naval symmories, cannot be put earlier than 354, in which year Ruschenbusch himself dates it (ibid. 399). Ruschenbusch (ibid. 420 n. 78) cites Jacoby (A 74 f.) for the date of Cleidemus, but calmly speaks of ‘die a. 355 vero¨ffentlichte Atthis des Kleidemos’, whereas Jacoby had referred to Cleidemus’ work as appearing ‘not earlier than 354 B.C.’. Ruschenbusch sees in Dem. XX and XXII the earliest surviving appearance of Cleidemus’s conception of Solon the inventor of radical democracy; but he himself dates these speeches to 355 (ibid. 399, 405 & 31), and in fact Dem. XXII is much more probably to be dated 357 (see D. M. Lewis in BSA 49 [1954] 39 ff., esp. 44). 8 9 [e.g. Busolt, GG II2 (1895), 418; Hignett, HAC 73.] Jac. i 65.

Aristotle and the Atthidographers

289

Cleisthenes: the former is a deliberate distortion of the truth about the Seisachtheia,10 and contains a crassly stupid comment upon the effect of Solon’s alleged currency reform;11 and the latter, which makes Androtion say that the Apodektai were put by Cleisthenes in the place of the Kolakretai, is either a bad mistake by Androtion or a misrepresentation of what he actually said. In the opinion of Wilamowitz and Jacoby,12 it is ‘self-evident’ that Harpocration or his excerptor is at fault; but an error of this kind would be unusual in Harpocration, and I should prefer to suspend judgement.13 I myself believe that F 6, if it is really Androtion at all (and this is very doubtful), comes from his account of Cleisthenes’ constitution: I have already dealt with this problem in an earlier essay,14 and I shall return to it later, since F 6, because of its close linguistic similarity to Arist., Ath. Pol. 22, is vitally important for the relationship between Androtion and Aristotle. Jacoby15 believed that F 4 must come from Androtion’s description of Solon’s constitution; but it is concerned entirely with the Areopagus and does not mention Solon, and the whole question is uncertain in the extreme, as Jacoby’s commentary shows. F 36, according to Jacoby,16 probably also comes from Androtion’s description of the constitution. ‘I suggest’, he adds, ‘that A[ndrotion] adduced some facts which would show us how Solon brought Athens into the sphere of Delphi.’ This may or may not be true. Two remarks may be added. (i) The actual fragment gives a law which specifies one of the duties of the Kolakretai and will therefore have been a law (n
omoB) in the narrower sense, rather than a part of Solon’s constitution (polite
ia) (see pp. 310–17 below). (ii) The statement in Ath. Pol. 8.3 about ‘the laws of 10

See Jac. i 144–6; ii 356–7 n. 7; cf. 454–67. As Jacoby (ii 467) rightly says, ‘Androtion expects us to believe that Solon’s currency reform greatly benefited the debtors without injuring the creditors who only got back two thirds of the silver they had lent out.’ He adds, ‘It is to the credit of Aristotle that he treated this astonishing statement of the ‘‘shrewd financier’’ with the contempt it deserves.’ 12 Wilam., AuA I 52 n. 19; Jac. i 118–19. 13 Jacoby’s statements in ii 113–14 n. 18 seem to me tendentious. His appeal to F 36, in i 119 (the end of the commentary on F 5), is entirely unavailing if one does not accept the assumption he makes here that Androtion supposed the law cited in F 36 to be Cleisthenic or post-Cleisthenic—an assumption Jacoby himself neglects in his commentary on that fragment: see Jac. i 147, and below. 14 15 16 Pp. 182–8 above. Jac. i 114–17. Jac. i 146–8. 11

290

Athenian Democratic Origins

Solon which they no longer use’17 is not merely derived from or dependent upon Androtion F 36: Aristotle says that in these laws pollaxo &y g
egraptai ‘‘toyB naykr
aroyB ei’ spr
attein’’ kai ’ ’ to &y naykrariko &y argyr
’ ‘‘anal
iskein ek ioy’’ [‘it is written in many places ‘‘let the Naukraroi exact’’ and ‘‘let them pay out from the Naukrarian treasury’’ ’], whereas Androtion gives a single passage only, in which the Naukraroi are not mentioned and it is the Kolakretai who pay the money. It is a characteristic example of the readiness of modern scholars to ascribe anything and everything in the Ath. Pol. to Androtion that Hignett, going beyond Wilamowitz,18 should say, after mentioning Androt. F 36, ‘Androtion was probably the source of the citation from ‘‘the old laws of Solon’’ in A.P. 8.3’. Jacoby19 says that Androtion and Philochorus ‘on the occasion of the constitution of Kleisthenes, not only gave the lists of the ten phylai and the (originally) hundred demes, . . . but they also explained the names topographically, etymologically, and mythologically; they enumerated the official cult-heroes with the genealogies of their descendants, and they probably added the position, and certainly the peculiarities, of each deme.’ This is to confuse Androtion with Philochorus. The only evidence Jacoby can give20 for Androtion’s having done all this is his F 66–8; and when we look at these fragments and Jacoby’s commentary upon them21 we see at once that there is not the least reason to think they are part of a description of the Cleisthenic constitution as a whole, although the corresponding fragments of Philochorus (328 F 24–29, probably 205–6) very probably are.22 I do not think we can even be certain, then, that Jacoby23 is right in believing that ‘Androtion24 treated the legislation and constitution of Solon in detail’—or the constitution of Cleisthenes either. He may have done so: he is bound to have said something about 17

18 Cf. pp. 320–1 below. HAC 12 n. 3; cf. Wilam., AuA I 51–2. 20 21 A 128. A 320 n. 174. Jac. i 170. 22 23 See below. Jac. i 30. 24 Jacoby has ‘Androtion and Philochorus’. I am not directly concerned with Philochorus here, but I should like to point out that of the fragments of that author to which Jacoby (ii 34 n. 19) appeals—i.e. F 20?, 21, 35 and 114—none affords good evidence for any extensive treatment of Solon by Philochorus. F 20 is simply Androt. F 4. F 21 need not have referred to Solon. F 35, as is now clear from A. Andrewes’s brilliant reinterpretation, in JHS 81 (1961) 1–15, has nothing to do with Solon. And F 114 tells us almost nothing. 19

Aristotle and the Atthidographers

291

both these important topics. But we do not know what he said, or even that he said much that was historically accurate. (d) Some light on the problem we are now considering is perhaps to be had from Philochorus, the ‘close connexion’ between whose Atthis and that of Androtion has often been emphasised, for instance by Jacoby,25 who says, ‘This connexion has been noticed already in antiquity as is shown by numerous quotations mentioning Philochorus and Androtion alongside of each other as the best witnesses, or as those agreeing on some special point.’ We ought, then, to look at the fragments of Philochorus, to see if there are any which refer, or may refer, to the constitutions of Solon or Cleisthenes, and whether there is any connexion between them and the fragments of Androtion—or for that matter the Ath. Pol. Apart from Philoch. F 20, corresponding to Androt. F 4 (which I have already dealt with above), there are only three fragments of Philochorus which seem to come from his treatment of Solon: probably (i) F 21 and certainly (ii) F 114 and (iii) F 200. In (i) F 21 (the oaths taken ‘near the stone’)26 Harpocration cites the Ath. Pol. (7.1; cf. 55.5) and Philochorus together. (The passage in Aristotle is in his account of Solon, and the one in Philochorus comes from his Book III, in which he probably dealt with Solon.) Jacoby27 thinks that ‘Androtion may be the source common to both.’ In fact Harpocration, who cites Androtion and Philochorus together five times and Aristotle and Philochorus together four times, here cites Aristotle and Philochorus, not Androtion; and there is no reason to suppose Androtion said anything similar, though of course he may have done. (ii) In F 114 Philochorus evidently agreed with Aristotle against Androtion (F 34) on the nature of the Seisachtheia. (iii) Of F 200 Jacoby28 says, ‘We may safely assume that Philochorus, when speaking about Athenian coinage, followed Androtion, whose account of the activity of Solon in this province underlies also the famous chapter 10 of Aristotle’s ’Aup.’ But for once we have the corresponding fragment 25 Jac. i 244; and see ii 178 n. 188, listing the parallels, viz.: Philoch. F 4/Androt. F 3; Ph. 20/ A. 4; Ph. 47/A. 24; Ph. 52/A. 52; Ph. 125/A. 39; Ph. 136/A. 43; Ph. 147/A. 18; Ph. 150/A. 48; Ph. 155/A. 30; and cf. Ph. 157 and A 53. 26 In addition to Jac. i 312, see Sandys, ACA2 25 n., who deals better with the variant in Plut., Sol. 25.3. 27 28 Jac. i 312. Jac. i 566.

292

Athenian Democratic Origins

of Androtion (F 34), and certainly the Philochorus fragment is nearer to Aristotle than to what we have of Androtion29—although the latter’s text may not have been transmitted to us accurately, coming as it almost certainly does through some intermediate source (Hermippus?) as well as Plutarch (Sol. 15.3). Of Philochorus’ account of the constitution of Cleisthenes we have no less than nine fragments, eight of which (F 24–9, 205–6) are from what seems to have been a list of the Cleisthenic demes, with notes explaining the names and sometimes making other comments. I think we can say it is unlikely that Androtion gave any such list,30 for Harpocration, who is responsible for F 24–9 of Philochorus, cites Diodorus the Periegete at least twenty-seven times on the demes, Hypereides (twice) and Pherecrates, but never Androtion, whom he does cite elsewhere in his lexicon, fourteen times in all. The last of our nine Philochorus fragments, F 30 (on ostracism), is likely to come from the account of the work ’ amenon nomouet
of Cleisthenes, since the words arj
ZsantoB Kleisu
enoyB [‘which began through the passing of a law by Cleisthenes’], though not quoted from Philochorus, would hardly have been included if he had assigned a different date to the introduction of ostracism. In view of the uncertainty about Androt. F 6,31 we cannot be sure whether Philochorus is here following Androtion as well as Aristotle. To sum up—the fragments of Philochorus, ‘the first scholar among the Atthidographers’,32 and the way they are used by Harpocration in particular, do nothing to confirm the view that Androtion’s work on the early history of the Athenian constitution was particularly important and lies behind much of Philochorus. Indeed, what little evidence we have rather militates against that view. The ‘close connexion’ between Androtion and Philochorus which is visible in fragments dealing with later times is not at all 29 ’ ZuZ de t Cf. Philoch.’s ekl
o n
omisma t o tetr
adraxmon . . . t vn & prot
ervn ’
didr
axmvn ontvn ktl. [‘the tetradrachm coinage was called . . . the earlier (standard)
d’ o‘ coins were didrachms etc.’] with Aristotle’s t Zn to &y nom
ismatoB ay
’ jZsin . . . Zn ’ arxa& ioB xarakt Zr d
idraxmon [‘the increase in (the value of) the coinage . . . the old (standard) coin-type was the didrachm’]. 30 Jac. thinks that Philochorus ‘evidently gave the total list’ of Attic demes, and that ‘his list presumably was the first’ (Jac. i 313), that of Diodorus the Periegete being presumably ‘later, if only by a few years’ (Jac. ii 227 n. 4 on F 24–9). 31 See pp. 182–8 and 289 above. 32 Jac. i 227. See also the impressive list of his works, ibid. 242.

Aristotle and the Atthidographers

293

apparent for the early period in which we are interested. There is in fact no trace of any connexion at all, unless we assume from Philoch. F 21 and 200 that because Aristotle as well as Philochorus said something, Androtion ‘must have’ said it first. Some people do make this assumption—Jacoby, for example, who says that Philochorus ‘did not use Aristotle’s ’Aup., nor could he be expected to do so: the ’Aup., on the contrary, depends upon the Atthides in its historical sections and on Androtion in particular, whom Philochorus himself always had before him.’33 Yet when we turn to the note on this passage34 we find no evidence given that Philochorus did not use the Ath. Pol. I do not myself know of any evidence either way. Of course Philochorus, like Aristotle, will have read Androtion, and it looks as if he may have drawn upon him heavily for the late fifth century and the fourth; but that is no guarantee that he regarded Androtion as a reliable source for the constitutional history of early Athens. 2. Apart from inferences from parallel passages in Plutarch’s Solon (to be considered in section 3 below), what is the evidence that Aristotle relied largely upon Androtion as an authority for the historical section of the Ath. Pol.? (a) I think we should set aside as a special case Ath. Pol. 29.2/ Androt. F 43 (the thirty Syngrapheis) and any further supposed— even highly probable35—dependence of Ath. Pol. 29–33 (dealing with the Four Hundred and the Five Thousand) upon Androtion, since on such a subject Androtion might be expected to have exceptional sources of information, through his father Andron, a leading ‘Theramenist’ member of the Four Hundred,36 and might reasonably be regarded as something near to a primary source, worthy of unusual confidence. Any use by Aristotle of Androtion here need not necessarily lead us to expect dependence elsewhere. (b) On at least one important factual question, the nature of the Seisachtheia of Solon, Aristotle gives a completely different account from Androtion, whose version (F 34) was clearly distorted, as Jacoby has shown,37 for the purpose of political propaganda. It is interesting to find that Philochorus also rejected Androtion here.38 33 36 38

34 35 Jac. i 231. Jac. ii 175 n. 89. See pp. 273–6 above. 37 See p. 283 above. See p. 289 & n. 10 above. Philoch. 328 F 114: see Jac. i 449.

294

Athenian Democratic Origins

(c) How far Aristotle’s statements about Solon’s reform of measures, weights and coinage, in Ath. Pol. 10, are actually derived from the passage in Androtion which we know as F 34, in the doubtless abbreviated version of Plutarch (Sol. 15.3–4), it is hard to say. The two accounts are quite similar, especially if we adopt (as Jacoby does) the certainly correct emendation of the Plutarch ‘ Zkont’ MSS proposed by Th. Reinach in 1928,39 and read ebdom
’
goysan [‘weighing seventy’] in place of the absurd ebdom
‘ a Zkonta
trivn & oysan [‘being of seventy-three’] of the MSS. (d) Apart from Androt. F 6/Arist., Ath. Pol. 22.3–4 (on which see pp. 182–8 above), quite a number of supposed borrowings by Aristotle from Androtion have been detected, but every one is based on pure assumption. For instance, Bruno Keil40 wished to see (i) Androt. F 10 behind Ath. Pol. 38 (the ‘Tens’);41 (ii) F 35 behind Ath. Pol. 15.3 (Pallene); and (iii) F 37 behind Ath. Pol. 28.2; but Jacoby has shown that no proof of the dependence of Aristotle upon Androtion can legitimately be based on any of these texts.42 Jacoby43 has also demolished Keil’s elaborate argument44 that an Atthis lies behind Plut., Sol. 19, and Arist., Ath. Pol. 8. 2,4. Jacoby himself singles out three points ‘in which the dependence of Aristotle on A[ndrotion] . . . , or the agreement between the two writers, is fairly certain’;45 but two of the three points of alleged correspondence are in fact very doubtful,46 although the third, the 39

40 In Hermes 63 (1928) 238–40. SVAVA 190–1. Keil must have meant ch. 38, though the number he gives is 29. 42 See for (i) Jac. ii 129 n. 7 (on F 10–11), with i 138; (ii) Jac. i 146; (iii) Jac. i 148. Cf. also Jac. i 152 on Androt. F 43. 43 Jac. i 113–17, esp. 114–15, with ii 110–11 nn. 40–1. 44 SVAVA 98 ff. 45 Jac. ii 91–2 n. 86. 46 It is a mere guess that (i) Arist., Ath. Pol. 29.3 (lines 13–14 OCT), derives from Androtion: it could easily be Aristotle’s own comment (and Jacoby admits this possibility in A 384 n. 30), although if the passage as a whole goes back to Androtion, as it may, the comment of course may come from him. Also, (2) Androtion may be responsible for the story in Ath. Pol. 23.1, but is it even probable that Aristotle took it from him? (Cf. Day and Chambers, AHAD 121, where Ath. Pol. 23.1 is said to furnish ‘positive evidence’ of the use of Androtion by Aristotle!) I would point out that Plut., Them. 10.6, quotes only Aristotle for the story in question—though it is certainly true that Plutarch quotes Androtion only once (in Sol. 15.3–4 ¼ Androt. F 34), and then probably not directly. Jacoby’s third point is dealt with in the text above. In my view the statement in Ath. Pol. 25.2 that Ephialtes took away from the ’ iueta [‘their additional, more recent powers’] is most unlikely to Areopagus t a ep
’ iueta have appeared in Androtion (cf. Jac. ii 106 n. 20; and for the meaning of t a ep
see Ath. Pol. 3.3). I must add that Androt. F 44 is not a good foundation for Jacoby’s 41

Aristotle and the Atthidographers

295

familiar theory that Androtion ‘supplied the documents for the reform of the constitution’ in 411 (Ath. Pol. 29–33)47 is probable enough. Again, the second board of Ten in 403, mentioned only by Arist., Ath. Pol. 38.3–4, is not ‘confirmed by a fragment of the work of Androtion’, as stated by von Fritz and Kapp48—unless of ‘ ZB & in Androt. F 10 is translated, quite unnecescourse kai tvn & ej sarily, ‘and the next Ten’. As Jacoby49 has said, ‘F 10–11 do not prove more than that A[ndrotion] gave an account of the election of the Ten ‘‘and the following events’’, and that he gave the full lists of names.’ It is impossible to tell whether Aristotle is following Androtion or not. No one would wish to deny that Aristotle used Androtion, and doubtless at many points his judgement will have coincided with that of the earlier writer. For the events of 412–10 Aristotle may well have considered him—if wrongly—the best available source. But of the actual fragments of the Atthidographers, only Androt. F 6 and perhaps F 10 afford any foundation at all for the theory that Aristotle depended largely upon Androtion and reproduced material for which he was the primary authority. 3. Those who wish to derive the Ath. Pol. mainly from Androtion have still one line of argument left. It is of course quite possible a priori that writers whose works we possess have drawn extensively from Androtion and the other Atthidographers, even where they have not quoted them by name and consequently no actual ‘fragments’ can be reconstructed. There is one work above all from which modern scholars have sought to prove the use of Androtion by Aristotle: namely, Plutarch’s Life of Solon. The theory which has been taken for granted by nearly everyone in the last 70 years is that Androtion must be the source of a particular group of passages common to that work and to the Ath. Pol. This theory is founded upon a series of assumptions: I believe all these are certainly unjustified and statement (ii 91 n. 86) that ‘the dependence of Aristotle on A[ndrotion] . . . does not seem restricted to the ’Aup.’: a ransom of one mina is quite likely to have been fixed upon other occasions than the one mentioned by Androtion, although certainly two minae is a more usual figure. 47 Jac. ii 91 n. 86 (giving a wrong reference to ‘ ’Aup. 19–23’). 48 ACA 184 n. 130. 49 Jac. i 138. Jacoby thinks it certain that ‘the second board of Ten is an invention of party politics’. He believes it to be unlikely, though not impossible, that Aristotle got it from Androtion. I agree.

296

Athenian Democratic Origins

some of them probably false. But before I examine the theory I should like to sketch its growth and dissemination. I shall refer to it as ‘Keil’s Theory’, because the first detailed exposition of it, less than two years after the publication of the London papyrus, was contained in Bruno Keil’s Die solonische Verfassung in Aristoteles Verfassungsgeschichte Athens (1892),50 a book which seems to me to have had an influence upon later writers altogether out of proportion to its merits. Since the publication of Keil’s book the great majority of scholars have been content to repeat his conclusions, without any re-examination of his arguments. Keil’s theory made a second appearance a year later in the very influential work of Wilamowitz, Aristoteles und Athen.51 Wilamowitz so rarely condescended to notice the opinions of other scholars that it is hard to say whether he was influenced by what Keil had written.52 At any rate, what I am calling Keil’s theory reappeared as part of the much more far-reaching reconstruction of the sources of the Ath. Pol. by Wilamowitz, according to whom (as is well known) the Ath. Pol. is largely derived from ‘the Atthis’ (especially in Androtion’s version),53 which in turn derives from an ancient chronicle kept by the priestly Exegetai. Jacoby’s demolition of the foundation of Wilamowitz’s theory, the supposed priestly chronicle, has vitally important consequences for the historical accuracy of the Ath. Pol.: if indeed the Ath. Pol. is scarcely more than Aristotle’s version of the latest Atthis—and one written by an exile at that—its value as a historical source becomes questionable, to say the least, and may be very small indeed. Keil’s theory soon took root strongly, although Eduard Meyer for one said he was not convinced.54 Two years after the publica50

See esp. 190–4, also 45–6, 49–50, 164–8, 170–3 etc. See I 260 ff. (esp. 276–7, 287–8), 305, cf. 123 n. 3. In his preface (I p. V) Wilamowitz seems to be claiming that the relevant chapters of his book were written before the publication of Keil’s. But there are several things in Wilamowitz’s book which certainly suggest that he had read Keil’s, e.g. ‘Androtion hat dem Aristoteles vermutlich sehr viel mehr geliefert als wir noch beweisen ko¨nnen’ (AuA I 305); cf. ‘Ich glaube, dass Aristoteles noch viel mehr, als wir nachweisen ko¨nnen, der Atthis des Androtion verdankt’ (SVAVA 191). 53 Note the interesting comment of Jacoby (ii 99 n. 127): Wilamowitz ‘did not work up ab integro the remains of Atthidography, because he was not much interested in the known Atthidographers as compared with the alleged Chronicle of the Exegetai.’ 54 GdA III2 600 n. 1 (on p. 601) ¼ II1 650 (end of § 405). And see Jac. ii 132 n. 2: although Keil was correct in assuming that Hermippus used Androtion directly, ‘he 51 52

Aristotle and the Atthidographers

297

tion of Wilamowitz’s great work, the theory was endorsed by Busolt, who accepted it without hesitation in the second edition of his Griechische Geschichte, Vol. II of which appeared in 1895.55 In 1920, in his Griechische Staatskunde,56 Busolt could state the theory, with a brief reference to Keil and Wilamowitz, as an accepted fact which scarcely needed proof. Meanwhile Adcock in 1912 had re-stated Keil’s theory, with some additional details,57 and the only scholar to attack it, Artur Ledl,58 had accepted most of Keil’s premises and agreed with him that Androtion was one of Aristotle’s sources; the only conclusion he disputed—on very weak grounds59—was that the common source of the Ath. Pol. and Plutarch’s Solon was Androtion. Keil’s theory, as stated by himself and later by Busolt, Adcock and others, depends entirely upon the assumption that neither Plutarch nor his ‘source’ (commonly taken, as by Keil, to be Hermippus the Callimachean,60 who wrote about 200 B.C.) knew the Ath. Pol, even indirectly, and therefore that the material which is common to the Ath. Pol. and Plutarch’s Solon must come from a source used both by Aristotle and by Hermippus–Plutarch; and in the absence of any other obvious candidate it is then assumed that this source ‘must have been’ Androtion.61 All this I would strongly challenge. I cannot undertake here a detailed refutation of Keil’s theory, and indeed, like many unfounded theories, I do not see how it can be actually proved wrong, except in certain particulars. failed to prove the assertion that he did not use as a source the ’Aup.’ But this assertion was a vital part of Keil’s theory. 55 See GG II2 32–45, esp. 41 n. 2 (on pp. 41–3), a long note which, taking Keil’s premises for granted, summarises the most important of Keil’s conclusions. 56 GS I 91–6, esp. 94: ‘Fu¨r die Bestimmung der Quelle ist es von wesentlicher Bedeutung dass Aristoteles, zwar nicht ausschliesslich, aber doch vorwiegend, die Atthis Androtions, die damals neueste Bearbeitung der Chronik, benutzt hat.’ 57 F. E. Adcock, ‘The Source of the Solonian Chapters of the Athenaion Politeia’, in Klio 12 (1912) 1–16. 58 SAAV 14–17. I cannot think of any other work to which Bloch might have been referring in the remark quoted in n. 63 on p. 285 above. 59 Jacoby (ii 99 n. 127) comments, ‘The results of Ledl’s ‘‘Quellenanalyse der ’Aup.’ . . . seem to me quite wrong.’ Cf. his commentary on Androt. F 3–4. 60 Jacoby, it is worth recalling, thought it reasonably certain that passages of a markedly learned character in Plut. Sol. (e.g. 19.3–4 & 25.2) come not from Hermippus but from Didymus who wrote a monograph on Solon’s Census: FGrH III B, Kommentar, on 340 Didymos, p. 92. 61 See esp. Keil, SVAVA 44, 49–50, 60, 97–100, 100–1, 155–7, 157–63, 172–3, 176–7, 185–9; Adcock, op. cit. (in n. 57 above) 14–16.

298

Athenian Democratic Origins

Some of the material common to the Ath. Pol. and the Life of Solon may have descended to Plutarch directly from the source of Aristotle, through Hermippus (who was certainly used by Plutarch) or another, and in some cases Aristotle’s source may be Androtion. But the theory has not been proved, as so many recent writers have innocently assumed, and it remains a conjecture, which should not be used to establish any conclusions about the sources of the Ath. Pol. or about Aristotle’s historical method. At the root of it lies what I believe to be a fallacious theory about Plutarch’s procedure in the composition of his Lives. The problem of the source of a Life by Plutarch is nearly always a difficult one, above all because there is no consistency in Plutarch’s methods.62 The situation would be less complicated if it were possible to accept a remark by Day and Chambers,63 that when a text names its sources ‘there is scarcely a problem’. This is far from the truth, especially in relation to Plutarch. Of course he did not read all the scores of writers he quotes by name.64 Many he will have read himself; some he is most unlikely to have seen; but about many others we simply do not know whether he had consulted them at first hand, for the purpose of the biography in question or generally, or whether he is repeating what an intermediate source said about them—although when he is not quoting directly he will often acknowledge the fact, as for example three times in the Life of Solon.65 He did, however, read very widely indeed: it is a great mistake to underestimate the scope of his literary research.66 ‘That the large majority of his quotations are 62 See the very sensible remarks of H. D. Westlake, at the end of his two articles on the sources of Plutarch’s Timoleon, in CQ 32 (1938) 65–74, and Pelopidas, in CQ 33 (1939) 11–22. Among other good analyses of Lives by Plutarch, see R. E. Smith, ‘The Sources of Plutarch’s Life of T. Flamininus’, in CQ 38 (1944) 89–95 [and see n. j on p. 327]. 63 AHAD 5. 64 See Ziegler’s Index I in the Teubner Plutarch, Vit. Parall. IV ii (1939). 65 ‘ l
egein wZsin ‘
ErmippoB Ey
’ anuZ t (a) vB on S
amion [‘As Euanthes of Samos ’ tfi Z& peri t vn says, according to Hermippos’] (11.2); (b) D
idymoB o‘ grammatik oB en & ’ onvn t vn ’ fi Z& pr oB t
eueike aj
& S
olvnoB antigraw oB ’AsklZpi
adZn Filokl
eoyB tin l
ejin ktl. [‘Didymos the grammarian in his ‘‘Refutation of Asklepiades Concerning the Axons of Solon’’ cites a passage of a certain Philokles etc.’] (the opening words of

‘ ErmippoB i‘ store&in wZsi P
1.1); (c) ta &yta men oyn ataikon [‘Hermippos says that this story is related by Pataikos’] (6.7). 66 An extreme example of this tendency is the article by J. E. Powell on the sources of Plutarch’s Alexander, in JHS 59 (1939) 229–40, the conclusion of which is that Plutarch used only two sources: a collection of Alexander’s Letters and ‘a

Aristotle and the Atthidographers

299

the result of his own reading should never have been doubted’, as Gomme says in the course of an admirable 30-page discussion devoted to Plutarch in the first volume of his commentary on Thucydides,67 a welcome corrective to some of the exaggerated estimates of the extent to which Plutarch was willing to rely upon only one or two sources for a given Life. But we must not expect Plutarch to remember accurately all that he read, and when he ‘ignores’ a passage we think relevant we certainly must not jump to the conclusion that he had not read the work in which it occurs. He may have read it, and forgotten—or omitted to make a note of—the passage in question, or he may have read it only in part, or for some other purpose, and failed to assimilate the relevant portion. Ancient literary manuscripts did not have the conveniences we nowadays take for granted, such as numbered pages and indexes; and finding one’s material in a large work—a roll, or series of rolls, at that—must often have been a very difficult process, which would have discouraged many a writer from consulting his authorities again and again. Many of the sources Plutarch used, or could have used, would have afforded material for several different Lives, as well as some of the other subjects Plutarch wrote about. He would hardly re-read each of them for each Life: some he might read and take notes from once only—the Ath. Pol., conceivably, which he refers to elsewhere;68 but I dare say he did not use it directly for his Solon. The considerations I have just advanced militate against several of Keil’s arguments, and Jacoby has already refuted some others.69 After quoting ‘the most decided pronouncements about the use of Aristotle made of A[ndrotion]’, from Keil, Busolt and Bloch, Jacoby concludes with a remark I have had occasion to quote already: ‘The general assumption of the importance of A[ndrotion] large variorum compilation on the history of Alexander, the same compilation of which Arrian’s Anabasis is principally a judicious epitome’. This judgement of Arr., Anab., reveals the absurdity of the whole theory, on which see W. W. Tarn, Alexander the Great II (1950), 306–9. On how Plutarch supplemented the meagre library resources available to him in his home town of Chaeroneia, about which he complains in Dem. 2.1–2, see the sensible remarks of Gomme, HCT I 77 ff. 67

HCT I 54–84, at p. 75. As Gomme says (HCT I 63 n. 2, with 76), Plut., Arist. 22.1, even if read with Arist. 1, ‘does not prove that he had not read Aristotle’s Constitution of Athens, which he quotes elsewhere more than once, only that he did not read it again before writing this Life’. 69 See Jac. i 114–17, with ii 109–11 nn. 35, 40, 41, 45; and cf. pp. 294–5 above. 68

300

Athenian Democratic Origins

for Aristotle has actually never been proved, and a systematic examination of the connexion between the two writers has never been made. It would require a full analysis of the historical part of the ’Aup.’70 I shall conclude this brief review of Keil’s theory with just one illustration of the way the theory crumbles as soon as its assumptions are challenged. One pair of parallel passages behind which Keil himself71 wished to see a single Atthis, that of Androtion, is that containing the stories about ‘Solon’s friends’, who profited (as did Solon himself, in one story) from the cancellation of debts: Ath. Pol. 6.2–4 (the palai
oploytoi [‘men of ancestral wealth’]) and Plut., Sol. 15.7–9 (the xrevkop
idai [‘debt-choppers’]). Wilamowitz72 realised that Androtion, at any rate, could not be the common source of these stories, because Androtion altogether denied that the Seisachtheia was a cancellation of debts, treating it as a mere currency wangle. To this Adcock73 triumphantly replied with what he thought to be additional proof of the attribution to Androtion: that what he considered the un-Aristotelian74 phrase oi‘ boyl
omenoi blaswZme&in [‘those wishing to slander him’], applied in Ath. Pol. 6.2 to those responsible for the version hostile to Solon, must come directly from Androtion, who represented the attacks on Solon to be groundless slanders—as of course they would be if there had been no cancellation of debts. Androtion’s account of the Seisachtheia, says Adcock, was an integral part of the ‘common source’, which (as reconstructed by Keil and himself) contained nothing that could not have stood in Androtion’s Atthis. Moreover, ‘Aristotle himself was neither partisan nor democratic’, whereas the incident is described in Ath. Pol. 6.2 from the ‘distinctly partisan and democratic standpoint’ which Adcock, following Wilamowitz, believed to be characteristic of ‘the Atthis’.75 70 Jac. ii 100 n. 127 (cf. pp. 285–6 above). Jacoby continues, ‘which work, as I may mention in passing, is in urgent need of a new commentary’. [No longer: see p. 325, nn. a and b]. 71 72 SVAVA 46–50. AuA I 63 n. 34. 73 Op. cit. (in n. 57 above) 15–16. 74 ‘The word blaswZme&in itself in this sense is never found in Aristotle, who always uses diab
allein’ (ibid. 16). But [Ste. Croix’s objection is lost. But he might have pointed out that the author used blaswZme&in to avoid repetition of diab
allein: on variation in the Ath. Pol. see Rhodes, CAAP 42–3.] 75 Op. cit. 15–16, cf. 4, 7, 8, 12, 13; and in CAH IV 39.

Aristotle and the Atthidographers

301

This defence of Keil’s theory will not do at all. First, the confused notion of an ‘Atthis’, characterised by a single standpoint, a ‘democratic’ one, has now been destroyed by Jacoby,76 who has taught us to think in terms of ‘Atthides’ (in the plural) having different standpoints, all in a loose sense ‘democratic’, but some of them decidedly ‘conservative’ and unsympathetic to radical democracy. As Jacoby says,77 ‘the majority of the authors of Atthides belonged to the conservative party: Androtion, Phanodemos, and Philochorus were more or less ‘‘reactionaries’’.’ Secondly, Adcock failed to notice that neither in Aristotle nor even in Plutarch (who does mention Androtion’s conception of the Seisachtheia) is there a trace of any such argument as Androtion must have employed: to the slander that Solon had made a profit for himself out of his cancellation of debts the reply in each case is not (as Androtion would have said) that there never was any cancellation of debts at all, but merely that although Solon’s friends did make such a profit, Solon himself did not. There are at least two sources behind Aristotle’s account and at least two more behind Plutarch’s,78 and none of these can be Androtion— although of course Androtion is likely to have had something to say on this topic, and may conceivably have used some such phrase as oi‘ boyl
omenoi blaswZme&in [‘those wishing to slander him’] when explaining how baseless all such slanders must be, because debts were not cancelled.79 But all this is speculation. 4. My next task is to explain why I believe that as regards the facts of Athenian constitutional history we should give greater weight to the statements of Aristotle than to those of any early Atthidographer, especially where there is a real likelihood that 76 A 71–9, 123–4, 129 ff., etc. Cf. Jac. ii 93 n. 86 (where Androtion is described as ‘a moderate conservative’, with a political outlook closely resembling Aristotle’s); also i 96, where Jacoby says ‘there can be no doubt as to his [Androtion’s] fundamentally conservative attitude: in this respect the Atthis speaks a clear language.’ Actually, the two-and-a-half-page note appended to this statement (ii 91–3 n. 86), beginning ‘I must admit that the fragments do not yield very much on this point’, shows that the language is not as clear as could be desired; but I think there can be no doubt that Jacoby is broadly right. 77 A 79. 78 One of Plutarch’s other sources is of course Polyzelus, and the other is the source which gave the figure of five talents (Sol. 15.9). 79 Cf. Jac. ii 132 n. 3, who warns against taking Androtion to be one of the tineB [‘certain people’] or oi‘ boyl
omenoi blaswZme&in [‘those wishing to slander him’] of Ath. Pol. 6.2.

302

Athenian Democratic Origins

documents may have been available. I say ‘the facts of constitutional history’ deliberately, to exclude in particular three things: first, the merely anecdotal material of which there is a good deal in the Ath. Pol. (as there evidently was in the Atthidographers); secondly, the attribution of motives to statesmen who appear on the political stage; and thirdly, the political judgements in general, which, as elsewhere in Aristotle, are a very mixed lot, some shrewd and sensible remarks lying alongside conventional superficialities. Of course political judgements cannot be entirely separated from the exposition of the facts of constitutional history, and sometimes they are indistinguishable; but sometimes it is possible and useful to draw the distinction. I can divide up what I have to say under three headings. A. Aristotle was above all a philosopher, yes. But he was also— we must never forget this—the man who interested himself in the constitutions of a very large number of Greek states, to the extent of collecting information about them, analysing it, and writing accounts of those constitutions which evidently contained (in at least some cases, if not all) historical material, in addition to a systematic treatment of the constitutions as they existed in his day. He also wrote a description of some barbarian customs, and he began a collection of laws which Theophrastus eventually published. Even more important than all this, for our present purposes, is the fact that Aristotle undertook the laborious and most unphilosophical task of investigating the records at Delphi and Olympia and Athens, and compiling lists of victors in athletic and artistic competitions80—this is particularly significant for us, precisely because it involved research into official records, and shows that Aristotle had a positive taste for this kind of activity, which he indeed seems to have initiated. Jacoby tends to set up a contrast between Aristotle’s historical ‘facts’ (including dates), which he thinks he would be likely to take over from Androtion, and his ‘judgements’, which would always be his own, even if, as Jacoby believes, they often happened to coincide with Androtion’s.81 To quote only two remarks of Jacoby’s—‘Even if he [Aristotle] had taken over all the facts (which 80

See p. 270 above. See Jacoby’s commentary on the fragments of Androtion, esp. i 103–4, 160; ii 99 n. 124; 99–102 n. 127. 81

Aristotle and the Atthidographers

303

he certainly did not) he did his own thinking and judging’;82 and ‘I do not believe A[ndrotion] to have ‘‘furnished’’ Aristotle with any political judgement: the philosopher followed the historian as to the facts, i.e. he took his Atthidographic material largely from A[ndrotion] because he sympathised with his political judgement, and because this judgement to a great extent agreed with his own.’83 But if a historian does not get his ‘facts’ right, no amount of what I might call ‘armchair activity’, such as ‘thinking and judging’, will be of any use to him, and may indeed result in misleading speculation. Jacoby,84 stressing the fact that Craterus and the Periegetai, unlike the Atthidographers, gave documents verbatim, mentions the generally received view that ‘the actual research among documents, by means of which the wording of them is brought into literature, was carried on in the Peripatos by the disciples and helpers of Aristotle, whose interest was primarily turned towards the polite
ia, the legislation, and the administration of a State. Subsequently, and without this restriction, the periegetai of the third century investigated the documents, Diodorus, Polemon (who was called stZlok
opaB [‘stele-mad’]), Heliodorus; so did the notabilities of antiquarian literature.’ But was it only ‘the disciples and helpers of Aristotle’, and not the master himself, who conducted this ‘research among documents’? And did they do it only after his death? Surely not. There is no evidence that any Greek before Aristotle had ever done any systematic research into archives. Here I agree entirely with Jacoby: although we may assume that the native Atthidographers ‘enlarged the account of Hellanikos by research among the documents’, we must beware of overestimating this side of their activity.85 ‘A systematic investigation of documents did not begin (roughly speaking) until after 350 B.C. and this was too late for Kleidemos and Androtion (perhaps even Phanodemos), who may be considered as sources of Aristotle, to make use of the results, even if they had wished to do so.’86 For the past, Jacoby rightly says,87 the Atthidographers ‘used their predecessors, i.e. the first Atthis of Hellanikos and the great historians who, from Hekataios onward, treated Athens more or less fully; and the later Atthidographers used the earlier: . . . Hellanikos used Herodotos; the 82 85

Jac. ii 100 n. 127. A 215, with 388 n. 65.

83

84 Jac. ii 101–2 n. 127. A 209. 86 87 A 198. A 202.

304

Athenian Democratic Origins

fourth-century Atthidographers took their form wholly and their matter largely from Hellanikos.’ And again,88 the Atthidographers ‘had no bias at all towards founding their narrative on documents.’ When passing beyond their own experience ‘they write on the basis of predecessors’ narratives, and where such a narrative does not exist, on the basis of the general conception accepted in their circles about the development of the Attic State. They never felt the need to correct this general conception by research in the documents . . . They evidently did not believe that the picture as it had been handed down would be changed in essential features if they consulted documents other than those readily accessible to all.’ Indeed, they ‘did not search systematically for the documents; they attached no importance to them, because they wrote the political history of Athens, and the conception of documentary writing of history had not occurred to the Atthidographers.’89 As for Androtion in particular, ‘We . . . cannot state with certainty whether A[ndrotion] made any research among documents at all. If he did, we must not overrate that side of his work . . . he had little need of documents for his political activities, and when he began to write the documents were not at his disposal.’90 B. This brings me to my second point, a very obvious one which, apart from the remark I have just quoted from Jacoby, has been almost entirely ignored: Androtion wrote in exile.91 Unless we are prepared arbitrarily to set this aside as false, or say that Androtion ‘must have’ begun to prepare for his task before he left Athens, we are driven to admit that he will have had little opportunity of doing any independent research, and none of consulting archives,92 and will have had to rely almost entirely on earlier literary sources, supplemented no doubt by oral information, par88

A 204–5. A 209. Hignett (HAC 17) is accordingly wide of the mark when he says that to conceive of Aristotle’s using ‘an official copy of Solon’s laws which previous writers had not troubled to consult’ is ‘to rate the industry of the earlier Atthidographers very low’. 90 Jac. ii 97 n. 110. 91 Plut., Mor. 605C (De Exilio 14) ¼ Androt. T 14. 92 Jacoby (A 91–2) says, ‘Androtion must have used the documents frequently’ [sc. ‘the documents about the elections and the reports about payments made to the strategoi’, which he has just mentioned]. But contrast the statements quoted above. I think Jacoby (A 91) is inferring far too much from Androt. F 38. The details may be exceptionally full in this case because, e.g., among the generals was a relative of Androtion, or of his authority. 89

Aristotle and the Atthidographers

305

ticularly that handed down in his own family—a source of knowledge by no means to be despised, at any rate for recent events. Androtion’s father Andron, and he himself, apparently played an active part in Athenian political life in the late fifth and fourth centuries, and for that period (and perhaps a little further back still, if Andron was not the first member of his family to enter political life) Androtion might reasonably be credited with special knowledge; but for earlier times his value as a historical source will have been no greater than that of his predecessors—Herodotus, Thucydides, Hellanicus, Cleidemus. He was evidently not an experienced historian already, when he came to write his Atthis: there is no trace of his ever having written any other work of a historical nature.93 Of the 68 fragments of Androtion accepted by Jacoby, all but nine or ten94 appear to relate to the second half of the fifth century or to the fourth, and the fragments dealing with earlier times are a very unimpressive collection indeed. I wonder how many of those who glibly repeat that Androtion is ‘of course’ the main source of Aristotle have examined these fragments. The three daughters of Cecrops; the foundation of the Panathenaea by Erichthonius (repeated from Hellanicus); two fragments about the judicial functions of the Areopagus, not very illuminating; the misrepresentation of Solon’s Seisachtheia; a mention of the battle of Pallene; F 5 on the Apodektai, a bad mistake unless Harpocration is in error;95 perhaps something about the foundation of Ennea Hodoi, which could have been taken from Thucydides—that accounts for eight of the ten fragments. The only ones of any real interest are F 6, on ostracism, and F 36, containing the quotation from an early law mentioning the naucraric fund, which I have already discussed. From about the mid-fifth century onwards the situation changes entirely; but I see no valid reason for supposing that the first two books ‘must have’ provided some good historical material, simply because the last six apparently did so. C. The second, the systematic, part of the Ath. Pol. (ch. 42 ff.) is based upon considerable research at first hand among documents by Aristotle, with the assistance at any rate of Theophrastus and probably other pupils as well: this is evident, even apart from the 93 94

On Androt. F 70–1, see Jac. i 171; and on F 75–82, see Jac. i 108 (on T 17). 95 F 1–6, 34–6, perhaps 33. See p. 289 above.

306

Athenian Democratic Origins

external evidence,96 and is now, I think, universally admitted,97 against the old view of Wilamowitz98 that this part of the Ath. Pol. depends upon an earlier Atthis. Accepting the conclusions of Bloch, Jacoby99 agrees that the second part of the Ath. Pol. ‘could not have been written without systematic investigation of archives and documents. After Aristotle had given the impetus and probably begun the work, this investigation was carried on by him and his disciples in common. The material collected in this way . . . provided the foundation of sources of Aristotle’s survey.’ I cannot understand why we should be invited to make a complete separation between the two parts of the Ath. Pol., which together form a single work. At many points in the second part— 45.1, 3; 49.3; 51.3; 53.1; 54.3, 7; 55.1, 2, 4; 56.3, 4; 60.2; 61.1; 62.1—Aristotle refers not merely to the situation existing at the time he is writing, but to an earlier state of affairs, about which also he had surely in many cases obtained his information at first hand. I cannot see why we should not give Aristotle the credit for making every effort to obtain his material for the constitutional history of Athens at first hand, from official documents. (How far he will have been able to succeed in this is quite another matter: I shall deal with this question in Section (iv) of this essay.) Even Jacoby, who says it is ‘agreed that the material for this historical part [of the Ath. Pol.] was taken mainly (if not entirely) from historical sources: Herodotos, Thukydides, and the Atthidographers’, also admits, a sentence earlier, ‘It is of course possible that even the introduction was the result of the research of Aristotle and his disciples, and in a number of polite&iai (which probably were all designed on the same model) this may have been the case.’ *** I hope I have now made it clear why I regard the choice between Aristotle and Androtion as the origin of much of our information about Athenian constitutional history as a significant one. To say that the source of A is B is often no more than to replace one literary authority by another, if not to substitute for a well96

See pp. 266–70 above. It will be sufficient to refer here to the admirable Introduction to the Bude´ edition of the Ath. Pol. (the second part, by B. Haussoullier), pp. xiv–xxx, and to Bloch (SHLFC 355 ff., esp. 367–76) and Jacoby (esp. A 210–11). 98 99 AuA I 214–16, 242–3, 256–9. See n. 97 above. 97

Aristotle and the Documentary Sources

307

known writer some shadowy figure of whom little or nothing is known. It is not so here. Unless Plutarch’s statement about Androtion’s having written in exile is false (and there is no reason to distrust it), Androtion had no opportunity at all of conducting documentary research; and therefore the value as historical evidence of the constitutional material in the Ath. Pol. is small if it was merely copied from Androtion without verification—except on one condition: that Androtion himself copied it from an earlier writer who did use the documents. It is very unlikely, however, as we have seen, that any earlier writer did do documentary research. On the other hand, we know that Aristotle and his school did precisely this, perhaps for the first time in human history, even if we cannot prove that such research on their part lies behind the historical portion of the Ath. Pol. Even if Aristotle, for his narrative, used the Atthidographers freely—as he surely did, especially for those stories in which they seem to have delighted—is it not probable that his research rather than theirs is the source of many of his statements of constitutional facts, and that even when he took such things from them, he verified the material at first hand, in so far as this was possible? After all, if they could unearth material from the records, Aristotle and his pupils could do it too. But the essential point is that we know he did go in for this kind of activity, whereas there is not a scrap of evidence that anyone before him had ever done so.

(IV) ARISTOTLE AND THE DOCUMENTARY SOURCES Finally, having established the probability that Aristotle was more likely than any of his predecessors to want to make use of documentary sources, I come to the question whether he had such sources available. 1. It is necessary to begin by following Jacoby1 in drawing a firm distinction between the sixth century and the fifth, the dividing line coming with the reforms of Cleisthenes. ‘There is an obvious difference in the two sections of the historical part of the ’AuZna
ivn polite
ia: from Solon to the overthrow of tyranny (chs. 5–20), and from Kleisthenes to . . . 401/0 B.C. (chs. 21–40). There 1

A 206–7.

308

Athenian Democratic Origins

is hardly a documentary date in the former that could not be read in the archons’ list or calculated with the help of oral tradition;2 in the latter there is an abundance of dates, references to, and occasionally detailed statements of the contents of, documents.’ (Jacoby’s sentence continues, ‘which therefore the Atthidographers used by Aristotle must have known’; but I do not believe any such wholesale inference is justified, since, as we have seen,3 we cannot be sure that the Atthidographers did much in the way of documentary research.) It seems very likely that ‘when the constitution of Kleisthenes made the Council instead of the archon the supreme executive in Athens’, as Jacoby puts it,4 ‘ ‘‘records of the Council’’ . . . began to exist, in whatever form.’ 2. I need not enter here in detail into the difficult problem of the Athenian state archives. The evidence was reviewed at length by Kahrstedt in 1938,5 and his conclusions have since been discussed approvingly by Jacoby6 and Harrison,7 with whose position in general I agree. It was almost certainly not until c. 403/2 that what we may call a State Record Office began to exist in the Metroo¨n. Before that date we cannot demonstrate, or even assume, the existence of anything like a unified Record Office, at the Metroo¨n or anywhere else. So far we may agree with Kahrstedt. On the other hand, as Jacoby has pointed out in an admirable note in his Atthis,8 the administration of Athens from 508/7 onwards, and still more that of the Empire when it came into being, ‘could not have been achieved without records (which may be distinguished, as the more general concept, from documents proper), and the records had to be kept somewhere. Actually the place is of no importance. But if there was no archive of the State, each official (certainly as early as the sixth century) had his own archive and there were also archives belonging to the administrative boards . . . ; also to public and private associations, phratries, 2 Jacoby (A 169) believed, I think rightly, that the motion of Aristion (Arist., Ath. Pol. 14.1) ‘was almost certainly preserved in oral tradition’. 3 4 Pp. 303–5, 307 above. A 206. 5 NLA. Mrs K. M. T. Atkinson, ‘Athenian Legislative Procedure and Revision of Laws’, in Bull. Jn. Rylands Libry. 23 (1939) 107–50 (cf. Gomme in CR 54 (1940) 38 and Kahrstedt in Gnomon 16 (1940) 377–9), argues against many of Kahrstedt’s conclusions, but has nothing to say that is really relevant to the question with which we are now concerned. 6 A 383–4 n. 27. 7 8 LMAEFC 27–9 [and n. k on p. 327 below]. A 383–4 n. 27.

Aristotle and the Documentary Sources

309

phylai, demes,9 or clans. From the time when writing existed, people deposited matters concerning the various officials; . . . the individual pieces of business and the cases with which they were concerned. The contents of the Didaskal
iai [‘‘Catalogues of Dramas’’] e.g. come from the archive of the archon . . . From the time when Kleisthenes made the Council the supreme executive the archive of the Council was that of the greatest importance, and the contents of it probably were simply transferred to the Metroon in 403/2 B.C.’ We have of course no means of telling how much accurate information was contained in these various records, or how far they would have survived the Persian sack in 480, or how easily available they would have been in the fourth century. But we certainly cannot afford to neglect this kind of documentary material as a possible source for the history of Athens, at any rate from the time of Cleisthenes onwards: the proof is in that very difference emphasised by Jacoby between the narrative in the Ath. Pol. of the periods before and after Cleisthenes. It is also worth noticing that Craterus probably devoted at least two whole books of his collection of Athenian decrees to those which were passed in the first half of the fifth century (or earlier).10 Craterus probably did

9 It is worth recalling here that in the mid-340s the demarch of Halimous apparently said he had lost the deme register, the vital lZjiarxik on grammate&ion: Dem. LVII 60. 10 At least 7 of the 21 fragments we have of Craterus (FGrH III B 342 F 1–3 and 5–8; cf. 19–20) probably come from assessment decrees or quota lists of the Athenian Empire (probably the former: see Jacoby, FGrH III b i 96, with ii 64 n. 33). F 1, from Book III, is the important one for our purposes: it mentions the Carian tribute district, which disappeared probably in 438/7, and should therefore come from a document earlier in date than that. The authors of ATL (I 203–4; III 9–12; and see I 483) are certain that it comes from the assessment decree of 454/3 (A 1 in ATL II 40); but (a) their argument that F 4, from Book IV, ‘may with virtual certainty be attributed to Perikles’ law of citizenship’ of 451/0 (ATL III 10) is valueless (cf. Jacoby, FGrH III b. i 101–2); (b) Doros is not certainly the Phoenician town (see Jacoby, ibid. i 100); and (c) their view entails the supposition that a Carian district was named in A 1, although the quota lists do not specify geographical districts until 443/2 (List 12, ATL II 18), and there is no particular reason for supposing that assessment decrees specified the districts much earlier than this. Jacoby (ibid. ii 65 n. 42) suggests 446/5 as the date of creation of the Carian district. In that case F 1 will come from the assessment decree of 446 or that of 443. See also Gomme, HCT I 371 n. 2. It remains possible that the authors of ATL are right, and that Craterus concentrated on the first and last assessment decrees, so that Book III, from which F 1 comes, included the year 454/3.

310

Athenian Democratic Origins

not copy all his decrees from stones but in at least some cases used the archives.11 3. For the period before Cleisthenes there is just one very special set of documents the possible use of which by Aristotle (and his predecessors) we must certainly discuss: I mean of course ‘the laws of Solon’. False conclusions have often been drawn from a failure to understand just what these ‘laws’ contained, and in particular how far Solon’s constitution, his polite
ia, was contained in his n
omoi, his laws. On the one hand, there are those who think that because Aristotle undoubtedly (as they assume) read ‘the laws of Solon’, therefore he must be right about such matters as the sortition of magistrates in Solon’s constitution (Ath. Pol. 8.1) and the qualifications of the census classes (7.3–4), about which he would have derived information from the laws. At the opposite extreme are those who cannot accept the sortition of archons in the early sixth century at any price, and who infer from the way Aristotle supports his statements on that subject and also about the qualification of the Hippeis in Ath. Pol. 7.4 that he did not have documentary information on these points; and who go on to conclude that Aristotle could not have had Solon’s laws before him at all. Neither position is justified, as we shall see. Another group of scholars has recently emerged which denies that Solon set up a new constitution: they admit he made laws (n
omoi),12 but believe the Solonian constitution (polite
ia) to be a later invention. This view too can be refuted. We have two closely allied problems to settle here: (1) how far did Solon create a constitution, a polite
ia, as distinct from passing laws to have effect within an existing constitution; and (2) if he was indeed responsible for a constitution, how far would it have been defined by his actual n
omoi? The best way to approach these questions is to make an analysis of the relevant uses of the words polite
ia and n
omoi in the fifth and fourth centuries. Let us begin at the end of the process, with Aristotle himself. In reading texts such as the Politics and the Ath. Pol. it is sometimes necessary to recognise a distinction between polite
ia and 11 Here I agree with Jacoby, GH 165 ¼ RE XI (1922) 1618; FGrH III b i 96 (and see ii 64 n. 26). 12 Some have gone so far as to say that Solon made no new n
omoi, but merely wrote down existing laws. This view is certainly mistaken: see below.

Aristotle and the Documentary Sources

311

n
omoi;13 but at other times we find the word polite
ia used in the broadest possible sense, to include what we should call the constitution, and any fundamental rule of the constitution can be called a n
omoB. There is no consistency in Aristotle’s usage. This can be illustrated from the Ath. Pol. In 7.1 (cf. 6.1; 10.1) ’
Aristotle says of Solon, polite
ian kat
estZse kai n
omoyB euZken ’ ’
lloyB14 . . . anagr
acanteB de toyB n
omoyB ei’ B toyB k
yrbeiB a ’ tfi Z& stoa&fi tfi Z& basile
ifiv kai
v’ mosan xr
Zsesuai e’
stZsan en p
anteB [‘he established the polite
ia and passed other n
omoi . . . and writing up the n
omoi onto the kyrbeis they set them up in the Stoa of the Basileus and everyone swore to observe them’]. If we take Aristotle’s words literally and interpret polite
ia in the narrow sense, it is the n
omoi which were written on the Kyrbeis, not the polite
ia. In the Politics also Aristotle makes a distinction between polite
ia and n
omoi, specifically with regard to Solon in II 1273b 32–4, where we are told that ’ enonto dZmioyrgoi m
among lawgivers oi‘ men n
omvn eg
onon, oi‘ de ffl kai polite
iaB, oiffl on kai Lyko &yrgoB kai S
olvn: oytoi g ar kai omoi n
omoyB kai polite
iaB kat
estZsan [‘some were authors of n
only, while others were authors of a polite
ia also, such as Lykourgos and Solon: for they established both n
omoi and polite&iai’]. In omoi alone, Pol. II 1274b 15–16,15 Aristotle credits Draco with n
‘ adding polite
ifia d’ yparxo
ysfi Z toyB n
omoyB e’
uZken [‘he added his n
omoi to an existing polite
ia’]. And see Pol. II 1265a 1–2; III 1286a 2–3; IV 1298a 17–18; and in particular IV 1289a 13–22, which it will be well to summarise. Laws, we are told, ought to be made, and are made, to fit polite&iai, and not polite&iai laws. A polite
ia is then defined as an organisation of the offices in states ‘ peri t ’ aB), prescribing how they are (a t
ajiB ta&iB p
olesin Z aB arx
to be distributed, what is to be the sovereign element of the consti‘ astZB tZB ’ in [‘and what is & koinvn
iaB est
tution, kai t
i t o t
eloB ek
the end of each community’].16 The definition of laws follows: 13 This was well understood in the 1890s (see e.g. Busolt, GG II2 20 n. 1, 32 n. 1), but has sometimes been forgotten in more recent times. 14 Day and Chambers, AHAD 79 n. 39, suggest that the reference to ‘other’ laws means ‘in addition to those concerning the constitution’; but the word could also have been inserted because of n
omoyB e
’ uZke [‘he passed laws’] in 6.1. 15 Cf. lines 18–19 on Pittacus. 16 ’
’ vn Cf. III 1278b 8–10: e
’ sti de polite
ia p
olevB t
ajiB t vn & te allvn arx & kai & kyr
iaB p
m
alista tZB antvn [‘A polite
ia is the organisation of the various offices of a state, and especially of the most important one of all’]. For Aristotle, when he

312

Athenian Democratic Origins

n
omoi d’ oi‘ kexvrism
enoi tvn & dZlo
yntvn t Zn polite
ian, kau’ o‘yB ’
’
de&i toyB arxontaB arxein kai wyl
attein toyB paraba
inontaB ’ yB [‘laws distinct from those which define the constitution, ayto
according to which the magistrates should exercise their powers and check those who transgress them’].17 Here laws proper, we might say, or laws in the narrow sense, are distinguished from those laws which define the constitution: n
omoi and polite
ia are different. On the other hand, the expression n
omoi (with nomoues
ia, nomou
etZB and related words) can often be used in a more general sense, to include the whole framework of the constitution. This can be illustrated from the Ath. Pol. In 22.1 ostracism is created by Cleisthenes by a n
omoB. In 7.1, as we saw a moment ago, Solon’s polite
ia is distinguished from his n
omoi; and in 9.1 the tr
ia dZmotik
vtata [‘three elements most favourable to the demos’] are features of his polite
ia; yet 10.1 calls these three rules the omoi and dZmotik
a [‘that favour the demos’ feature’] among his n
refers to his constitution-making as his nomoues
ia [‘legislation’], ’ and 8.2 uses the word enomou
etZsen of the rule regarding the election of magistrates contained in 8.1. Undoubtedly, it is this usage which is the older one: as far as I can discover, there is no trace of any explicit distinction between polite
ia and n
omoi until well into the fourth century, when the writers on political theory required a more elaborate terminology. Certainly in the fifth century it was the usual practice to refer to the constitution existing in a state, or devised by a legislator, as its or his n
omoi. It is necessary to make this clear, because a failure to realise it has led more than one recent historian to some false conclusions about the reliability of our sources for early Athenian history in general and about the was writing the Politics (cf. pp. 258–60 above), Solon left the Areopagus and t
Zn ’ vn t vn & arx & a‘
iresin [‘appointment to offices’] as they were; but (in spite of Pol. II 1273b 41--74a 2) the introduction of the t
elZ [‘property classes’] alone, even apart from the creation of the Heliaea, would surely make Solon’s kat
astasiB [‘establishment’] a distinct polite
ia in Aristotle’s eyes, a polite
ia being to him above all a ’ vn t
ajiB t vn & arx & [‘organisation of offices’]. 17 Cf. Plato, Laws VI 751A, where there are said to be d
yo e’
idZ ta &yta peri ’ vn ’ ontvn, o‘
saB te polite
iaB k
osmon . . . , pr vton & men katast
aseiB arx & te kai arx
’ aB ei
nai de&i kai tr
Z toyB n
omoyB ta&iB ayt opon o‘
ntina kauistam
enaB: e
’ peita oy
‘ tv d ’ ‘ astaiB apodot
’ arxa& iB ek
eon ktl. [‘the following two branches concerning the organisation of the constitution . . . , first the appointment of magistracies and magistrates, how many of them there ought to be and elected in what manner; then the laws are to be assigned to each magistracy in such a way etc.’].

Aristotle and the Documentary Sources

313

nomoues
ia [‘law-giving’] of Solon in particular. In a recent study which is valuable for its collection of material from the orators concerning Theseus, Draco, Solon and Cleisthenes, Eberhard Ruschenbusch18 declares that down to 356 Solon is mentioned only as one of the Seven Wise Men, as poet or as lawgiver (Gesetzgeber), but ‘never yet as the maker of a constitution’ (Scho¨pfer einer Verfassung). He infers that this is not accidental, and that contrary to the opinion of modern scholars, the constitution of Athens was not regarded in the fifth century as the work of Solon. He ends his article by asking whether we should not now regard all accounts of early Athenian constitutional history as ‘reine Konstruktion’ [‘sheer construction’].19 Ruschenbusch here draws a distinction between polite
ia and n
omoi which is too rigid even for the fourth century and has no place at all in the fifth. The conclusive evidence against him can easily be displayed. Cleitophon referred in 411 to the p
atrioi n
omoi [‘ancestral laws’]20 by which Cleisthenes set up the democracy.21 These n
omoi can only be the constitution of Cleisthenes, who is never credited with any laws of a non-constitutional character. In Thuc. VIII 76.6, when the democrats in the fleet say in 411 that the men in the city have done away with the p
atrioi n
omoi, they mean their political constitution, their democracy. In the light of these passages we can see that several references to ‘the laws of Solon’ between the years 403 and 399 cannot be taken, as by Ruschenbusch, to exclude constitutional provisions: the decree of Teisamenus (And. I 83), and other passages in And. I (81, 82, cf. 95, 111) and Lys. XXX (2, 26, 28—especially the last, where Solon is bracketed as a nomou
etZB [‘law-giver’] with Themistocles and Pericles). The decree of Teisamenus (polite
yesuai toyB ’AuZna
ioyB kat & a t a p
atria, n
omoiB de xrZsuai to&iB S
olvnoB [‘the Athenians are to conduct their public affairs in the ancestral manner, and employ the laws of Solon’]) has been quite wrongly 18 PP, esp. 407–8, followed by Sealey, in Historia 9 (1960), at p. 159; and by Day and Chambers, AHAD 16–17, 72–3. The latter, however, seem to have misgivings: see their p. 16 n. 65. 19 Op. cit. 424. 20 On the usual meaning of this expression I agree with Jacoby (A 244–5 n. 46): ‘p
atrioB n
omoB is an entirely vague expression which only means ‘‘an old law’’, ‘‘a law from the time of our fathers’’ ’. 21 Arist., Ath. Pol. 29.3.

314

Athenian Democratic Origins

interpreted on this point by Fuks22 and Ruschenbusch.23 When we cease to take the opening clause in isolation, and read it with the rest of the decree, we can find yet another confirmation of the fact that the expression ‘the laws of Solon’ came to mean simply the corpus iuris Atheniensium [‘body of Athenian law’]:24 the word n
omoi occurs no fewer than six times in the remaining part of the decree, dealing with possible innovations, and here it would be absurd to pretend that it does not cover constitutional changes among others. The word polite
ia does not occur in the decree of Teisamenus, and there is no opposition between polite
yesuai & [‘conduct their public affairs’] and n
omoiB xrZsuai [‘employ the laws’]. The only reason why it was necessary to insert the polite
yesuai clause, I suggest, is that everyone would know that the Solonian n
omoi (that is to say, the laws of Solon himself and later amendments and additions) did not explicitly mention some fundamental rules of constitutional procedure which rested entirely or mainly upon custom.25 According to Antiphon (III a 1), o‘ n
omoB [‘the law’], surely in the sense of ‘customary law’ (note the singular case), and t a cZwism
ena, ‘what has been voted’, are & polite
iaB [‘have authority over the whole k
yrioi p
asZB tZB polite
ia’].26 Dracontides in 404 proposed a resolution that the Thirty should toyB patr
ioyB n
omoyB syggr
acoysi, kau’ o‘yB polite
ysoysi [‘write up the ancestral laws, according to which they would conduct their public affairs’].27 Describing the same meeting of the Assembly, Lysias (XII 73) could say in the ’ eleysen ym ‘ &aB . . . tfi Z& polite
ifia following year that Theramenes ek
‘ Drakont
idZB ap
’ ewainen [‘recommended that & xrZsuai Zn you . . . employ the constitution proposed by Drakontides’]. And when Xenophon (Hell. II iv 42) makes Thrasybulus tell the ’ & Athenians in 403 that they must to&iB n
omoiB to&iB arxa
ioiB xrZsuai [‘employ the ancient laws’], he was thinking above all of their democratic constitution. A polite
ia at this period is something omoi: the two words made up of n
omoi, or conducted according to n
22

23 AC 38–9. PP 408. See J. Schreiner, De corpore iuris Atheniensium (Diss. Bonn, 1913), esp. 48–50, at p. 49, endorsed by Harrison, LMAEFC 30 n. 34, and Fuks, AC 39. 25 Cf. pp. 315–17. 26 I cannot understand why Day and Chambers, AHAD 72 n. 24 (cf. n. 65) want to see here a hint of a ‘distinction between laws and a constitution’. 27 Xen., Hell. II iii 2. 24

Aristotle and the Documentary Sources

315

are never contrasted. It is very wrong to assume that Solon was not already credited with a large, if entirely undefined, proportion of the constitutional laws of Athens, even if the historical conception of a strictly Solonian polite
ia, as distinct from that of Cleisthenes, did not arise until much later28—the Athenians did not begin to think and enquire about the history of their constitution until later. An explicit distinction between polite
ia and n
omoi, as far as I know, only emerges in the writings of the fourth-century political theorists. This distinction, for all that it arose rather late, nevertheless did embody a truth which now becomes important for our present purposes. We have no right to expect to find the whole of Solon’s polite
ia in his n
omoi, and we must allow for the probability that ‘the laws of Solon’ which were known in the fourth century and later may not have included some of the provisions which actually set up the constitution in the first place. In those spheres in which Solon accepted the existing n
omoB (in the sense of customary law) and made no change in the situation, he might well make no specific enactment at all. More than that, much of what he laid down once for all, for example the Seisachtheia29 or the rules prescribing the method of election of magistrates or the qualifications of the census classes, may not have been included among his n
omoi, which were surely intended above all to govern daily life, and were probably often cast in the form of instructions to magistrates: ‘the Naukraroi are to exact’ or ‘to spend out of the naukraric fund’,30 or ‘the Kolakretai are to give money for travelling expenses out of the naukraric fund to the Theoroi going to Delphi’,31 or the rule (on Solon’s First Axon) that the archon must put a curse upon anyone exporting natural products other than olive oil, or pay a fine32—although of course many laws will have been cast in a more general form, as we find, for example, in Plut., Sol. 20–4; Lys. X 16–18. I think Jacoby33 is right, therefore, in saying Aristotle realised that Solon had not deposited his constitution in a special document: 28

Cf. Jac., i 29. According to Jac., i 121, the Seisachtheia ‘is known not to have appeared in the laws of Solon’. This of course is only an inference, if a compelling one, from the fact that Androtion could completely misrepresent the nature of the Seisachtheia (see p. 289 with n. 10 above). 30 31 32 Ath. Pol. 8.3. Androtion 324 F 36. Plut., Sol. 24.1. 33 A 333–4 n. 21 (with 333 n. 20); cf. Meyer, GdA III2 600 n. 1. 29

316

Athenian Democratic Origins

‘the constitution was not inscribed on the first axon, but must be inferred from the laws’; and this was not an easy task. ‘It was necessary to know the form of the State, for which Solon gave the laws; to know the alterations to which the institutions had been subjected in the course of time (mostly by psephisms . . . ); in particular to interpret the wording of the laws, which was always succinct and often vague; and a true understanding was possible only if one knew the existing Law, which Solon’s laws changed or superseded. It was inevitable to draw conclusions; we can recognize Aristotle’s procedure even in the most succinct summary of the main points in ’Aup. 7–9: it is evident everywhere that he had the laws before him, inferring from them the main lines of the ‘‘constitution’’.’34 There is reason to think that Aristotle’s inferences were by no means always right.35 In particular, as I show elsewhere in this book,36 I myself believe (without claiming to be able to demonstrate it beyond all possible doubt) that he was wrong about the introduction of sortition for the choice of magistrates 34 ’ Z Jacoby’s sentence ends, ‘and accordingly determining its nature as arx dZmokrat
iaB’. He is of course referring to Ath. Pol. 41.2 (cf. 9; 22.1), where Aristotle, mentioning the metabol
Z or met
astasiB [‘(constitutional) change’] ffl ’ i S
’ ’ Z which occurred ep olvnoB [‘in the time of Solon’], adds aw’ ZB arx ’ eneto [‘from which arose the beginning of democracy’]. But surely dZmokrat
iaB eg
it is a mistake (committed also by Day and Chambers, AHAD 67 etc.) to make ’ Z dZmokrat
iaB [‘beginning of demAristotle call the constitution of Solon the arx ocracy’], implying that it was itself a democracy. What Aristotle actually says is that from the constitution of Solon the beginnings of democracy arose—a rather different matter. In his curiously discursive remarks about Solon’s constitution in Pol. II 1273a 35–74a 21, Aristotle seems deliberately to refuse to ‘label’ that constitution, although he is certainly giving it his approval in 1274a 11–21. In 1273b 35 ff. we have merely what e
’ nioi men o’
iontai [‘certain people think’]; and the short sentence beginning e
’ oike de S
olvn [‘it seems that Solon’] (1273b 41 ff.) is continued by ’ v another beginning di o kai m
emwonta
i tineB ayt fi & [‘for which reason some people blame him’], starting a train of thought that goes on to 1274a 11. At 1274a 11 ff., however, he exculpates Solon from responsibility for the later developments which we know he disliked: these, he says, were not foreseen and planned by Solon but were chance developments. This seems to me to go very well with the description of ffl arx ’ ZB ’ Z dZmokrat
iaB eg
’ eneto Solon’s constitution in Ath. Pol. 41.2 as the one aw’ [‘from which arose the beginning of democracy’]. 35 See Meyer, GdA III2 600–1 n. 1: after saying that ‘authentisches Material’ was available, in the form of 1. Solon’s poems and 2. his ‘Gesetzbuch’, Meyer continues, ‘Dagegen 3. u¨ber die Solonische Verfassung besassen die Spa¨teren keine authentischen Nachrichten. Sie waren, wie Aristoteles pol. Ath. 8 mit den Verweisen auf sZme&ia und auf die Solonischen Rechtssatzungen zur Evidenz zeigt, lediglich auf die bestehenden Institutionen und, wo diese gea¨ndert waren, auf Ru¨ckschlu¨sse angewiesen, die nicht immer richtig ausgefallen sind.’ 36 Pp. 92–103 and 215–23 above.

Aristotle and the Documentary Sources

317

(Ath. Pol. 8.1) and about the qualifications of the census classes (7.4), though not about the fact that it was membership of these classes which determined access to office (7.3) or about the creation of a Council of Four Hundred (8.4). We can now answer, as far as the unsatisfactory nature of our sources permits, the two questions which made this examination of the terms polite
ia and n
omoi necessary. There is not the least reason to think that Solon did not create a polite
ia, but it will have been defined only partly in his n
omoi.37 From this two important results follow: even when we find reason to suppose that Aristotle did not possess the actual text of certain constitutional provisions of Solon, we need not conclude, as many have done,38 that he did not have before him the text of Solon’s laws in the narrower sense; and conversely, if we find reason to think that ‘the laws of Solon’ were available to Aristotle, this does not mean that he will necessarily have had the text of all, or even most, of the laws defining the constitutional framework. 4. A fourth-century writer might in theory derive his information about Solon’s laws either from the original physical objects, the much discussed Axones or Kyrbeis, or from copies. Only in the latter case need we consider whether the Solonian material was distinguishable from later accretions and alterations. I begin with the fact that the anonymous Vita Menagiana39 credits Aristotle with a treatise in five books, Peri tvn & S
olvnoB ’ onvn [‘On the Axones of Solon’]. Did this otherwise unattested aj
work in fact exist; and if so, was it by Aristotle? I do not see how we can be certain. A powerful argument against the existence of any such work by Aristotle was advanced long ago by Busolt:40 it would be very surprising that not the least fragment should survive of a work by Aristotle on such a scale and of such interest to antiquarians. The authenticity of the work, Busolt concluded, is extremely doubtful. Jacoby, after referring to the work as one ‘of which we 37 Bengtson (GG2 122) puts it very well: ‘Eine geschriebene Verfassung, ein ‘‘Verfassungsgrundgesetz’’, hat Solon nicht hinterlassen. Wohl aber griff seine Gesetzgebung bei dem Unvermo¨gen der Hellenen, zwischen Recht und Verfassung eine Trennungslinie zu ziehen, auf viele Gebiete der Staatsverfassung u¨ber.’ 38 e.g. most recently Day and Chambers, AHAD 79; cf. Weil, AH 125 (text to n. 192); Hignett, HAC 12–27. 39 V. Rose, Arist. Fragm. (1886) p. 16 line 140 (Hesych.). 40 GG II2 44 n. 1.

318

Athenian Democratic Origins

cannot simply doubt the existence’,41 later had second thoughts, for in the notes to his commentary on Androtion F 3–442 he says, ‘I should not like to affirm that the five books Peri tvn & S
olvnoB ’ onvn are genuine. All other books about the Axones (no. 339– aj
41)43 are from late Hellenistic or Roman times.’44 It is significant that Jacoby was nevertheless ‘confident that Aristotle read the Solonian laws himself’:45 it appears from another passage46 that he conceived Aristotle as doing so ‘either in the original or in the Ms. of Theophrastos’ N
omoi’—although he then47 shows he is not certain that ‘there were no ‘‘interpolations’’ (in the sense which the word bears in the textual criticism of the Digest) in the Axones’. He also feels obliged to ‘admit that Aristotle had no documentary tradition about the various measures of Solon, apart from the legislation proper: we do not know on what grounds he asserts so confidently that the seisachtheia preceded, or particularly that the reform of the currency followed, the legislation’.48 With these opinions I would concur. 5. I do not propose to enter into the controversy about the nature of the Axones and Kyrbeis—a controversy which received a new lease of life in 1879 on the publication of the first fragment to be discovered in modern times of the Athenian ‘Code’ of 410–399, and which has been stirred up again during the last thirty years by the finding of several more fragments of the Code.49 As Sterling Dow has said,50 ‘To reconcile all that the Classical authors wrote on the subject is impossible, because they contradict each other.’ He himself argues that the Axones contained secular law, and that the name Kyrbeis came to be applied to the sacred law.51 For this 41

A 385 n. 51. Weil, AH 125–7, has since concurred. Jac., ii 110 n. 39. 43 In FGrH III B. 44 Cf. pp. 319–20 below. 45 Jac., i 47. 46 Jac., ii 134 n. 8 (on Androt. F 36). 47 id. n. 9. 48 Jac., i 145. 49 A complete bibliography up to 1959 is given by S. Dow in Historia 9 (1960), at pp. 292–3. Add now Dow in Hesp. 30 (1961) 58–73 and Pl. 9–12; and in AJA 65 (1961) 349–56; A. R. W. Harrison in CQ 55 ¼ n.s. 11 (1961) 3–5 [and p. 326 below]. 50 ‘The Law Codes of Athens’, in Proc. Massachusetts Histl. Soc. 71 (1953–7, published 1959) 3–36 and Pl. I, at p. 28. There is a recent collection of most of the ancient literary evidence by L. B. Holland, ‘Axones’, in AJA 45 (1941) 346–62; but in my view Holland’s discussion does not make very profitable reading. 51 Op. cit. (in n. 50 above) 27 ff., esp. 28–31. 42

Aristotle and the Documentary Sources

319

there is some ancient authority;52 but other explanations have been offered which seem to me at least equally possible.53 I shall speak of the original physical objects as the ‘Axones’ and of the provisions they contained as Solon’s ‘laws’. Of the Axones, in Plutarch’s day (seven centuries after Solon), the Athenians still preserved in the Prytaneum le
icana mikr
a [‘slight remains’],54 a phrase which reminds me irresistibly of those fragments of wooden roof-beams, resembling honeycombs, sometimes placed in English church porches in support of an appeal for funds by woodrestorers, as a vivid demonstration of the ravages of death-watch beetle. By what stages this dilapidation had progressed, and for how long the Axones had remained legible, it is impossible to say. We have laws quoted by various authorities, with the numbers of the Axones (1, 5, 13, 16 and perhaps 21) on which they had appeared;55 but unfortunately there is nothing to tell us at what date the copies from the Axones were made. We hear of several accounts of the Axones by Hellenistic scholars: Apollodorus, Aristophanes of Byzantium, Asclepiades of Nicaea, Didymus, Eratosthenes, Euphorion of Chalcis, Polemo, and Seleucus of Alexandria.56 Through these writers there survives a certain amount of evidence, of very varying quality, about the composition, size, shape and appearance of these objects. But only in one source is there any mention of legible writing upon the Axones, which were already some two hundred and fifty years old by the time of the fourth-century Atthidographers: Euphorion, in the latter part of the third century B.C., gave the entirely credible 52 Aristophanes of Byzantium (see FGrH 241 F 37b); cf. Plut., Sol. 25.2. But Ar., Av. 1354, does not fit this explanation. 53 ’
e.g. Jacoby, A 309 n. 64 (k
yrbeiB is an archaic word and stands to ajoneB as uesmo
i to n
omoi; so also Keil, SVAVA 58 n. 1); or J. H. Oliver, in Hesp. 4 (1935) 5 ff., at pp. 9–10. 54 ‘ &aB [‘still in my Plut., Sol. 25.1. I cannot believe that Plutarch’s e
’ ti kau’ Zm day’] is simply copied from his source, e.g. Polemo, as Hignett (HAC 23 n. 3) seems to think possible. 55 [Plut. Sol. 24.2; Schol. Hom. Il. XXI. 282; Plut. Sol. 19.4; ibid. 23.4; Harp. s.v. o‘
ti oi‘ poiZto
i: see Rhodes, CAAP 131.] 56 Perhaps the most convenient way of citing the texts is to give the references to FGrH: 241 (Eratosthenes) F 37; 244 (Apollodorus) F 107; 339 (Asclepiades) F 1; 340 (Didymus) F 1; 341 (Seleucus) F 1. Cf. 72 (Anaximenes) F 13, but here Jacoby omits the important citation of Euphorion by Didymus (Harp., Phot. and Suid., s.v. o‘ k
atvuen n
omoB ¼ Euphorion fr. 6 in J. U. Powell, Collectanea Alexandrina (1925) 30).

320

Athenian Democratic Origins

information (known to us only through a citation by Didymus, quoted by the lexicographers) that the writing upon ‘the Axones and the Kyrbeis’ was boustrophedon [alternately left to right and right to left]57—a characteristic which could have been apparent even if very little of the script remained visible. At least three of the scholars named above (Asclepiades, Didymus and Seleucus) wrote special monographs on the Axones,58 and presumably therefore discussed the laws they contained; but this does not necessarily mean that they derived their information about these laws from the actual Axones: even if they saw the Axones themselves (and it is very doubtful if they did),59 they may very well have relied for such knowledge as they claimed to possess of their contents upon copies, or alleged copies. Even if we do not take seriously the fragment of Cratinus60 which speaks of the Kyrbeis of Solon and Draco as being used to parch barley, we may well feel that by the second half of the fourth century B.C. the Axones themselves may have been at least partly illegible. If so, however, copies must already have been made of the laws, or at any rate some of them, in the actual form in which they appeared on the Axones. For this there are two main pieces of evidence: first, laws are cited by ancient authors (as we have already seen) according to the numbers of the Axones of which they formed part;61 and secondly, phrases are quoted from laws which had certainly gone out of use well before the recodification at the end of the fifth century, as by Androtion (FGrH 324 F 36) and Aristotle (Ath. Pol. 8.3), the latter actually saying that ‘in the laws of Solon which they no longer use it is written in many places 57

See n. 56 above. See FGrH 339 F 1; 340 F 1; 341 F 1. 59 The statement quoted from Didymus by Harp. etc. is phrased in such a way as to suggest that Didymus himself had not actually seen the Axones: ‘o‘
ti g
ar’ wZsi
’
(sc. D
idymoB) ‘boystrowZd on Zsan oi‘ ajoneB kai oi‘ k
yrbeiB ’ ’ tv gegramm
enoi, ded
Zlvken E ywor
ivn en fi & ’Apollod
vrfiv’ [‘(Didymus) says that Euphorion in the ‘‘Apollodorus’’ has explained that the Axones and Kyrbeis were written boustrophedon’]. It is impossible for us to tell whether Asclepiades or Seleucus had seen the Axones. Of the two writers who give the most plausiblesounding descriptions of their physical appearance, namely Aristophanes and Polemo (see FGrH 241 F 37bc), only the latter can have seen them himself, unless Aristophanes is right in saying that Axones and Kyrbeis were alike, and that the Axones contained n
omoi, the Kyrbeis uys
iai [‘sacrifices’]. 60 Fr. 300 K/A (274K) ap. Plut., Sol. 25.2. 61 See n. 55 above. 58

Aristotle and the Documentary Sources

321

(pollaxo &y) ‘‘the Naukraroi are to exact’’ and ‘‘to spend out of the naukraric fund’’.’62 6. The trouble is that as soon as we begin to think in terms of copies of the laws of Solon the question arises whether the copies were accurate, or whether there may not have been ‘interpolations’ (to use Jacoby’s phrase)63 and even outright forgeries. And is it not all too likely that, for example, where a Solonian law had been modified, the modification would tend to be incorporated in copies of the ‘Solonian’ law?64 It was probably Aristotle himself, as most people now admit, who was deceived by the ‘constitution of Draco’ (Ath. Pol. 4), a fourth-century forgery.65 I have also argued in earlier essays that Aristotle was mistaken about the qualifications of the census classes ’ prokr
itvn [‘sortition from preand about Solonian kl
ZrvsiB ek selected candidates’]. There are, broadly speaking, three ways of accounting for these errors: (1) Aristotle may have been taken in by inauthentic copies of the laws in question, whether inaccurate, ‘interpolated’ or actually forged; or (2) in the absence of any surviving Solonian law, genuine or faked, he may have drawn false inferences from the other material before him; or (3) he may be relying upon ‘established traditions’, embodied in the work of one or more of the early Atthidographers, which had themselves been corrupted in one of the two ways just mentioned. Above all, as I have indicated before,66 Aristotle was perhaps particularly likely to be deceived by what looked to him like ‘documentary’ material. The general conclusion to be drawn from all these considerations is that Aristotle must not be conceived as necessarily having available to him either all the laws of Solon or none of them. We must at least entertain another possibility (which I believe to be the truth): that he had an accurate record of many of Solon’s laws but was misinformed about others, or unaware of them, and that for some of the provisions creating the framework of Solon’s polite
ia, 62 Most of the reasoning of Hignett, HAC 19–22, is beside the point if we admit no wholesale recodification of the Athenian laws between Solon and c. 410—as I certainly would not. 63 See n. 47 above. 64 Cf. Hignett, HAC 23, 313, on Plut., Sol. 19.4. (Cf., however, p. 316 above.) 65 The alternative, of course, is to suppose that Ath. Pol. 4 (and consequential changes) were added after Aristotle’s death. See p. 273 above. 66 Pp. 276–7 above.

322

Athenian Democratic Origins

which were not included in the laws on the Axones at all, he will have had to rely at least partly on tradition and inference, both of them all too liable to error. In so far as the original Axones were legible in the fourth century, the early Atthidographers are surely as likely as Aristotle to have made use of them; but the search for documents would be a very different matter, for reasons I hope I have made sufficiently clear. 7. It is time to sum up on the question of Aristotle’s documentary sources. First of all, like Hellanicus and the other Atthidographers, he certainly had the archon list,67 going back at least to Solon (594/3) and probably to Creon (683/2). This may well have been no more than a mere list of names: I myself doubt the existence on the list of even those brief annotations ‘referring to the list itself’ which Jacoby was prepared to accept.68 It was probably an extract from this list69 which was displayed on stone in the Agora about 425 B.C.70 Apart from the archon list it is difficult to think of any preCleisthenic document Aristotle might have had except ‘the laws of Solon’. We cannot be at all sure that he was able to read the Axones themselves: he will probably have had to rely at least partly on copies of the laws; but evidently he had copies of some sort, although whether they were complete and accurate or whether they contained gaps and errors and ‘interpolations’, even deliberate forgeries, we have no means of telling. We have seen that before Cleisthenes ‘archives’ may have been almost non-existent, and that for something like a hundred years after Cleisthenes the records may have been very incomplete and were not centralised. But there is one conclusion which does force itself upon us. If we have reason to suspect disagreement between Aristotle on the one hand and Androtion or any other early Atthidographer on the other, then, in the absence of confirmatory evidence on either side, there is every reason to prefer Aristotle, because he is very much more likely than any of his predecessors to have done research at first hand among the documents—such as they were. 67

See Jacoby, A 171–6, 197 etc.; Jac., i 14 ff.; Cadoux, AAKH, esp. 77–8. A 174–6, 177, 185–6, 188, 196, cf. 94, 96–7, 197. At least Jacoby has shown that it is impossible to conceive of the archon list as containing any detailed record of contemporary events or constitutional reforms. 69 See esp. Jacoby, A 172–3. 70 M/L 6 ¼ Fornara 23. 68

Endnotes

323

ENDNOTES i (from p. 260 n. 27). Ross (DAT 71) on the dating of certain passages in the Politics Newman’s three points were put forward more diffidently than Ross’s citations of him would indicate. (1) In PA IV 439 (on Pol. V 1312b6–7) Newman merely says ‘it is possible that . . . ’. (2) In PA II 333–4 (on Pol. II 1270b11–13) he begins, ‘It is quite unknown to what Arist. here refers, but I venture to suggest whether it is not possible that . . . ’. He continues, ‘If events of 333 B.C. are really referred to . . . ’, and ends, ‘I mention the hypothesis for what it is worth’. (3) In PA IV 255 (on Pol. IV 1299a14–19), after saying that Aesch. III 13–19 is ‘probably present to Arist.’s mind here’, he qualifies this a little farther on, with the words, ‘If cc. 13–19 of Aeschin. c. Ctes. are here present to Arist.’s mind . . . ’. ii (from p. 260 n. 28). Weil (AH 181–210) on the dating of certain passages in the Politics I will give just three examples. (1) Arist.’s references to the Persian empire (see Weil, AH 198–9) are timeless, and afford no grounds for supposing that he must be writing before Alexander’s conquests, as Weil infers. (2) I know of no reason whatever for dating the causeway that joined Clazomenae to the mainland precisely in the late 330s (as Weil, AH 199–201). And in any event Arist. is speaking of a time before the causeway can possibly have existed, so that he would quite ’ n
naturally have written oi‘ en Zsfiv [‘those on the island’] even if by his time the causeway had been built. However, it now seems that Chytrion or Chyton should probably be located not at the mainland end of the causeway but several miles farther to the south-west: see J. M. Cook, ‘The Topography of Klazomenai’, in ’Arw: ’Ej. (1953–4), m
eroB de
y teron, 149–57. And if Chytrion was thus well inland, it would have ’ i X
y trfiv [‘those from been even more natural for Arist. to contrast oi‘ ep ’ n
Chytron’] with oi‘ en Zsfiv [‘those on the island’], quite irrespective of whether the causeway existed or not. (3) It seems to me absurd to say, with Weil (AH 195–8, at p. 196), following Keil (SVAVA 123), that aper QZba
ioiB [‘as at Thebes’]) ‘can only be in Pol. VI 1321a28–9 (kau
the present: it is impossible that the past should be implied’, and that after the destruction of Thebes (in 335) Arist. ‘could only have written QZba
ioiB pot
e [‘once at Thebes’] or some similar expression’ (my italics in each case). It is a pure supposition that the situation depicted in Pol. III 1278b25–6, admittedly past, was restored between 338 and 335. And Arist.’s usage elsewhere shows that either a past or a present

324

Athenian Democratic Origins

verb may be implied with our kau
aper QZba
ioiB. Here VI 1318b21–7 is illuminating: according to the rule imagined by Keil and Weil, it would be ‘impossible’ to understand a past tense in line 25, with v
‘ sper ’ Mantine
ifia [‘just as in Mantineia’], yet the rider which Arist. en ’ Mantine
ifia happens to add in lines 26–7, where we have v
‘ sper en
[‘just as was once the case in Mantineia’], shows that we must. pot’ Zn
uasin [‘they are accusAgain, the main verb in V 1302b18–19 (ei’ v tomed to’]) is naturally taken as a present, yet ostracism had not been used in Athens for nearly a hundred years, and was almost certainly dead at Argos also. In V 1316a29–39, too, the two verbs are both in the present, while the examples are from the past. iii (from p. 261 n. 31). Quadriremes in the Athenian fleet IG II2 1627 (330/29) b. 266 ff., at lines 275–8, mentions 18 quadriremes. Unfortunately, those of the ‘navy lists’ for the 330s which survive at all are very fragmentary, and the fact that quadriremes do not appear in them is not decisive. In IG II2 1624 (late 330s) c there is a vacant space after the ‘total number of triremes’ in lines 125–9, and this may suggest that at the time in question no quadriremes existed. But it is probable that quadriremes were in commission just before 330/29, for we know from IG II2 1631 (323/2) c. 430–41 that Stesileides of Siphnos had been commander of a quadrireme, and IG II2 1627 (330/29) a. 194–6 shows that Stesileides was dead by 330/29. (This was realised by Bo¨ckh, SA1 III ¼ Die Urkunden u¨ber das Seewesen des Attischen Staates (1840), 75–6.) I doubt myself if the Athenians would have built their first eighteen quadriremes all at once (when they turned to quinqueremes, they began with seven: IG II2 1629 (325/4) d. 783 ff., at line 811); but it does not look as if they began to build them much before 330. iv (from p. 264 n. 46). Jaeger on the dating of Aristotle’s Polite&iai The statements I have quoted are all the more strange, in that earlier in ’ tvn the same book (Arist.2 265–6) Jaeger takes the expression ek & synZgm
envn politeivn & in Arist., EN 1181b17, to mean ‘in the light of the constitutions we have collected’, and remarks that ‘much useless ingenuity has been expended on denying’ that this is a reference to ‘the collection of 158 constitutions’. He goes on to emphasise that Arist. is saying, ‘Now I have at my disposal the copious material of the 158 constitutions, and I am going to use it in order to give to the ideal state a positive foundation’; and he adds, ‘This was written at the end of the latest version of the Ethics, that is to say, in his [Arist.’s] last

Afterword

325

decade. The collection of constitutions came into being at the same time’. But how did 158 constitutions all ‘come into being at the same time’? v (from p. 279 n. 13). Xenophon as a source for the Ath. Pol. According to Day and Chambers (AHAD 11), Xenophon is ‘ruled out by sharp discrepancies between him and Aristotle’; but the only authority they give for this statement is Hignett (HAC 386–9), and Hignett, although he sets out the discrepancies, does not say that they ‘rule out’ the use of Xenophon in the Ath. Pol.—see e.g. p. 386, where Hignett declares that the Ath. Pol.’s dating of the occupation of Phyle to the beginning of winter is ‘simply a deduction from Xenophon’s reference to the snowstorm’ (in Hell. II iv 3). One might just as well conclude that Thucydides is ‘ruled out’ as a source for the Ath. Pol. by the discrepancies between his and Aristotle’s account in 29–33. Wilamowitz (AuA I 166), who denied the use of Xenophon by Aristotle, was driven to invent a lost work of Theramenes as the common source of Xen. and Arist. In view of resemblances such as Xen., Hell. II iii 19 and Ath. Pol. 36.2, I find it hard to believe that Arist. did not use Xen. directly; but this can scarcely be proved. Certainly Arist. also used a different source for the years 410–403, but I do not see how we can decide what it was—one or more Atthides, or the Oxyrhynchus historian’s continuation of Thucydides, or Ephorus (using the Oxyrhynchus writer), and conceivably a speech or speeches delivered shortly afterwards (cf. Lys. XII and XIII, which however cannot have been the source in question).

AFTERWORD by P. J. Rhodes This paper was written before the publication of my commentary on the Ath. Pol. in 1981a and that of M. Chambers in 1990;b indeed, to my great benefit, while I was working on my commentary Ste. Croix lent me copies of this paper and of other papers in this collection, and I am ashamed to discover that, although I referred to these papers at various points, I omitted to include a general acknowledgment in my Preface. Ste. Croix’s paper antedates work on Androtion and the other Atthidoa Rhodes, CAAP; cf. id., Aristotle: The Athenian Constitution (Penguin Classics, 1984), whose title does not indicate a change of mind about the authorship of the work. b Aristoteles, Staat der Athener, u¨bersetzt und erla¨utert von M. Chambers (1990).

326

Athenian Democratic Origins

graphers by P. Harding,c who reacted strongly against Jacoby’s view that each Atthis had a political slant, usually opposite to that of its immediate predecessor: Harding shares my view (cf. below) that Androtion was one source of Ath. Pol. among many; he believes, as Ste. Croix did not, that Philochorus made considerable use of Androtion.d It also antedates work on the kyrbeis/axones of Solon, in particular by R. S. Stroud,e who argued that the axones and the kyrbeis were separate sets of objects (the axones being the original objects on which Draco’s and Solon’s laws were inscribed), and were still available for consultation in the fourth century, and by A. Andrewes,f who supported the view of E. Ruschenbuschg (which was taken into account by Ste. Croix: cf. pp. 318–19 above) that axones and kyrbeis were alternative words for the same set of objects.h The authorship of the Ath. Pol. continues to be debated. Ste. Croix, though starting section (i) with the statement that Aristotle can hardly have written all 158 Politeiai, nevertheless was prepared to believe that he had written this one: Aristotelian authorship is rejected by me but accepted by Chambers. Aristotelian scholars would not quarrel with Ste. Croix’s view that the Politics as we have it dates from 335–322 but parts of it are likely to be older.i I believe that a first version of Ath. Pol. was written in the late 330s and the text which we have incorporates revisions made in the early 320s; Chambers is sceptical of revisions and dates the whole work 328–325. On sources for the historical part of Ath. Pol., Section (ii) expresses doubt about what had become the standard view, that Androtion’s Atthis was the principal source; (iii) doubts whether Androtion on the early Athenian constitution was important or served as a basis for Philochorus, doubts also whether Androtion was the source of the material found both in Ath. Pol. and in Plutarch’s Solon, and argues that the first part of Ath. Pol. embodies original historical research as the early Atthides did not—just as the second part embodies original research into the current constitution of Athens. I sympathise with the attack on excessive attribution to Androtion, and regard the first part as based on a variety of written texts including Androtion’s and perhaps c Culminating in Harding, ATA: his earlier articles are listed in the Bibliography of that book. d On the Atthides cf. also Rhodes, in Purposes of History . . . 24–6.v.1988 (Studia Hellenistica 30, 1990), 73–81, adopting a position not far from Harding’s. e R. S. Stroud, The Axones and Kyrbeis of Drakon and Solon (U. Calif. Pub. Class. Stud. 19, 1979). f A. Andrewes, in F
oroB: Tribute to B. D. Meritt (1974), 21–8. g E. Ruschenbusch, S
ol!noB n
omoi (Hist. Einz. 9, 1966), 14–25. h I follow Andrewes, Chambers follows Stroud. i Cf. I. Du¨ring, RE Supp. xi (1968), 159–336 at 290–1, 334–5.

Afterword

327

other Atthides (with problems arising when the author did not successfully resolve disagreements between his sources), but I do not believe that the author engaged in original research as opposed to making the best use he could of books available to him; Chambers believes in Aristotelian authorship, and in stronger authorial control of the material than I do, but his position on the sources used by the author is close to mine. On the particular issue of the Solonian chapters, I have suggested that the author used some earlier work on Solon (deriving his knowledge of Solon’s poems and his laws from that), and that Plutarch used both that work directly and Ath. Pol. Plutarch’s methods of work have been studied, with particular reference to the Roman Lives, by C. B. R. Pelling.j Section (iv) asks what documentary evidence would have been available to a writer researching into early Athenian history, and concludes that for the pre-Cleisthenic period there would have been only the archon list (probably a bare list of names, without annotations) and a version of the laws of Solon, probably a copy whose relationship to the original must be left uncertain. I agree with Ste. Croix on a bare archon list, but am more willing to believe that an authentic text of Solon’s laws did survive in the fourth century and that, despite the orators’ habit of ascribing later laws to Solon, it was possible for those who wished to do so to distinguish between Solon’s laws and later laws; despite the Persian sack of Athens in 480, I think it possible that the sixth century generated a few other documents which survived to the fourth.k On the other hand, M. H. Hansen has argued that Solon’s axones did not include any constitutional laws, and that Ath. Pol.’s account of the Solonian constitution reflects not the facts of the sixth century but the wishful thinking of the fourth.l j See especially C. B. R. Pelling, JHS 99 (1979), 74–96 with postscript in B. Scardigli (ed.), Essays on Plutarch’s Lives (1995), 312–18; JHS 100 (1980), 127–40 ¼ his Plutarch and History (2002), 1–44 ch. 1; 91–115 ch. 4. k Cf. Rhodes, in M. Pie´rart (ed.), Aristote et Athe`nes/Aristoteles and Athens . . . 23– 5.v.1991 (1993), 53–64. See also J. P. Sickinger, Public Records and Archives in Classical Athens (1999), chs. 1–2. l M. H. Hansen, C&M 40 (1989) [publ. 1993], 71–99; this and accompanying papers published also as a separate book, W. R. Connor et al., Aspects of Athenian Democracy (1990), same pagination. For another sceptical treatment of fourth-century accounts of Solon see C. Mosse´, Annales (ESC) 34 (1979), 425–37.

8 The metra in Aristotle, Eth. Nic. V vii 5 1134b35–1135a3

’ t a de kat a synu
ZkZn kai t o symw
eron tvn & dika
ivn o
‘ moi
a esti to&iB m
etroiB. oy’ g ar pantaxo &y ’
isa t a oi’ nZr a kai sitZr a ’ ’ &yntai, me
izv, oyffl de pvlo &ysin, el
’ attv. m
etra, all’ oyffl men vno [‘Rules of justice based on convention and expediency are similar to measures; for wine and corn measures are not everywhere equal, but are larger where they buy, and smaller where they sell.’] This passage is usually understood to say that corn and wine measures are ‘larger in wholesale and smaller in retail markets’,1 or at any rate that ‘wholesale and retail measures differ’;2 and this is often explained as by Stewart:3 ‘Dealers buying up corn or wine in large quantities compute by means of large units of measurement; but when they retail their stock they find it convenient to use smaller units—e.g. bottles instead of hogsheads.’ I suggest that this cannot possibly be right, for the following reasons: 1. It does not fit the context, which surely requires that the measures concerned be prescribed by law, not merely chosen by agreement between the parties, or by one party with the aim of defrauding the other. At the beginning of the chapter (1134b 18 ff.), Aristotle says there are two kinds of t o politik on d
ikaion [‘political justice’], i.e. t o wysik
on [‘by nature’] and t o nomik
on [‘by 1 As by Sir David Ross in the Oxford translation, H. Rackham in the Loeb, H. Jackson in his edition of Book V. 2 J. Burnet, The Ethics of Aristotle (1900), 235 n.; cf. J. A. K. Thomson, The Ethics of Aristotle (1953), 138. And see n. 3 below. The interpretation adopted by A. Grant, The Ethics of Aristotle4 (1885) II 128 n. is a little better, if one reverses his ’ &yntai as referring to ‘the producing’ and oyffl de pvlo &ysin to conception of oyffl men vno ‘the consuming countries’. 3 J. A. Stewart, Notes on the Nicomachean Ethics (1892), I 496. Essentially the same interpretation is given by R. A. Gauthier and J. Y. Jolif in their edition (L’Ethique a` Nicomaque, II, Commentaire, 1e Partie, 395–6).

The metra in Aristotle

329

law’], the former being that which is valid pantaxo &y . . . kai oy’ tfiv& ’ m
doke&in Z Z [‘everywhere . . . and not through opinion one way or the other’]. The word pantaxo &y throughout this chapter means ’ P
ersaiB [‘in Persia’] in 1134b 27), literally ‘everywhere’ (note en but especially, of course, (as so often elsewhere in Aristotle) ‘equally in each and every polis’.4 That the pantaxo &y of 1135a 1 ’ i oyd’ ’ ai‘ in particular has this sense is put beyond doubt by the epe polite&iai [‘since neither are constitutions’], explanatory of oy’ ’ a pantaxo &y [‘not the same everywhere’] in the next sentence tayt (1135a 4), which is explicitly parallel to the one with which we are concerned. Introducing his measures-simile, Aristotle begins a de kat a synu
ZkZn kai t o symw
eron tvn & (1134b 35--35a 1), t dika
ivn [‘rules of justice based on convention and expediency’], a phrase which is merely a synonym for t a de n
omima d
ikaia [‘legal ’ a nomik justice’] (cf. all on kai synu
Zkfi Z [‘but legal and conven’ tional’], 1134b 32; also the further synonym, anur
vpina d
ikaia [‘human justice’], 1135a 4). Returning from his simile, with ‘ ivB [‘similarly’] (1135a 3), Aristotle says that t a m Z wysik a omo
’ ’ all’ anur
vpina d
ikaia (i.e. t a n
omima d
ikaia) [‘justice which is ’ a not natural but human’ (i.e. ‘legal justice’)] are oy’ tayt ’ i oyd’ ’ ai‘ polite&iai [‘not the same everywhere, pantaxo &y, epe since neither are constitutions’]: they are different in different states, since constitutions differ. Unless Aristotle is drawing a false and misleading parallel, he must also have in mind measures prescribed by law, which differ from state to state, with larger measures in states ‘where they buy’ and smaller ones ‘where they sell’. The standard interpretation, however, conceives the measures as differing merely by the choice of the parties to the various transactions: the ‘wholesaler’ prefers to buy by large measures, the ‘retailer’ to sell by smaller ones, and that is all. 2. The interpretation in question also makes the parallel drawn by Aristotle break down over oy’ pantaxo &y [‘not everywhere’], which occurs in both parts of the simile (1135a 1 and 4): in the ‘ ivB [‘similarly’] clause, as we have seen, it means ‘not in all omo
’ states’; but within the measures-simile the oy’ pantaxo &y . . . all’ oyffl men . . . oyffl d
e [‘not everywhere . . . but where . . . and where’] will (on 4

Of course Aristotle will always be thinking primarily of Greek poleis, but he can speak of at least one non-Greek state, Carthage, as a quasi-polis, having a polite
ia [‘constitution’]: see Pol. 1272b 24.

330

Athenian Democratic Origins

this interpretation) be referring to different transactions—or at most, if we suppose (as I would not) that ‘wholesale markets’ and ‘retail markets’ existed separately, to different places, irrespective of state boundaries. 3. In every purchase there must ex hypothesi be a seller, and in every sale a purchaser. Precisely as much as is bought in ‘wholesale markets’ is also sold there, and in ‘retail markets’ there is similarly the same quantity of selling as of buying. Why should Aristotle ’ &yntai [‘where choose to ignore entirely in the one case (oyffl men vno they buy’]) the sellers, and in the other (oyffl de pvlo &ysin [‘where they sell’]) the buyers? ’ &yntai and pvlo &ysin are entirely inappropri4. The words vno ate, on the interpretation we are considering. They mean ‘buy’ and ‘sell’; they do not mean ‘deal in a wholesale market’ and ‘in a retail market’ or anything of that kind. It is not in the interests of buyers as such to use relatively large units of measurement, of sellers as such to use relatively small units. As we shall see presently, there is one sense, and one sense only, in which a buyer will, by the very fact that he is a buyer, wish to use measures which are me
izv ’ attv [‘larger’], and a seller, similarly, those which are el
[‘smaller’]; but this will lead us to an entirely different interpretation of Aristotle’s simile. We must therefore reject the standard interpretation as certainly false, and look for another, in which the measures are (A) prescribed by law, (B) differently in different states; and by which we can explain (C) Aristotle’s curious concentration first on one side and then on the other of transactions all of which are essentially two-sided, and (D) how buyers as such can be thought of as ’ attv preferring their measures to be me
izv [‘larger’], sellers as el
[‘smaller’]. I think this investigation can be pursued without venturing too far into the marshlands of metrology, a subject the entire study of which, as Pritchett5 has recently remarked, ‘is now very much in a state of flux’. Let us first clear out of the way one or two interpretations of our passage which might be suggested. Aristotle’s m
etra cannot be different grades of measures within the same system (tons and pounds, or gallons and pints), as in the interpretation we have 5

In Hesperia 25 (1956) 192. Many useful observations about Greek measures of capacity are made by Pritchett, op. cit. 192–6, 200–1, and by D. A. Amyx in Hesperia 27 (1958) 168–95, cf. 287–307. [See also Afterword, pp. 347–8 below.]

The metra in Aristotle

331

already rejected. The whole cast of the sentence, especially the oy’ pantaxo &y ’
isa [‘not everywhere equal’], is dead against this— of course all systems of measures must have different grades. And it is certainly not in the interest of buyers as such to use relatively large measures and of sellers as such to use relatively small ones, but rather the reverse. Nor can the m
etra be different systems, as for example the Attic and the Aeginetan, within each of which the corresponding grades have the same names but different values. Greek cities used systems of weights and measures en bloc: they did not use one system (the Attic, let us say) for corn, another (the Aeginetan, perhaps) for wine, and a third (the Pheidonian, for example) for other things; nor could they have prescribed one system for buying and another for selling, because there must be sellers in every purchase and buyers in every sale, and they (the other parties in each case, whom Aristotle here feels he can afford to ignore) would simply demand more money for the commodity, or give less money, as the case might be, if the parties Aristotle is thinking of attempted some sharp practice involving the standard of measurement. Business would become impossible under such conditions. At Athens a medimnos or a drachma meant an Attic medimnos or drachma, at Aegina it meant an Aeginetan one. If, let us say, an Athenian merchant, selling a medimnos of corn at Aegina, handed over an Attic medimnos, he would be told that it was less than a medimnos (an Aeginetan medimnos); and if he insisted on using his own, Attic measure, the price would be correspondingly reduced, so that he would gain nothing. And in any case, as we have seen, Aristotle must be thinking of purchases in one group of states and sales in another. Now m
etra, in a passage such as ours, might a priori have either (or both) of two quite different meanings, unfortunately not distinguished by LSJ, s.v. m
etron, I 2: it could mean (a) theoretical measures, standards of measurement (as, e.g., in Hom. Il. VII 471; Od. II 355; IX 209; Hdts VI 127.3), or (b) the physical containers used in applying those standards to actual objects (as, e.g., in P. Tebt. 105.40).6 These physical m
etra, of course, may or may not be of the proper size in any given case. What I am going to suggest ’ attv refer to physical containers, is that Aristotle’s me
izv and el
6 Similarly, one might distinguish the physical m
etra of Il. XII 422 from the theoretical m
etra of the other passages cited by LSJ, s.v. m
etron, I 1.

332

Athenian Democratic Origins

and mean greater or less than they ought theoretically to be. This at once satisfies our condition D as no other interpretation will, for it is indeed in the interest of buyers as such to use physical measures larger than the theoretical measures they are supposed to represent, and of sellers as such to use smaller measures. Aristotle refers specifically to ‘wine and corn measures’, not just ‘measures’ or even ‘liquid and dry measures’. Of course it would be possible to have only one official container for each important grade of the city’s measures, dry and liquid: one could test a doubtful measure by pouring corn or water into it and then transferring the contents to the official container.7 But we have ample confirmatory evidence that some Greek cities, at any rate, did have separate official containers for wine8 and for corn.9 Today there are laws in all countries against the use of weights and measures not of the prescribed size, with official inspectors to see that the laws are observed. So it was at Athens, for example, where the Metronomoi were the responsible magistrates,10 and must (even before the date of the inscription of c. 100 B.C., IG II2 1013, which will be discussed below) have kept a whole set of official weights and measures, so that they could settle disputes about the validity of private ones used by buyers and sellers. The Metronomoi (and the corresponding magistrates elsewhere) would have had the final word, and if a dispute arose it would be settled by the production of the official measures kept by the 7 IG II2 1013 (see below), lines 10–11, says that all those who sell in the Agora ’ v[ij & m
et]rvi. But (a) the inscription is more are to measure out all liquids tvi :& ayt ’ v than two centuries later than Aristotle, and (b) the words tv fi & ayt fi & m
etrfiv, far from meaning ‘with one single physical container’, can hardly be limited even to a series of containers of the same size, for there would of course have to be measures of different grades (from kot
ylZ, say, to metrZt
ZB); and the expression can only mean ‘according to the same measure’, in the sense, ‘with a measure of the same standard’. 8 e.g. (a) the Chian wine-se¯ko¯ma published by W. G. Forrest in BSA 51 (1956) 63–7; (b) the se¯ko¯ma inscribed oi’ nZ[r aj e]i’ kosak
ot ylon st
amno[B] [‘twenty-kotyle wine measure’], also from Chios, referred to on p. 65 of Forrest’s article; (c) the Thasian se¯ko¯ma, bearing the word oi’ nZr
a, published by J. Pouilloux, Recherches sur l’histoire et les cultes de Thasos II (1958) 101–2, no. 194 & Pl. XV; cf. id. I (1954) 405– 6, no. 153 & Pl. XLIV 1, 2. 9 e. g. (a) IG II2 1013, lines 21, 27; (b) an inscription of about the same date, from Delos, ID 1820 (¼ Choix 143), which speaks of a s
Zkvma sitZro &y med
imnoy [‘measure of a medimnos of corn’]. 10 ffl See esp. Arist. Ath. Pol. 51.2: kai oytoi t vn & m
etrvn kai t vn & staum vn & ’ epimelo &yntai p
antvn, o‘
pvB oi‘ pvlo &ynteB xr
Zsontai dika
ioiB [‘and they are responsible for all weights and measures, ensuring that sellers use fair ones’].

The metra in Aristotle

333

Metronomoi. All privately owned measures used in a particular state would therefore tend (in so far as they were not fraudulent) to approximate in capacity to the official physical measures, whether or not these were of exactly the right theoretical size. We know from various kinds of evidence that there existed in antiquity—as we should expect—physical measures used by private individuals which were not of the size they ought properly to have been; so that, for instance, a man handing over a prescribed quantity of corn or wine in payment of rent in kind, or in repayment of a loan, might be deceived or forced by his landlord or his creditor into giving more than the right quantity. Thus in A.D. 603 (to take a rather extreme case) we find Pope Gregory the Great11 writing to a notary, Pantaleo, of his indignation at the discovery that certain coloni Ecclesiae have been obliged to hand over their frumenta according to a modius-measure containing no less than 25 sextarii instead of the lawful 16: he expresses his pleasure at the news that Pantaleo has broken up the iniquitous measure ‘et iustum fecisse’ [‘and made a proper one’]. It would be interesting to know how many sextarii the new ‘modius iustus’ contained, in view of Gregory’s order, in another letter (to Peter, a Sicilian subdeacon),12 that the rustici Ecclesiae were not to be compelled to hand over their frumenta according to a modiusmeasure containing more than 18 sextarii. Again, an almost contemporary contract from Byzantine Egypt,13 entered into by the steward of part of the vast estates of the Apion family, stipulates that the steward is to pass on to his employer the 15% profit he will make out of employing the artabus-measure in use on the estate, where most rents would be paid in kind. In order to avoid later disputes, the parties to a contract for payment in kind might agree that measurement was to be according to a particular container, as in a lease from Ptolemaic Egypt, dated 103 B.C.,14 which provides for payment in new wheat, ‘measured by the 6-choinix 11 Epist. XIII 37 (MGH, ed. L. M. Hartmann, II, p. 400) ¼ XIII 34 (Migne, PL 77). 12 Epist. I 42 (MGH I, p. 62) ¼ I 44 (Migne). 13 ‘ vn ‘ P. Oxy. I 136, lines 27–9: prosomolog v& de lZmmat
isai tfi Z& ym & yperwye
ifi a ‘ er paramyue
iaB to &y paralZmptiko &y m
etroy t vn ’ ‘ ’ abaB d
eka yp & artab vn & ekat on art
p
ente [‘I agree further to gain for your magnificence, as compensation for the measure used in receiving payments, fifteen extra artabae on every hundred’ (tr. Grenfell/Hunt)]. 14 P. Tebt. 105, lines 40–1.

334

Athenian Democratic Origins

dromos measure of the Temple of Souchos in the aforesaid village by just measurement’. It may be objected that these are late sources to cite in elucidating a passage in Aristotle. But there are few early Greek writers in whom we might expect to find evidence of the fraudulent or extortionate use of false measures. Fortunately, there happens to be a clear reference in Hesiod’s Works and Days15 to the possibility of obtaining some advantage by using a greater or smaller measure: the poet advises the use of the same mxtron for both sides of a transaction (the loan and its repayment), implying that some men did cheat by using different measures.16 And it was not only private individuals in antiquity who used incorrect weights and measures. The modius-measure of the reign of Domitian, now in Chesters Museum, found in 1915 at Carvoran, and published by Haverfield,17 bears an official inscription stating that it was ‘tested to a capacity of 17 and a half sextarii’ (exactus ad S XVIIS). Now this is already almost 10% above the proper capacity of the modius (17 and a half instead of 16 sextarii), and in fact the measure contains about a sixth more again than its inscription indicates: nearly 20 English pints, against the 16.8 pints which would be the equivalent of 17 and a half Roman sextarii. A blunder of this magnitude would be scarcely credible, and we must surely accept Haverfield’s explanation, that ‘we may have here a device by which Roman officials robbed provincials’.18 Later legislation expressly forbade tax-collectors 15 Lines 349–51. Cf. Matt. 7.2; Mark 4.24; Luke 6.38. In the OT, see esp. Deut. 25.13–15; Prov. 20.10, 23; Amos 8.5; Mic. 6.10–11; also Lev. 19.35–6; Prov. 11.1; Ezek. 45.9–12. 16 [A ms. note refers to Theophr. Char. 30.11, where, though details are obscured by textual corruption, it is clear that the miser gives short rations to his slaves through trickery over measures.] 17 In Arch. Aeliana, 3rd Series, 13 (1916) 85–102. Cf. the measure of A.D. 367–75 found at Ponte Pun˜ida, described by Haverfield, p. 93. [But see R. Duncan-Jones, ZPE 21 (1976), 61, for the possibility that the measure published by Haverfield is based on a larger modius, known from metrological sources.] 18 The alternative suggestion of A. F. Berriman, in Arch. Aeliana, 4th Series, 34 (1956) 130, in so far as I understand it, seems to me implausible. I take it that it was useful to have the weight of the vessel recorded, as on this modius-measure (38 Roman pounds), so that it could be used if desired for weighing substances which required a container. Roman containers are often so marked. Measures of weight and of capacity were officially equated at Rome in respect of wine, at any rate, from a very early date, by the Lex Silia (Riccobono, FIRA I2 p. 79). In this particular case, of course, any intention to cheat would result in an understatement of the

The metra in Aristotle

335

to extort more than the standard amounts of tribute by using measures or weights exceeding the norm.19 Again, it may indeed be objected that in Roman Britain we are very far removed from the Greece of Aristotle’s day. There is, however, a well-known Athenian decree, IG II2 1013,20 dating probably from the very last years of the second century B.C. or the early years of the following century, which strongly suggests that Athenian magistrates had been making and using weights and measures (official ones, therefore) which were not correct. A main purpose of the decree was to enforce the use of standard weights and measures, of which official specimens (sZk
vmata) were to be preserved by the magistrates concerned,21 who are not named in the surviving portion of the decree but must surely have been (or have included) the Metronomoi. These magistrates were to compel all those selling in the Agora to use these weights and measures. ’ estv m The inscription goes on to say,22 kai mZ[k]
: eti ej
: Zdemi : &ai ’ Zi & p[oi]
arx Zsasuai m
Zte m
etra m
Zte st
aumia [me
izjv m]Zde ’ at]tv to
y tvn [‘and it is no longer permitted for any magistrate :el[
to make measures or weights [larger o]r smaller than these’]:23 the word mZk
eti [‘no longer’] (which seems certain) is significant. weight—and this is indeed what may have happened: the vessel (in its present state, at any rate) weighs 25 and three-quarters English pounds, instead of the 27 and a half which would represent 38 Roman pounds (see Haverfield, op. cit. 94–9). 19 C. Th. 12.6.21 ¼ C. J. 10.72.9, of A.D. 386. It is interesting to find the very same imperial constitution prescribing the rates of commission which are to be paid to the collectors in respect of various commodities—it may well be that the surplus one and a half sextarii of the Carvoran modius represents the commission (of between 9 and 10%) which had to be paid to the collector. The Armenian collectors are allowed a commission of 6 and two-thirds in respect of wine and lard, as against 5% elsewhere: in making this provision, the emperors describe themselves as ‘humanitatis necessitate commoti . . . longinquitatis causa’, a consideration which might also be thought to apply to the frontier of Britain. 20 For a new restoration of lines 49–62, with the aid of a fragment of another copy of the inscription, see B. D. Meritt in Hesperia 7 (1938) 129–30. This inscription has often been discussed, seldom very satisfactorily. References are given by John Day, An Economic History of Athens under Roman Domination (1942), 111 n. 349; add S. Young in Hesp. 8 (1939), at pp. 279–80; M. Crosby in Hesp. 18 (1949) 108–13. [Tr. M. M. Austin, The Hellenistic World (1981), 191–3, no. 111.] 21 Lines 7–11, cf. 54–5. 22 Lines 11–12. 23 I suppose it is possible that what this clause forbids is the making of measures or weights of a higher or lower grade on the standard scale (than those actually provided for in the lost portion of the decree); but if that were intended I would have expected a more sweeping prohibition against the making of measures or weights of any grade other than those specifically authorised.

336

Athenian Democratic Origins

Quite a number of actual Athenian measures24 and weights25 have come to light in the course of the excavations of the American School, and some of them are stamped in such a way as to show that they were official ones. It is not yet possible to say anything very positive about discrepancies among the measures,26 but there seems to be no doubt that the weights show marked irregularities. ‘A lack of any close adherence to a fixed standard on the part of the Athenian officials’ has been suspected, as a result of the examination of some early bronze weights of the late sixth or early fifth century.27 ‘The many weights in the Agora offer a fertile field for investigation, but the divergencies from a single standard are so frequent and so great that it is doubtful if any satisfactory conclusions can be reached.’28 There is good reason, then, to suspect that even the official measures and weights of Greek cities might depart widely from the norm. It is my belief that the passage in Aristotle we are now considering shows (and this is its real interest) that Greek states might sometimes make their official physical measures, forming the standard by which their theoretical measures were applied or interpreted in practice, larger or smaller than they ought to have been. And the fact that Aristotle attributes ‘larger’ measures to cities ‘where they buy’ and ‘smaller’ ones to cities ‘where they sell’ proves that the variations were not (or not always) mere random ones, caused by inefficiency, or the difficulty—which must often have been a serious one—of deciding precisely what the desired standards were. (I shall return to this question later.) Now any dispute about the size of official measures, and any complaint that 24

See (1) V. Grace in Hesp. 3 (1934), at p. 297; (2) T. L. Shear in Hesp. 4 (1935), at pp. 346–7; cf. H. A. Thompson in Hesp. 6 (1937), at p. 166; (3) O. Broneer in Hesp. 7 (1938) 222–4; (4) S. Young in Hesp. Suppl. 8 (1939) 274 ff., esp. 279–80. The first three are measures authorised by the State, the fourth is a clepsydra which must have belonged to the tribe Antiochis. The fruit measure published by M. Crosby in Hesp. 18 (1949) 108–13 is apparently a private ‘three half-choinix’ measure, stamped as approved on behalf of the State: see IG II2 1013, lines 17–18, 22, 63–7. 25 See T. L. Shear in Hesp. 4 (1935), at pp. 347–8; Hesp. 7 (1938), at p. 362; Hesp. 9 (1940), at p. 307. 26 It is sometimes difficult to be sure what measure a given container represents; and even when this is reasonably clear, we often cannot know exactly how far it was intended to be filled—cf. IG II2 1013, lines 18–23, with 23–5. 27 By T. L. Shear in Hesp. 7 (1938) 362. 28 Shear in Hesp. 9 (1940) 307.

The metra in Aristotle

337

the magistrates responsible, or the workmen they employed to make the measures, had acted negligently or improperly, would ultimately have to be settled by a decision of the supreme political or judicial body of the city in question. Therefore, in each case where the variations from the norm were not random, the decisive factor could only be the interests of the majority of the citizens, or rather of those citizens who had effective political power—in the case of an oligarchy, the pol
iteyma [ruling group],29 or the dominant members of it; in the case of a democracy, the whole body of citizens, or those who in practice were able to make their will prevail in the assembly or the courts. What Aristotle implies, then, is that in states where the majority of the citizens having political power were preponderantly buyers of wine or corn, the measures used in dealing in those commodities (the official measures, and therefore, as we have seen, the private measures as well) would be larger than they ought in theory to have been, but smaller in states where those same citizens were mainly sellers.30 Citizens, at least those of any political importance, will have been largely sellers and not buyers of corn and wine in many Greek cities: one thinks first, perhaps, of the Thessalian cities for corn, of Chios for wine. Apart from those few foreigners to whom ’
& egktZsiB the privilege of gZB [‘right of land-tenure’] had been granted, only citizens could own land in freehold,31 and only they would produce corn or wine. But circumstances differed greatly from city to city, and even if it is hard to believe that more than a few imported the bulk of their corn, as Athens did, many may have imported much of their wine. Athens, as so often, is a special case: here I would expect that the landowners (the majority of the citizens) would always try to grow enough corn for their own households, but would rarely sell corn on the market, while the considerable urban population, both citizen and metic, would

29 See Arist., Pol. 1305b 22--35; 1293a 12--17, 21--4;1321a 26--31, etc.; also SEG IX I, lines 5, 6 etc. 30 In each case there would of course be a minority whose interests would be just the reverse. But Greek states, even the democracies, paid much less attention to the rights of minorities than modern democracies do. 31 ’ ’ isuvsa [t From Lys. VII 10 ( ’Alk
ifia ’Antisu
enoyB apeleyu
erfiv em
o xvr
ion] [‘I leased (the plot of land) to Alkias the freedman of Antisthenes’]) I think we must infer that metics could be leasehold tenants at Athens, if not elsewhere. This is often overlooked.

338

Athenian Democratic Origins

regularly buy large quantities of imported corn. Athenian corn measures would therefore tend to be on the large side. I would stress the political element, which alone can explain why Aristotle felt able to ignore sellers in one group of transactions and buyers in another (our condition C). He ignored them because not they but the other parties to the transactions were able, indirectly, to decide exactly how much larger or smaller than the norm the physical measures were to be. It is not necessary to suppose that there was much intention to cheat deliberately. When it is difficult to produce a measure of exactly the right size—when, indeed, there might well be some doubt what the ‘right’ size was—the natural tendency would be to err on the safe side. One can easily imagine an Athenian Metronomos saying to the workman he was employing to make his m
edimnoB sitZr
oB [‘grain medimnos’], ‘You know how angry the Athenians will be if they think we’ve given them a scant medimnos, because of course they will nearly always be buying and not selling with it, so at all costs don’t make it too small.’ The main fact behind Aristotle’s statement, then, may simply be that in a world where precise standards of measurement were difficult to achieve, citizens thought of their own interests and took care to err on the safe side. It may be that sometimes there was a more deliberate attempt to cheat, by manipulation of the measures. Clearly, the more they departed from the norm, the more obvious it would be, and the greater the likelihood that allowance would be made for the discrepancies. In the Athenian corn trade, in particular, any noticeable exaggeration in the size of the measures would certainly be compensated for by an adjustment of prices all along the line: if the medimnos-measure were found to ‘ be as much as, say, seven instead of six Attic ekte& iB [‘sixths’], the emporoi who imported the corn would simply raise their prices by an obol in the drachma; and since the prices of corn and its derivatives were then strictly controlled (in the late fourth century, ’
at any rate),32 right up to the point at which the alwita [‘flour’] or ’
rtoi [‘loaves’] reached the eventual consumers, no one would a really profit from the larger measure. But if the corn-measures were only a fraction over-sized, it might be hoped that they would be taken as correct—and since Athens in the mid-fourth century 32

See Arist., Ath. Pol. 51.3.

The metra in Aristotle

339

was probably importing each year a million medimnoi of corn or more,33 the total profit might be quite substantial. The ‘laws of economics’, as we know them, were but imperfectly understood in antiquity. Economics, in so far as it is something more than the oi
koB- or household-management, to which alone the Greek words oi’ konom
ia, oi’ konome&in, oi’ konomik
oB and the like primarily apply, was one of the few branches of thought to which the Greeks made little contribution: what we call economics was to them entirely subsumed under the headings ethics and politics. I do not believe that the citizens of a Greek state who might be called upon to decide the capacity of a standard measure would think much beyond the immediate, practical issue: ‘We shall be mainly buying with this measure, so we had better have it larger rather than smaller’, or the reverse. An important consideration, which cannot be discussed here, is that the vast majority of sales in the Greek world were cash sales, in which delivery of the goods was made on the spot, in return for money: sales on credit, or with deferred delivery, were very much rarer than in the modern world34 and as a general rule35 were not legally enforceable in the courts of Greek cities. A buyer, therefore, unless he was prepared to trust the seller completely, would not normally offer to take, for example, so many amphoras of Thasian wine, according to sample or not, but rather certain specific amphoras which could be designated on the spot. This made precision of measurement infinitely less important than it is with us. And indeed we find an astonishing degree of variation in capacity, even among those ancient containers which look alike.36 33 See Dem. XX 31–2, with the comments of A. W. Gomme, The Population of Athens (1933), 32–3. 34 See on the whole subject F. Pringsheim, The Greek Law of Sale (1950), esp. pp. 90–2, and for a brief summary J. W. Jones, The Law and Legal Theory of the Greeks (1956), 227–31. Cf. L. Gernet, Droit et socie´te´ dans la Gre`ce ancienne (1955), 201–36. 35 Unless, e.g., a credit sale was dressed up as a loan, by the purchaser making formal acknowledgement of the receipt, as a loan, of a sum equal to the unpaid purchase money. It may be that in some Greek states, in the fourth century B.C., most if not all evasions of this kind were impossible: see Arist., EN 1162b 29–31; 1164b 12–15; with 1131a 1 ff.; also Plato, Rep. VIII 556ab (with Laws VIII 849de; IX 915de). 36 Virginia Grace, in an admirable article on ‘Standard Pottery Containers of the Ancient Greek World’, in Hesp., Suppl. 8 (1949) 175–89, at p. 176, has pointed out that ‘even as between jars of the same shape and fabric, and identical height, a difference in dimensions and outline that is scarcely noticeable to the eye can

340

Athenian Democratic Origins

Finally, there is one rather technical metrological problem which must be dealt with: did Greek cities give official specification of their standard measures of capacity by diameter and height, in terms of linear measure, or did they at any rate construct their standard measures in practice according to such formulae? Recent writers have tried to detect an example of the former procedure in late fifth century Thasos,37 and of the latter in Hellenistic Chios;38 but both these attempts are based on misconceptions. I shall discuss the whole question in an Appendix. Here it is only necessary to remark that the assumption of a necessary relationship between measures of capacity and linear measures is entirely unfounded. There has never at any time been any natural relationship, or any legally recognised relationship, between (for example) any British measure of capacity and any volumetric unit derived from linear measure:39 all British measures of capacity, both dry and liquid, are derived from the ‘imperial gallon’, which contains a certain weight of distilled water at a certain temperature under a certain barometric pressure.40 The Greek measures of capacity, like the Roman, were not originally derived from or related to units of linear measure,41 and it looks as if the Greeks were very slow to begin thinking of volumes in terms of cubic units based on linear measures.42 Now until the ordinary educated Greek became used to doing this, specifications of containers by diameter and height were obviously out of the question, even when the geometrical relations subsisting between linear measurements and volume were known in theory for bodies such as the sphere, cylinder, spheroid and cone—this simple fact is often overlooked. As far as I can see, measures of capacity were never regularly expressed in terms of cubic units produce a difference in capacity of about an eighth of the total.’ She gives a reference to two second-century Cnidian amphoras (Plate 19, nos. 7 & 9: see p. 186), both 95 cm. high, and differing by only 3 mm. in diameter (33.7 & 34 cm.), which have capacities of just under 30 litres and about 33 and a half litres respectively. 37

Mabel Lang, in BCH 76 (1952) 18–31. W. G. Forrest, op. cit. in n. 8 above. According to A. E. Berriman, Historical Metrology (1953), 175, the Standards Department of the Board of Trade now accepts 277.418 c.in. as the equivalent of a gallon; but there is no statutory authority for this. 40 Weights and Measures Act, 1878, sec. 15: 10 lb. weight at 628 F. and 30’’ pressure. 41 See F. Hultsch, Griechisch u. ro¨mische Metrologie2 (1882). 42 See the Appendix to this chapter. 38 39

Appendix

341

derived from linear measures, and I do not know of any evidence that they were ever so expressed, except perhaps by mathematicians, before the Christian era. We certainly cannot afford to assume that any Greek city ever defined its measures in this way until unimpeachable evidence of it is produced. I would paraphrase our passage as follows: ‘Rules of justice based on convention and expediency resemble measures. For the official containers used for measuring wine and corn are not the same in all states, but in those where wine and corn are mainly objects of purchase by those citizens who have political power, the containers will be larger than they ought to be, and in states where these commodities are mainly objects of sale by such citizens, the containers will be smaller.’ Like so many of Aristotle’s terse remarks, the passage is obscure to us at first sight, but it would have been common knowledge to the educated and travelled Greek, to whom variation in weights and measures may have been about as familiar as the goodness or badness of the wine or the coffee in this or that country to the experienced tourist of today.

APPENDIX Measures of Capacity 1. Dealing with a new inscription, probably of the late fifth century B.C., found at Thasos in 1947, Miss Lang (op. cit. in n. 37 above) seeks to demonstrate the use of what she calls ‘Hero’s formula for determining the capacity of pithoi’, taken from pp. 202 ff. of the edition by F. Hultsch of the remains of the ancient metrological writers. In the mysterious numerals of the Thasian inscription Miss Lang sees maximum and minimum figures for internal height (60–55 Ionic d
aktyloi [‘fingers’]) and diameter (44–22 ditto) of a standard pithos.43 Taking an average figure of 33 Ionic fingers for what she supposes to be the diameter, Miss Lang makes two calculations, from what she takes to be the maximum and minimum heights, and then strikes a mean between the two, giving a capacity of 311 litres, or (on her reckoning) 8 metretai. 43 It is perhaps worth remarking that the order of the numerals is not 60.55 j 44.22, but 55.60 j 44.22.

342

Athenian Democratic Origins

She then proceeds to make up several imaginary ‘pithos-shapes’, according to the formula: height 55–60, diameter 22–44, which have exactly the right capacity. This reconstruction is open to serious criticism in three different ways. First, we have no right to look here for an application of ‘Hero’s formula’. It is not so much that Hero himself lived some five hundred years later than this inscription,44 but that the passage in question cannot safely be attributed even to him: it comes from a Byzantine compilation, little or none of which may go back to Hero. Hultsch’s incomplete edition of 1864 has been out of date for over forty-five years, and we must consult Heiberg’s edition of Hero, Vol. V (1914), where we can compare it with many more formulae of the same kind. It is not quite clear to me which formula Miss Lang had in mind, but I imagine it is the one used in Hultsch’s §§ 19, 21, which we can now refer to as Stereometrica II 22, 24, and identify as coming from a part of the ‘Heronian’ corpus which has scarcely any claim to represent the work of Hero himself, as Heiberg and others have demonstrated. There can be no justification for picking out of such a work a particular formula for calculating the capacities of vessels—one of several such formulae—and trying to see it behind a document more than a thousand years earlier in date.45 In the second place, it is in principle wrong (see pp. 340–1 above) to expect to find a Greek city giving mathematical formulae for determining capacity in terms of units derived from linear measure, above all as early as the fifth century B.C. I should like to draw attention to an interesting passage in the ‘letter of Eratosthenes’, quoted in the commentary of Eutocius on Archimedes, On the Sphere and Cylinder: ‘ vn dynZs
omeua de kai t a tvn & ygr & m
etra kai jZrvn, & l
ogfiv de oiffl on ’ m
edimnon, ei’ B k
ybon kau
istasuai kai di & to
y toy metrZt Zn Z a tZB ’ ’ in t a to
y tvn dektik a agge& ia, p
oson xvre&i [‘We pleyr &aB anametre& will also be able [by use of Eratosthenes’ new method] to convert wet and dry measures such as the metretes and the medimnos into a cube and measure the capacity of vessels by reference to its side’].46 At whatever date the letter of Eratosthenes was concocted, we find its author, a mathematician, speaking in future terms (dynZs
omeua [‘we shall be able’])—i.e., future to Eratosthenes—of the practicability of stating such quantities as the metre¯te¯s and medimnos in cubic units derived 44 Probably in the second half of the first century A.D.: see O. Neugebauer, The Exact Sciences in Antiquity2 (1957), 178, with the references there given. 45 This is so even if we admit, as we must, that much of the ‘Heronian’ corpus of elementary geometrical writings is ‘merely a Hellenistic form of a general oriental tradition’ (Neugebauer, op. cit. 146). 46 Archim., Vol. III (ed. Heiberg, 1915), p. 90.

Appendix

343

& from linear measure (ei’ B k
ybon . . . kai di a tZB to
y toy pleyr &aB ’ anametre& in), and thus calculating the content of vessels containing these measures. Thirdly, Miss Lang does not seem to have realised either how feeble the Heronian formula is, or that her extension of it, with the height as well as the diameter variable, makes it even more elastic. What the ‘Heronian’ formula does is to take the arithmetical mean between the extreme diameters; but unless this happens also to be the average diameter (and it may be far from that), the sum will not work out right in practice—everything depends on the contours of the vessel. Introducing a variable height naturally makes things even worse. Although we can construct, according to this formula, pithos-shapes having the ‘right’ capacity, we can also devise, just as easily, any number of equally characteristic shapes which have a much greater or smaller capacity. In other words, such a formula would be useless. Now whoever invented the formula in question, and comparable formulae found in the Stereometrica, was certainly not concerned with making jars of a precise capacity (if he had been, he would soon have abandoned or modified his formulae), but with giving a rule of thumb for calculating the approximate capacity of existing vessels. Some of the pot-formulae in the Stereometrica are quite ludicrous.47 Any competent potter, on the other hand, to whom the Heronian formula was available, would soon have discovered by experience that the formula would not give a precise capacity but only a rough approximation; and it is impossible to believe that the Thasians or any other Greek state ever published such a formula as the specification of their standard measures. It is much better to admit that at present we have no means of interpreting the Thasian inscription. 2. In a useful paper entitled ‘Numerical Notation on Greek Vases’, in Hesperia 25 (1956) 1–24, Miss Lang lists 61 inscriptions on vases found in the excavations of the Athenian Agora which are, or can reasonably be suspected of being, indications of capacity. Her careful analysis can only evoke gratitude, but at some points it is open to criticism. For example, her reference (p. 4) to Chian jars ‘manufactured in accordance with minimum specifications of dimensions’ (my italics) is not warranted by the reference she gives, to her earlier paper in BCH, discussed in the preceding paragraph. Even on her own showing, there would be no justification whatever for calling the 47 e.g. II 23 (Hultsch, § 20), where the volume would come out in square feet; I 53 (Hultsch, § 18), where p
odeB and pZxe&iB are intermingled; and II 21 (not in Hultsch), where the po &yB (i.e. o‘ stere oB po &yB) is equated with the k
yauoB, instead ’ ’ of the amwore
yB or amwor
iskoB or 3 Roman modii, as in most other passages in the same work (e.g. II 22–4).

344

Athenian Democratic Origins

specifications ‘minimum’ ones. Again, on p. 6 she says, ‘Complete Chian jars, as was seen above, hold somewhat more than 7 Athenian choes’; but although this is certainly true of some Chian jars, she has not shown, and does not show, that it was true of the majority: of those she herself lists, bearing marks of quantity, three out of six (nos. 10, 30, 58, as against 8, 15, 17) are said to hold just 7 Attic choes,48 and a seventh (no. 2), of which only the neck survives, bears a mark which she accepts as indicating a content of 7 choes. The essential fact is that, ‘as one might expect, there is considerable variation of capacity among amphoras with the same stamped guarantee’ (Miss Lang, p. 4).49 3. In BSA 51 (1956) 63–7, W. G. Forrest, partly under the influence of Miss Lang’s papers, has attempted to explain the construction of two cavities in the Chian wine-se¯ko¯ma which he republishes. He suggests that the upper half of each cavity should be considered as a cylinder, and the lower half as half a spheroid; and he then supposes that by working back from the desired capacities the maker calculated mathematically the dimensions he required. This explanation is open to a whole series of objections, even if we make the very doubtful assumption that Archimedes’ formula for the volume of a half-spheroid was known and used in Chios at the date of this inscription (the late third or early second century B.C.). The cavities are quite differently shaped, and in neither case is there anything to suggest that their form was conceived as a cylinder plus a spheroid: the outlines, in their very different ways, are homogeneous throughout, and any attempt to divide the containers into two quite different halves is merely arbitrary. The upper halves, as Forrest admits, are not really cylinders, and the lower halves are very far from being half-spheroids—it is the whole of the smaller cavity, and roughly the lower two-thirds of the larger, which might be conceived as half-spheroids! The ‘diameter’ which the application of the formula is supposed to give in each case is of course the diameter of the ‘cylinder’ (or rather, the average diameter of the quasi-cylinder), which is also the base of the half-spheroid; but in the case of the smaller cavity that base is 15.2 cm ¼ 8.2 fingers instead of the 9.2 fr. ¼ 17 cm given by the calculation. If the maker was 48

Or 8 Chian choes, if Miss Lang is right about the Chian standard, as she may be—although the fact that Chian coins stood to Athenian ones in the ratio 7 : 8 is not, to my mind, relevant in considering standards of capacity. 49 Virginia Grace, in Hesp. 3 (1934), at pp. 296, 303–4; and in Hesp. Suppl. 8 (1949) 182, describes six reconstructed Chian jars of the third quarter of the fifth century B.C. as all having a capacity of about 22 litres. According to Hultsch’s figures, accepted by Miss Lang (pp. 2–3), this would be almost 6 and three-quarters Attic choes; but the figure was based on measurement with wheat, and if, as it seems, two of Miss Grace’s amphoras are Miss Lang’s nos. 8 & 15, they apparently hold 7 choes 3 kotylai or a fraction more. See also D. A. Amyx in Hesp. 27 (1958), at p. 194.

Appendix

345

trying to use this formula, he certainly misapplied it; and (unless his errors happened to compensate for each other) he could only have got the receptacle right in the end by precisely that method of trial and error which Forrest thinks ‘expensive in time and material’. Finally, the calculation would not give the maker the one important measurement which he really needed from the start: the diameter of the cavity at the top  19.7 cm ¼ 10.6 fr.50 in the case of the smaller cavity (where the calculation gives 9.2 fr. ¼ 17 cm.), and in the other case 28.4 cm ¼ 15.4 fr. (instead of 16 fr. ¼ 29.6 cm., by the calculation). It is far more likely that the maker began with a rough idea of the desired dimensions,51 and then proceeded to increase the diameters by small degrees, until he obtained the right capacities, tested with water. There is no reason for calling such a method ‘expensive in time and material’, let alone speaking in exaggerated terms of ‘an improbable situation in which the potter turned out pots with hit-or-miss abandon in the hope that some would chance to be accurate measures’ (Miss Lang, op. cit. in n. 37 above, p. 23). 4. In her article on the Thasian metrological inscription (op. cit. 23–4), Miss Lang says, ‘We should assume even without evidence that the laws introducing or enforcing measures of capacity must have included specifications of interior dimensions.’ We have already seen (pp. 340–1 above) that no such assumption can possibly be justified. But Miss Lang goes on to produce what she takes to be ‘evidence confirming this assumption’, from IG II2 1013, lines 18–25, where she finds an indication that there was a particular specified diameter for these vessels, ‘constant for at least a certain range of measures’. The lines in question direct sellers of certain specified fruits (Persian nuts etc.) to use ‘a measure holding three half-choinikes of grain levelled off, selling them heaped up in this measure, which is to have a depth of five fingers and a lip one finger wide’; and similarly, the sellers of certain other commodities (green almonds etc.) are to sell ‘in a heaped-up choinix measure twice the size of the one mentioned above, with a lip three halffingers wide’. Now this is strange. The fruits are to be sold heaped up, but in a vessel the capacity of which is determined by the amount of 50 Actually, anyone intent on seeing here some application of the formula for the volume of a half-spheroid has a much better alternative than the one Forrest suggests: he might assume that the whole of the smaller receptacle (the only one which survives complete) was treated as half a spheroid, so that the complete formula becomes 2=3  11=14  14  d2 ¼ 770 c.fr. This gives d ¼ 10:25 fr: ¼ 19 cm., a close approximation to the actual diameter of the cavity at the top (19.7 cm. ¼ 10.6 fr.), which is the vital measurement. But this hypothesis seems to me to be ruled out by the shape of the other receptacle. 51 Just possibly with the aid of some rule-of-thumb formula, if only p ¼ 3, rather than 1=4p ¼ 11=14.

346

Athenian Democratic Origins

grain which it holds levelled off. The amount which the vessel will hold up to the level of the rim, namely 1 and a half choinikes (measured in grain), is fixed, but above this there is a ‘heap’ which is to some extent variable: the greater the diameter of the rim, including the lip, the greater the ‘heap’. The one dimension we must know, therefore, is the diameter at the rim, but this is not stated: what we are given is the depth, and the width of the lip, without even the corollary (essential in the absence of a diameter) that the container is to be cylindrical—a feature which, the depth and capacity being known, would fix the diameter of the rim also. (By making the vessel not cylindrical, but broader at the top and narrower towards the base, one can broaden and thus increase the ‘heap’ even if the height of the vessel is fixed.) Miss Lang argues that her theory of a prescribed diameter, ‘constant for at least a certain range of measures’, is ‘confirmed by line 24, where a double measure is specified, since doubling the height while keeping the diameter the same doubles the capacity’. But we have no right to expect verbal precision from Athenian inscriptions such as this. The word diplas
ioni in line 24 would naturally be taken to mean ‘of 3 choinikes and 10 fingers deep’. And with a constant diameter all measures would have to be cylindrical; yet a contemporary Athenian measure52 which, though apparently a private container, is validated by an official stamp (doubtless under the provisions of lines 63–7 of our inscription), is not cylindrical but slopes inwards perceptibly from top to bottom. Everything points to there being no official definition of the Athenian measures of capacity in terms of linear measure.53 5. How did Greek states decide what the capacity of their standard measures ought to be? The answer is that we do not know for certain. The vital step, I submit, was the making of official se¯ko¯mata, the capacity of which would probably in most cases be copied from the standard measures of some other city (with or without variation), but would in any event be decided by the magistrates concerned, subject to the final approval of the supreme governing body of the city, without reference to any linear measure. Many Greeks, of course, must have known that their se¯ko¯mata, and therefore the measures in everyday use, were not precisely the same as the corresponding measures in other cities using the same standard; but this, I have suggested, would not trouble them as long as the discrepancies worked out, on the whole, for ‘their own’ advantage—that is to say, for the advantage of the majority of the citizens having political power. 52

Published by Margaret Crosby in Hesp. 18 (1949) 108–13, with Pl. 5, nos. 1–2. The official Athenian containers of the fifth century, published by O. Broneer, were evidently tested with grain, or something of the sort, before the pots were fired: see Hesp. 7 (1938) 222–3; Hesp. 18 (1949) 111 n. 15. 53

Afterword

347

In time—perhaps only in the Christian era, under Roman influence?—some Greeks did come to accept equivalents for their measures of capacity in terms of cubic units of linear measure. Only when this happened would the application of mathematical formulae in the construction of pots etc., designed to ensure precise capacities, become possible. And I know of nothing to suggest that the practice ever became widespread or had any importance. It may be significant that in the Heronian corpus, for example, it is only in those compilations which have least claim to represent the actual work of Hero, i.e. those printed in Volumes IV and V of the Teubner edition, that we find—as far as I can see—such expressions as o‘ stere oB po &yB [‘cubic foot’], or even p
odeB in the sense of cubic feet.54

AFTERWORD No new interpretation of the Aristotelian text is known to us. This Afterword relates rather to the archaeological data discussed by Ste. Croix. A corpus of the evidence from the Athenian agora is provided by Lang and Crosby. They discuss (42– 8) the practical difficulties of making accurate measuring vessels out of clay: adjustments were made prefiring by the addition of strips of clay inside the vessel, but the extent of shrinkage caused by firing must just have been guessed. They note (ibid.) that a preference for particular diameters is observable in measuring vessels, and they explain this by use of a precursor of the ‘Heronian formula’; but there would obviously be a tendency for measuring vessels to have standard forms by whatever system they were made. An Attic black-figure amphora (c. 510–500 B.C.) has recently been published that bears the dipinto dem
osioB (SEG xlvii 108). It seems likely to be the earliest instance of a public measuring vessel for liquids, though its capacity is not in fact an attested measure. Grayson touches on many related topics (e.g. the Metronomoi, 463–74; IG II2 1013, 295–7), though his concern is with weights, not measures. He notes that complaints about the possibility of swindling through false measure were commonplace: Ar. Eq. 1009; Nub. 640; Thesm. 347–8; Plut. 435–6. The main thrust of his thesis is to downplay the importance of scientifically precise weighing in early Greece. He 54 See again n. 45 above. In the present state of our knowledge it is impossible to say how soon this general oriental (i.e. Babylonian and Egyptian) tradition, associated with numeral scripts unintelligible to nearly all Greeks, penetrated the thinking of the Greeks to an extent sufficient to lead to widespread interest in the ‘Heronian’ formulae.

348

Athenian Democratic Origins

surveys surviving Attic weights in detail (66–163), and concludes (162) ‘Permitted variation was exceedingly high . . . ‘‘Demosion’’-marked weights were marginally more carefully made, while those on the heavy (commercial) standards varied most of all even where marked ‘‘Metronomoi’’ or ‘‘Agor[anomoi]’’.’ This conclusion needs to be read with the caution of Lang and Crosby (17) that surviving weights may be lighter or even, in the case of metals, heavier than they originally were. Figueira, 302–15, discusses the adoption of Athenian measures by allied states in the 5th c.; on p. 305 he appears to accept Lang’s application of the concept of cubic measure to the Greek world. In correspondence with Ste. Croix in 1976, Dr R. P. Duncan-Jones wrote ‘Although there is bound to have been plenty of finagling in the use of capacity measures, I am sceptical of our ability to deduce specific examples from the surviving evidence (except where it is categorical and explicit, as in the Gregory letter to Pantaleo [p. 333 above]). POxy 136 seems to me a straightforward indication about the use of two different artaba standards, one for exacting rent from tenants, another smaller one for internal accounting purposes on the estate concerned. Though it may have been harsh, it was not exactly cheating to say that tenants had to pay their rent by a designated large artaba measure. The state seems to have habitually used an artaba measure for tax-purposes which was larger than the commonest one in domestic use [Chiron 6, 1976, 257– 60].’ He went on to point out that there were so many standards for capacity measure (‘with the 20-odd Egyptian artabae as the reductio ad absurdum’—see his article of 1979) that it is difficult to adduce ‘diversity between surviving unmarked measures as evidence of chicanery’. The book by Kula is a splendid comparative study by a distinguished Polish economic historian. Duncan-Jones, R. P., ‘Variation in Egyptian Grain-measure’, Chiron 9 (1979) 347–75 Figueira, T., The Power of Money (1998) Grayson, C. H., Greek Weights, diss. Oxford 1974 Kula, W., tr. R. Szreter, Measures and Men (1986) Lang, M., and Crosby, M., The Athenian Agora X: Weights, Measures and Tokens (1964)

9 How Far was Trade a Cause of Early Greek Colonisation?

I want to begin by doing two things. First, I shall discuss very briefly the other possible causes of the colonisation movement. And secondly, I shall try to analyse the concept of trade, and bring out the fact that when we speak of trade we may be referring to one or more of three different things. First, then, the other possible answers to the question what caused the colonisation movement. To my mind, all the factors which have any consistent weight at all really boil down to one: land hunger due to population pressure. The other factors do not amount to much more than aggravations of that. Fortunately, we need waste no time over this, because all scholars, I think, have admitted that the great movement of early Greek colonisation was above all agrarian, directed towards the seizure of good agricultural land. Even those who put much more emphasis than I do on trade as a cause of colonisation admit that land hunger was the prime cause in the majority of cases. Most of the evidence about early colonisation is not literary and tells us little or nothing about the motives for colonising. We have only one reasonably detailed foundation story which we can trust: this is Herodotus on the foundation of Cyrene.1 The bulk of the evidence about the early colonies, being purely archaeological, does not tell us much about causes or motives; but there are, as we shall see, some scraps of what I may call ‘circumstantial evidence’ which do help. By ‘circumstantial evidence’ I mean features of the site, or the fact that colonists went there particularly early, in preference to what might seem to us more obvious places, or did not go there until later than we might have expected. All this evidence together does not amount to very much, but fortunately, such as it is, it does virtually 1

Hdts IV 150–8; 159–65.

350

Athenian Democratic Origins

all point in the same direction: to land hunger, aggravated by various economic, social, political and legal factors, as the main driving force of the movement. May I just pick out three of the more significant scraps of evidence. 1. Cyrene, colonised from the island of Thera about 630. There is no need to go into details here, because the whole story is in Herodotus. I need only emphasise that the background is a long period of drought on Thera, and that the colony was clearly not at all due to adventurousness or love of exploration or anything like that. On the contrary, in order to send out a mere two penteconters full, conscription had to be resorted to, and the men were chosen by lot and forbidden to return. They did try to come back almost at once, whereupon the rest of the population came down to the beach and drove them away. (It was the usual story: ‘I’m all right, Jack.’) And when the emigrants did eventually settle down, it was at Cyrene, an excellent agricultural site, well inland from the sea and high above it. 2. My second piece of evidence is strikingly like that last incident. In his Greek Questions no. 11,2 Plutarch is out to tell us who ’ the aposwend
onZtoi were. He explains that when the earliest colonists of Corcyra, who were Eretrians, were driven out by the Corinthians (this was about 733) and sailed off home, their fellowcitizens came down to the beach with slings and prevented them from landing. Again, it was a case of ‘I’m all right, Jack’. The ’ unfortunate colonists thus got the name of the aposwend
onZtoi, ‘the men who got slung out’. Evidently they had been sent out to Corcyra, like the Theraean colonists to Cyrene, because their home territory could not support the whole of its population. 3. The third piece of evidence is particularly striking, because it concerns one of the key sites for trade in the whole area of GraecoRoman civilisation: Byzantium. Surely, one might think, anyone going to found a colony in the region of Propontis or the Black Sea coast would go for the site of Byzantium first, if he had the least interest in developing trade, for this is the ideal position for a great seaport. Indeed, as Greek sailors knew very well, the current through the Bosphorus naturally brings one over to the North West shore and the Golden Horn: it is difficult, as Polybius3 says, to put in at all at Chalcedon (Calchedon), on the opposite 2

Mor. 293ab.

3

Polyb. IV 39 & 43–4; cf. Strabo VII 6.2, p. 320.

Trade and Colonisation

351

side of the strait; and even the tunny shoals passing through, according to Strabo, Pliny and Tacitus,4 were carried by the current in such a way as to give Chalcedon a wide berth, but provided a lucrative source of revenue to the Byzantines. Yet, acording to a tradition which there is not the least reason to doubt, the Megarians, who founded Byzantium about 668, settled seventeen years earlier at Chalcedon. When the Persian Megabazus heard that Chalcedon had been founded before Byzantium, he remarked (Herodotus tells us)5 that the colonists must have been blind. The explanation, however, is simple: what the emigrants were looking for was a good agricultural site, and from this point of view (and this alone) Chalcedon is superior to Byzantium. (It does not seem likely, by the way, that the minor copper deposits on the neighbouring island of Chalcis can have been worked before colonisation.)6 Those are some characteristic scraps of evidence for the essentially agrarian character of some of the early colonies. I shall pass very quickly over the other possible causes of the colonising movement, because, as I said earlier, I think they did not do much more than aggravate the land hunger which everyone admits to have been the basic factor. There is, for instance, first the inheritance system, on which some historians have laid much stress.7 On the death of a Greek, all his property was divided equally among his sons, and he had no right to make a will preventing this. So one might expect a constant fragmentation of landed property, with the holdings growing smaller and smaller by division. But I do not think we need attach much independent weight to this: such a system of inheritance is almost universal the world over, except in countries where something like a feudal system has prevailed and there has been a need to keep fairly substantial pieces of land undivided, so that they may support a warrior. Greek inheritance law was not at all unusual. And if the population was increasing rather rapidly from the early eighth century onwards, as it appears to have been, and the process of fragmentation was going on apace, then we should simply find ourselves with a contributory cause of the land hunger I have already accepted as the primary cause of colonisation. 4 5 7

Strabo, loc. cit.; Pliny, NH IX 50–1; Tax., Ann. XII 63. 6 Hdts IV 144.2. See Roebuck, ITC 103–4 & n. 102. e.g. J. B. Bury, Hist: of Greece3 (ed. R. Meiggs), 86–7.

352

Athenian Democratic Origins

Or, secondly, we might direct our attention to the aristocratic political regimes which in and around the eighth century concentrated all political power and a great proportion of the available wealth in the hands of a small group of hereditary nobles, Hesiod’s & dvrow
agoi basilZeB [‘gift-devouring kings’]. Certainly one independent cause of colonisation may have been the desire to escape to another region where one would not be entirely at the mercy of a narrow oligarchy of birth. But peasants all over the world down the ages, preferring above all to stick to their ancestral land, have put up with oppression and exploitation by the ruling few, who have commonly chastised them with whips if not with scorpions; and in my opinion the way in which the oppressive aristocracies of the eighth and seventh centuries contributed most to the colonising movement was by engrossing much of the best land, by force or unjust judgement, and thus increasing the number of the land-hungry. Thirdly, there are the cases in which some aristocrat is said to have led out a colony because he was discontented and frustrated at home (like Dorieus of Sparta, the younger brother of King Cleomenes)8 or was under a cloud (like the Corinthian Archias, the founder of Syracuse).9 This may account for some of the leaders of expeditions, and even for odd members of their followings; but we should still need a more universal motive force behind the bulk of the emigrants. We might, I think, allow independent status to a fourth type of colonist: emigrants driven from their homes for one reason or another, by foreign invasion or by internal strife. But colonists of this class will have been few: the most obvious examples are the Messenian (though not the Chalcidian) founders of Rhegium,10 and the Spartan Partheniae who went to Taras,11 respectively at the beginning and the end of the first Messenian war in the late eighth century. *** I turn now to the concept of trade. I want to suggest that concealed in this word there are three very different types of activity, and that if one speaks of a ‘trading colony’ one must be quite sure which of these one has in mind. 8 10 11

9 Hdts V 42.2. See T. J. Dunbabin, The Western Greeks 15 & n. 3. Strabo VI 1.6, p. 257; Diod. VIII 25.2. Arist., Pol. 1306b 22–31; Strabo VI 3.2–3, pp. 278–80.

Trade and Colonisation

353

In making this analysis of the concept of trade, I shall be taking into account above all the objectives of trade. It is here above all that political factors enter: economic issues were if anything even more inextricably tangled up with politics in the Greek world than they are today, and if we want to form clear ideas about state policy in regard to trade (which is our special objective at the moment) we must be quite sure that we understand in what light trading issues were seen by the Greeks, more particularly (I would emphasise this) by those Greeks who enjoyed political power in the cities. What we call broadly ‘trade’, I suggest, could present itself to the Greeks in one or more of three aspects, which in a way are all connected, but are also separable—and must for our purposes be separated. First and secondly, trade can mean imports or exports: it can of course mean both at once, and naturally in the long run imports must broadly balance exports and vice versa. But sometimes we shall find the policy of a Greek city concerned entirely with the first aspect of this process, ignoring the second (with imports, ignoring exports). And thirdly, trade can be viewed solely from the profit angle: buying cheap and selling dear, with the aim of making a profit on the exchange. For convenience, I shall call these three aspects of trade ‘imports’, ‘exports’, and ‘commercial exchange’. Individual traders and merchants and shop-keepers, of course, were interested primarily in the third aspect, to which the first or second or both might contribute. But a Greek city collectively, I would maintain, was not interested in the very slightest degree in whether ‘its own’ traders and merchants made profits through commercial exchange. This sounds almost paradoxical to us, but there were very good reasons for it, as I shall explain. Nor— and this is just as hard to understand at first—did any Greek city known to me ever show any concern whatever about exports: whenever we can see foreign trade becoming a factor in Greek city policy, it is always the import aspect which is predominant. The Greek would have laughed at our notion that a balance of trade is to be considered ‘favourable’ when exports exceed imports in value: what the Greek citizen was interested in was getting what he wanted to make life more pleasant, and this of course meant being able to procure a full supply of necessaries, and as many luxuries as he could get on top of that; and if, like Athens in the fifth and fourth centuries, you could pay for all these things partly with silver which you were lucky enough to be able to dig up from

354

Athenian Democratic Origins

your own territory, or even at times with imperial tribute—well, you were fortunate indeed. A few Greeks did realise that if you want to have imports you must have sufficient exports to balance them; but the only passages I have been able to discover on this topic (in Plato, Aristotle and Plutarch)12 show plainly that it is the imports which are in the forefront of the picture: the exports come in simply as the economic consequence of the need for essential ’ arkeia, complete self-sufficiency, was the ideal; but, imports. Ayt
as Plato says in the Republic,13 it is virtually impossible to find a place where no imports are needed, so we must have imports; but if the trader goes out empty-handed he will come back emptyhanded too, so our city must produce enough of a surplus to pay for its imports. Aristotle14 rather grudgingly says that a city ought to have access to the sea, in order that there may be a good supply of necessaries; and then he goes on to show that because every state must be able to import commodities it does not itself produce, therefore it must export the surplus of its own products. But he does not at all like the idea of deliberate production for export: a city, he says, should not act as a merchant for others, but only for itself.15 The writer of the Pseudo-Xenophontic Athenaion Politeia, the so-called ‘Old Oligarch’,16 lays great stress on the fact that a city which has command of the sea can prevent other cities rich in timber for ship-building or in iron or copper or flax from exporting these products, except to itself. But the point here is the ability of the ruling power to get all it wants, and deprive its enemies of what they want. The Old Oligarch is looking at trade from the point of view of the buyer, the receiver of goods: if he notices that the rulers of the sea can stop a given state from exporting something, it is because he is thinking of the other end of the transaction, of the desire of the ruling power to obtain these goods for itself or deny them to its enemies. I would particularly emphasise that although the Old Oligarch is very conscious of the advantages of thalassocracy, he never considers using it as a means of obtaining a monopoly of (or even preferential conditions for) the sale of one’s own products. And there are many other passages one could quote. When merchants are spoken of as a good thing, as 12 13 14 15

[e.g. Pl. Leg. 847d; Arist. Pol. 1327a 11–31; Plut. Sol. 22. 1.] II 370e–71a. And see Laws IV 705ab for a moral objection to export trade. Pol. 1327a 19, 25–8; cf. 1257a 32–3. 16 Pol. 1327a 27–8. II 11–12.

Trade and Colonisation

355

by Xenophon,17 it is their activity as importers which is noticed. In the Funeral Oration, Thucydides18 makes Pericles congratulate the Athenians because owing to the greatness of the city ‘the produce of all the world comes to us, and to the Athenian the good things of other countries are as familiar a luxury as those of his own’. And so on.19 Of course, we must not make the old familiar mistake of treating the outlook of Plato and Aristotle and their like as being necessarily the outlook of ‘the Greeks’; but I know of no evidence to suggest that practical politicians thought more clearly or more effectively about economic problems of this sort than the intellectuals did. The Athenians even charged customs duties on their own exports at the same rate as on imports—2% in the fourth century. It is the accumulation of negative evidence which to my mind is ultimately decisive. City legislation and treaties (two obvious sources of evidence) often concern themselves with imports, especially of staple necessities, but never—except to restrict them— with exports. In very strong contrast is the keen interest shown by a real ‘trading city’, Carthage, to protect the interests of her own merchants by excluding the foreign trader from her dependencies, or severely limiting his activities: this appears clearly from the two earliest treaties between Carthage and Rome,20 when Carthage was the stronger power. No similar provisions exist, as far as I know, in any surviving treaty between Greek cities.21 Nor is there a single scrap of evidence, as far as I have been able to discover, that any Greek city ever took any action to increase its exports. Finally, and perhaps most important of all, there is a source of evidence which seems to have been entirely neglected, even by Hasebroek:22 the origins of wars between Greek states. Could wars ever be fought for objectives that can be called in any sense ‘commercial’? I have been trying for some time to collect the evidence: there are some 17

Hipparch. IV 7. Cf. also Plut., Lysand. 3 (Ephesus, 407). II 38.2; cf. Hermipp. fr. 63 K/A, ap. Athen. I 27e–28a. See n. 12 above. 20 Polyb. III 22.5 & 23.1 ff. (508 B.C.); 24.2 ff. (348 B.C.). Contrast I 62. 8–9; 63.3 (241 B.C.) and XV 18 (201 B.C.). See T. Frank, Econ. Survey of Ancient Rome I 6–8, 35–6, 68, 96. 21 Or, as far as I know, in any treaty dictated by Rome. The Treaty of Apamea between Rome and Antiochus III, in 189/8 B.C., contained a commercial clause in favour of Rhodian imports (Polyb. XXII 23. 17 ¼ XXI 42.17). 22 Hasebroek, TPAG. 18 19

356

Athenian Democratic Origins

interesting general statements about the principal reasons for the outbreak of wars, and there are accounts of the origins of particular wars. In almost every case (this seems to surprise many people) the evidence relates to just one kind of war: wars about the ownership of land, whether on a large scale, like Sparta’s Messenian wars of the eighth and seventh centuries, or petty squabbles over bound& o
‘ rvn).23 Of wars having anything in the nature of aries (peri gZB commercial objectives there is not a suggestion anywhere, as far as I know, until well into the Hellenistic period—and even then the example I am thinking of is unique.24 In the few cases in which we are informed about the origins of wars between supposed pairs of ‘commercial competitors’ (Athens and Megara, Corinth and Megara, Samos and Miletus, Chalcis and Eretria) we find that the cause of the dispute, when it is mentioned, is the ownership of land. A state at war with another may attack a third party in order to compel it to deny to an enemy some essential product, as Athens sent an expedition to Sicily in 427 to stop the export of Sicilian grain to the Peloponnesians.25 But we never hear of a war fought in order to capture markets for exports, or of a war arising out of commercial rivalry. The common assumption that wars between Greek cities could arise from causes such as these is largely due, I think, to the hypothesis put forward some forty years ago by Cornford and Grundy, and still by no means dead, that the real cause of the Peloponnesian war was commercial rivalry in the Western trade between Corinth and Athens—a theory which rests entirely on a priori assumptions and is unsupported by one single piece of solid evidence [cf. OPW 214–20]. Greek cities, then, (and by Greek ‘cities’, of course, I really mean those who had political power in the cities, the governing class) might often be keenly interested in imports, especially imports of essential commodities such as corn, timber and metals; but they were not in the least interested in increasing their exports or in furthering what I have called ‘commercial exchange’, except perhaps in so far as they might profit by the customs duties they charged at their ports. The main reason is very simple: the governing class in all Greek cities, with one or two possible exceptions (not including Athens or Corinth or Aegina), was always predominantly agrarian in character. Merchants as such were 23 24

See [OPW 218–20 and] n. 12 above. 25 Polyb. IV 47.1 (220 B.C.). Thuc. III 86.4.

Trade and Colonisation

357

politically insignificant. They were not organised, and they exercised virtually no influence in politics, individually or collectively. No single statesman is known to have been a practising merchant, and no merchant is known to have played a prominent part in politics, even at Athens. The merchants were not all, as Hasebroek absurdly claimed, both non-citizens and men of little or no property; but the evidence we have (mainly derived from the orators) strongly suggests that in the fourth century anyway the majority of merchants were not men of much substance. Their influence on politics, as merchants, was certainly infinitesimal. Moreover, they would tend to be away from their cities during the sailing season, which was also roughly the campaigning season, when most of the vital decisions would have to be taken. Greek city governments did not identify themselves with ‘their own’ merchants—or, for that matter, with ‘their own’ craftsmen and artisans. We must not think of any Greek city as enjoying a ‘commercial monopoly’, just because goods made within its walls dominated a particular market, as Corinthian pottery dominated the Western market from the end of the eighth century to the middle of the sixth, and Athenian pottery thereafter. Similarly to speak of the ‘merchant marine’ of a particular state or states (and we have all done it) is to be guilty of a misconception: trade was as detached from the flag as it could well be. Individual merchants and groups of merchants would of course be rivals, but we must not project this rivalry on to the cities in which they lived. Contrary to what we know so well of many modern states (and even some late mediaeval and Renaissance trading cities), rivalries between merchants living in Greek cities (of which they often were not even citizens) were not reflected at the political level, as far as my knowledge of the evidence goes. It is true that I have been utilising evidence drawn largely from the fifth and fourth centuries. Is it fair to read all this back into the archaic period? I think for present purposes it is; because, first, trade was certainly much more developed in the fifth and fourth centuries than in the eighth, seventh and sixth, for which we have virtually no evidence except the archaeological; and secondly, if merchants had scarcely any influence in the democratic cities of the fifth and fourth centuries, they would have been even less influential under the aristocratic regimes of the archaic period. When people ignore the fifth/fourth-century evidence, it is not because they have better evidence for the archaic period (as I said, there is

358

Athenian Democratic Origins

almost none) but because they have got used to thinking in entirely anachronistic terms, making use of mediaeval and modern analogies, and talking about ‘merchant classes’ and ‘commercial policies’ and so forth. In a recent article which makes an outstanding contribution to the archaeology and history of the eighth and seventh centuries,26 Mr John Boardman remarks that the Lelantine war between Chalcis and Eretria in the late eighth century cannot have been primarily, as Strabo and Plutarch say or imply,27 a struggle for the plain of Lelanton lying between the two cities, which he describes as ‘a handful of vineyards in what amounted to Chalcis’s back garden’: he thinks this may have been an excuse, but not a reason, and he wishes to see ‘commercial jealousies’ as the real reason behind the war. I would suggest that the Lelantine plain (which, if I am not mistaken, is the largest and most fertile plain in Euboea) was not merely the most likely source of friction between the two states (which normally kept on good terms with each other)28 but was actually by far the most important thing they could choose to fight about, and that no reference to ‘commercial jealousies’ between two Greek cities can make any sense at all, especially in this early period, when all political power was concentrated in the hands of a land-owning aristocracy. And I should like to draw attention to an interesting event which occurred about 545/4: according to Herodotus,29 the Phocaean emigrants wanted to buy the Oenussae islands from Chios, but the Chians refused to sell, ‘fearing that the islands would become an emporion, and that their own island would consequently be shut off’. The only possible explanation is that the Chians were thinking entirely in terms of imports to Chios: they did not want the flow of goods to their own city to be diverted to a new Phocaean tradingstation on the Oenussae islands. We ourselves might consider it an advantage to Chios to have an additional flourishing emporion nearby; and indeed it would have been an advantage if the Chians had been at all concerned about their exports—in that case, the more emporia there were within easy reach, the better. But if the Chians were thinking in terms of imports (and on that assumption 26

In BSA 52 (1957) 1–48 at p. 27. Strabo X 1.12, p. 448; Plut., Mor. 153. On the Lelantine war, see also W. G. Forrest, ‘Colonisation and the Rise of Delphi,’ in Historia 6 (1957) 160–75. 28 29 Strabo, loc. cit. I 165.1 27

Trade and Colonisation

359

alone), their refusal is intelligible, because a Phocaean emporion on the Oenussae islands would merely serve, from the Chian point of view, to intercept goods which might otherwise have gone directly to themselves, and raise their price in Chios—if they ever got there at all—by the addition of the Phocaean middleman’s profit. We are now in a position to begin to look for ‘trading colonies’. As I have said, it is generally admitted that land hunger was by far the most important cause of colonisation. And I do not myself know of any direct, positive evidence, from good or bad sources, that any of the early colonies was founded with a view to trade in any of my three senses, although in certain cases there is strong circumstantial evidence, and of course even agricultural colonists must normally have been guided in their choice of site by merchants who had been trading previously in the area concerned. Certainly there is nothing in any of the foundation legends to suggest that a colony was founded by merchants, or primarily for commerce. Thucydides30 describes the Phoenicians (rightly or wrongly) as occupying headlands and islets on the coasts of Sicily ‘for the purpose of trading with the Sicels’; but he never says such a thing of the Greek colonies of which he proceeds to give a description. I think this is significant. Moreover, if Greek state policy was as I have described it, then can we imagine any reason why any Greek city should ever want to send out a ‘trading colony’? If we recall our threefold classification of ‘trade’, then we might conclude, assuming our principles are sound, that a Greek city might send out a colony in order to ensure supplies of some important commodity like corn, but that it would never found colonies in order to provide itself with markets for exports, or to promote what I have called ‘commercial exchange’. When people speak loosely of ‘trading colonies’, as they very often do, I think they nearly always have in mind the second and third of these activities, the very two which seem to me to be excluded on historical grounds. In fact I know of no case in which there is explicit evidence proving beyond doubt that the object of a foundation was trade for the benefit of the mother city, except for some of the Athenian colonies and cleruchies of the fourth century; and here the evidence is only for trade in the first of my three senses: trade as imports, the supply of important commodities to the mother city—in this case, corn. 30

VI 2.6.

360

Athenian Democratic Origins

The best evidence is an Athenian inscription dated 325/4:31 here, the Athenians resolve to send a colony to the Adriatic to ’ provide its own emporia and supply of grain (empor
ia oi’ ke
ia kai [sit]opomp
ia), with a naval base to guarantee protection against Etruscan piracy and a fleet for the benefit of all seafarers, Greek and non-Greek. Were there ‘trading colonies’ in this or either of my other two senses in the great period of colonisation: the eighth, seventh and early sixth centuries? Let us take as a test case Syracuse, founded from Corinth about 733. (Syracuse is a good test case, because if it turns out not to have been a trading colony it is unlikely that we need consider the claims of any of the other foundations, in Sicily anyway, unless there is some particular reason to look for a trading motive.) Blakeway and Dunbabin,32 to mention only two of the most prominent archaeologists who have emphasised the importance of trading motives in colonisation, had no doubt that the primary object of the Bacchiad founders of Syracuse was to promote trade. Insisting, rightly, that ‘the colony was an official venture of the Corinthian state’, Dunbabin declared that ‘the main object of the rulers of Corinth can hardly have been other than to form a trading colony.’33 (Incidentally, Dunbabin adds, ‘The interest of the Bacchiads in trade is testified by ancient authorities as well as by the wide distribution of Protocorinthian vases’; but he gives no references to these ancient authorities, and why the wide distribution of vases made at Corinth should testify to its rulers’ interest in trade, or indeed what ‘interest in trade’ means in this connection, I do not know.) I suggest that there is probably no truth at all in this theory about the origin of Syracuse. Instead of proceeding from musthave-beens, let us look at the few small scraps of evidence we have about the foundation of Syracuse. First, there is Aithiops and his honey-cake. Athenaeus34 refers to (though he does not actually quote) a poem of Archilochus describing how Aithiops, one of the settlers, bartered his lot of land for a honey-cake on the voyage out, and thus became a byword for Tod II 200 ¼ Harding 121, lines 165 ff. A. Blakeway, ‘Prolegomena to the Study of Greek Commerce with Italy, Sicily and France in the 8th and 7th centuries B.C.’, in BSA 33 (1933) 170–208; Dunbabin, op. cit. (in n. 9 above), ch. i. 33 34 Op. cit. 15. IV 167d. 31 32

Trade and Colonisation

361

prodigality. This shows that the settlers had land allocated to them before they started out; and it was the giving away of this land which made Aithiops into the proverbial spendthrift. This story is far from proving that the colony was essentially an agrarian one, but it surely suggests this. Secondly, there is the fact recorded by Strabo35 that the bulk of the colonists came from the inland village of Tenea. Again, this does not prove that Syracuse was not founded as a trading colony, but again, it suggests it. Thirdly, we know from Herodotus36 that political power at Syracuse in the early fifth century was held by the Gamoroi, evidently a body of landowning aristocrats; and it seems likely that the Gamoroi were the original ruling class. Again, it looks like an agrarian colony. My fourth piece of evidence seems to me a strong one, although I have not found it used in this connection before, and as far as I know it has been entirely overlooked by those who would see Syracuse as a trading colony. It is simply this: that according to Thucydides,37 Archias and the settlers began by driving out the Sicels by force from the island on which the inner city ‘ p
‘ ent
’ oB) stood in Thucydides’ day (that is to say, (Z oliB Z Ortygia)—an action which is also recorded of the colonists at Leontini and Locri, and which probably, as Dunbabin says,38 occurred at nearly every Sicilian and South Italian site settled by Greeks. At least half the Greek colonies in Sicily and South Italy were built on sites previously occupied by native villages, and Dunbabin thought it likely that most were, the one apparent exception being Selinus, where there are no pre-colonisation remains. Now I cannot believe that colonists who came to trade in the usual sense of the word would begin by making war on their potential customers. Does not this confirm the view which we might have reached anyway, on the basis of what I said earlier about the attitude of Greek states towards trade, that if Syracuse was in any sense founded as a trading colony, then ‘trade’ must be understood in the first of my three senses only, and we must think of a foundation partly designed to supply the mother city with some essential commodity?—which in this case could hardly be 35 37

36 Strabo VII 6.22, p. 380. VII 155. 38 VI 3.2. The Western Greeks, 43.

362

Athenian Democratic Origins

anything but corn, Sicily having nothing much else I know of to offer at this date, except perhaps slaves; and one does not need to found a colony in order to trade in slaves. I personally doubt very much whether this was the objective of the Bacchiad founders of Syracuse: I do not believe that in the mid-eighth century a government of noble landowners would have wanted to plan for regular imports of corn, to compete with the produce of their own estates. Their situation was quite different, I submit, from that of fifth/fourth-century Athens, where (as I believe) very few citizens grew corn for sale on the market, internal or external, and a large part of the urban population was fed on imported corn. I have no difficulty in seeing Syracuse as an essentially agrarian colony, sent out to relieve population pressure at Corinth, and planted at Syracuse rather than on the great corn-growing plains to the North (where Leontini and Catana were founded by the Chalcidians a few years later), for two very good reasons. First and foremost, Syracuse provides, in the form of the island (as it then was) of Ortygia, the best strategic site of the whole of the east coast of Sicily, ideal for defence against the Sicels, who were only too likely to resent a Greek intrusion involving not just a settlement of a few merchants but the absorption of much good land. As Dunbabin shows,39 Syracuse from the first extended to the mainland as well: there was good land there, stretching southwards;40 but it was Ortygia which commanded the whole coastal area—a situation, as people have often remarked, destined by nature to give control over the whole south-eastern area of Sicily.41 And secondly, there were on Ortygia exceptionally copious sources of fresh water: the famous spring of Arethusa (much admired by Cicero) and some others as well.42 The fact that Syracuse also possesses a magnificent natural harbour (two harbours, in fact) would not, I think, have appeared nearly as important in the mid-eighth century as one might assume; for at this date it looks as if merchant ships, as well as warships, would normally be drawn up on a beach. Now it may seem that I am getting very near the point of claiming that in the early period Greek cities never founded colonies for trading purposes at all, or at any rate that if they did, ‘trade’ is to be understood solely in the sense of imports, the supply 39 41

40 The Western Greeks, 17, 48–50. Ibid. 95, 211. 42 e.g. Ibid. 48, 95. See E. A. Freeman, Hist. of Sicily, I 353–6.

Trade and Colonisation

363

of some essential commodity to the mother city. I do make that assertion, but it contains one essential concealed qualification, which I shall make clear presently—because I do accept the fact that there were some ‘trading colonies’, even in my third sense of the word ‘trade’, namely commercial exchange. This seems to me quite certain. I will just refer briefly to the most obvious examples. 1. Naucratis, founded in the Nile Delta about 615–610 (the pottery evidence now enables us to date the foundation very narrowly).43 Naucratis seems to have lacked the surrounding belt of agricultural land owned by citizens which was an essential feature of every proper Greek city. It was above all what the Greeks called an ‘emporion’, a place of trade, of commercial exchange; and I suspect that more than two hundred years may have elapsed before Naucratis actually became a proper polis.44 (The word ‘emporion’ is sometimes applied loosely to a proper city which was a centre of trade; but it is often used either in relation to the port of a city, like the Peiraeus, which was just a part of the city of Athens, or to mere trading-stations, Greek or non-Greek, which were something less than proper cities.) 2. Al Mina, at the mouth of the River Orontes, in Northern Syria. A great deal has been written about Al Mina since the first publication of Woolley’s excavations over twenty years ago, notably by Woolley himself, Sidney Smith the Assyriologist, Dunbabin, and Boardman.45 The view of all these scholars is that Al Mina is the city of Posideion mentioned by Herodotus;46 but this must remain uncertain. 3. Cumae, in Campania. (It is convenient to call this colony in Italy by its Latin name, to distinguish it from the Greek city, Cyme, which provided some of its settlers and gave it its name.) Now Cumae is both the farthest Greek colony on the west coast of Italy, and at the same time the earliest (the traditional date is roughly 757). The colonists first settled at Pithecussae,47 the modern Ischia, and then transferred to the mainland. The city of 43

See R. M. Cook, ‘Amasis and the Greeks in Egypt,’ in JHS 57 (1937) 227 ff. On the ‘prostatai of the emporion’ (Hdts II 179), and other questions, see pp. 399–406 below, esp. 401–4. 45 C. L. Woolley, in JHS 58 (1938) 1–30, esp. 1–16, 28–30; S. Smith, in Antiquaries Jnl 22 (1942) 87–112; Dunbabin, The Greeks and their Eastern Neighbours (Hellenic Soc., Suppl. Paper no. 8, 1957), at pp. 25–9, 56–7; J. Boardman, op. cit. (in n. 26 above), at pp. 5–9, 24–5. 46 47 III 91. See Strabo V 4.9, p. 247 (Pithecussae); 4.4, p. 243 (Cumae). 44

364

Athenian Democratic Origins

Cumae was planted on land not very good for agriculture, whereas there was plenty of very good land available along the coasts of Italy very much nearer to Greece—land on which Greeks settled in large numbers after about 720, the traditional date of the founding of Sybaris. The evidence is purely circumstantial, but many scholars believe that we must think of Cumae as a trading rather than an agricultural colony and I think this may well be right, even if Cumae never had a real harbour. However, in view of the objections raised by R. M. Cook,48 the question should be allowed to remain open. 4. One might add more than one Greek city actually called Emporion, the name being surely sufficient proof of the original character of the foundation. The best known is the town on the east coast of Spain, a secondary foundation from Phocaean Massalia, which still preserves its ancient name in the form Ampurias. The Athenian colonies lying on the corn-route from the Black Sea are also possible examples of trading colonies, in the first of my three senses of the word ‘trade’, i.e. imports. It may be that this is what they were, but the question cannot be decided in the present state of the evidence.49 Now there is one feature I would like to stress in at any rate the first three of the probable trading colonies I have just mentioned: they are all right up against some advanced civilisation, rich enough to be able to pay well for whatever the Greeks brought, and to give something in return. The position of Naucratis speaks for itself. Al Mina is at the very place best fitted for the transmission of Greek goods across Northern Syria to the great bend of the Euphrates, and thence to the thickly populated area of Babylonia, and the reception of Mesopotamian goods travelling in the reverse direction. And Cumae was very possibly put down where it was because it was as near as the Greeks could get to the centres of Etruscan power in Italy, and because of its potentialities as a centre of trade in iron and copper. If we had not known of the existence of Greek colonies in these three places, they are the very spots in which we should have expected to find them. But the great majority of the Greek colonies, including all those in Sicily and South 48

In Historia 11 (1962) 113–14. Cf. G. Vallet, Rhe´gion et Zancle (1958) 56–7. For the view that trade with the Black Sea has nothing to do with Sigeum, at any rate, see Denys Page, Sappho and Alcaeus (1955), 158 n. 2; cf. ATL I 547–8. 49

Trade and Colonisation

365

Italy, were not near to any centres of high civilisation. The Sicels and the South Italian tribes, as the archaeologists tell us, had no great wealth and no very valuable artifacts, and the Greeks could not have found much to bring home from Sicily or South Italy except corn and slaves—and it seems unlikely to me that there would have been very much corn for export until the Greeks began to grow it themselves. We are left, then, with an apparent contradiction. There were some trading colonies, not limited (apparently) to the first of my three senses of the word ‘trade’. But, on my premises, we would expect not to find any Greek state sending out trading colonies, except perhaps in the first sense of ‘trade’. The way in which I would resolve this apparent contradiction is quite simple: I would suggest that although there were trading colonies in the general sense, they were not founded in the first place by a deliberate act of state (in the way Cyrene and Syracuse and so many others were), but that they grew up originally from below. Greek merchants trading with a particular area would first establish a tradingstation, an ‘emporion’, and would in time come to spend part of the year there; and they might soon bring with them a few of the craftsmen who made the sort of goods the natives liked (potters and vase-painters, for instance). After a time, when the emporion was well established, some of the merchants and craftsmen would wish to make their permanent home there—and at that point, or very soon afterwards, they would want to turn the settlement into a proper polis. The natural thing for them to do then was to send back to the city or cities from which the majority of them came, to ask for a distinguished oikiste¯s who would give their city a good send-off and could be buried in the agora and worshipped as a hero after his death; they would also ask for the formal transmission of religious cults, and for the other trappings of a city founded with all proper ceremony. No Greek city would be likely to refuse a request of this sort, at least if the prospects of the new city seemed reasonably good. And since every Greek city was surrounded by an agricultural territory from which it normally fed itself, there would probably be a chance of sending out some landless citizens or impoverished farmers to join in the new venture. Of course, if many new agricultural settlers arrived, and wanted to extend the territory of the colony, there might well be strong resistance by natives who had welcomed a mere trading-station not involving

366

Athenian Democratic Origins

the taking of much land: we may compare what happened at Cyrene.50 I can offer no positive evidence for this theory of mine, in the form of foundation-stories or anything else. But it seems to me very probable. It has been asserted51 of the colony of Emporion in Spain that according to the archaeological evidence ‘it seems to have existed as a small port of call for two generations before being converted into a regular colony and growing to some prosperity.’ That is exactly what I have in mind. I should like just to mention a controversy not irrelevant to our subject which has arisen among the archaeologists concerned with the Western Greek colonies. Blakeway and Dunbabin, following in the footsteps of Orsi and others, stated it as a definite fact that the Greek colonies in the West were preceded by a generation or more of Greek trade with some of the sites later colonised; they thought that Syracuse and some of the other colonies, as well as Cumae, were founded mainly for purposes of trade—and when they spoke of ‘trade’ they often seem to have meant exports, mainly (of course) of pottery, because all the evidence consists of pottery finds. This theory was challenged by R. M. Cook in 1946:52 he thought the Greek vases alleged to be of pre-colonisation date were in fact not earlier than the Greek foundations. And now, within the last few years, two leading French archaeologists, Vallet and Villard, who know the archaeological evidence well at first hand, have launched a detailed attack on the Blakeway–Dunbabin thesis,53 along the same lines as Cook. Now I am not an archaeologist, and I can claim no value for my own opinion. But for my part I think Vallet and Villard have successfully undermined the foundations of the hypothesis of pre-colonisation trade with the West, without, on the other hand, being able to prove (as they seem to think they have) the opposite hypothesis, that there was no pre-colonisation trade. The point is that Geometric and Protocorinthian vases cannot 50

Hdts IV 158–9. I cannot now remember where I read this. Cf. P. Bosch-Gimpera, in CQ 38 (1944) 53 ff., at p. 54. 52 In JHS 66 (1946) 67 ff., at pp. 70–80. 53 F. Villard and G. Vallet, ‘Ge´ome´trique grec, ge´ome´trique sice´liote, ge´ome´tri´ tude sur les premiers contacts entre Grecs et indige`nes sur la coˆte que sicule. E orientale de Sicile’, in Me´ls. d’arch. et d’hist. 68 (1956) 7–27; G. Vallet, Rhe´gion et Zancle (1958), passim. 51

Afterword

367

often yet be dated with confidence to within a decade or two; and if we have to abandon (as I think we must) the notion of a long period of trade before colonisation, we can still insist on leaving open the possibility of a few years of such trade, which could not as yet be proved or disproved from the archaeological record, on the information now available. (As a matter of fact, I am told that the most recent finds on Ischia and the neighbouring coast of Italy suggest that there was a distinct period of pre-colonisation trade.) The facts about Cumae which I mentioned earlier show that this at any rate may well have been a trading colony. On the other hand, for reasons I gave (the absence of an advanced civilisation to trade with) there may well have been no trading colonies in Sicily and South Italy—except perhaps for Zancle-Messana and Rhegium, which form a rather special case, because of their unique strategic situation, controlling the Gulf of Messina. But I still think it very likely that the early colonists of Sicily and South Italy were guided to their destination by Greek merchants who had already been trading in that area, if perhaps for a short period only. In the great majority of cases where we find a Greek city making an official colonial venture, however, there seems to me a high degree of probability that the circumstances will prove to very like those visualised by Plato:54 ‘When men who have nothing and are in want of food show themselves ready to follow their leaders against the possessions of the propertied class, they are treated as a disease of the state and are sent away in the kindest possible spirit, with the euphemistic title of a colony.’

AFTERWORD by C. M. Reed This paper, the earliest one in this volume, was delivered as a talk to the Oxford University Classical Society in 1959. To take in turn the three principal propositions Ste. Croix advances (with the vigour and finality characteristic of his talks): First, The primary cause of colonisation was land hunger due to overpopulation. Graham in his CAH2 chapter on archaic colonisation affirms the continuing consensus on Ste. Croix’s point. (Graham devoted 54

Laws V 735e–6a.

368

Athenian Democratic Origins

his career to studying colonisation; his collected papers have recently been published, and his 1964 book on the subject reappeared in 1999, sadly without its 1983 update.) Cawkwell endorses the need for food as the main cause but thinks it due to drought, not overpopulation. So currently discounted is a causal connection between archaic trade and colonisation that recent works on colonisation neglect to mention trade, just as recent or forthcoming treatments of archaic trade or traders ignore colonisation: on colonisation see for example the recent collection edited by Descoeudres; on archaic trade and traders, the 1983 overview by Cartledge, TPR, and Reed’s recent book. Ste. Croix in this talk as elsewhere pays little attention to the archaeological evidence—a hint of the tension that would persist well after 1959 between sociologically-minded British ancient historians and their counterparts in archaeology. Boardman’s ‘Epilogue’ to Greeks Overseas4 and the essays edited by Tsetskhladze and De Angelis update the relevant archaeological record. Brill has announced a forthcoming comprehensive history of Greek colonisation, edited by Malkin. Second, exceptions to Ste. Croix’s first thesis above must be made for ‘ports of trade’ such as Naucratis, Pithecussae, and Al Mina. Recent work confirms Ste. Croix’s claims about the Greek experience at Naucratis and Pithecussae; the Greek connection with Al Mina remains more controversial. Austin’s monograph on Naucratis preceded Mo¨ller’s new study, which ranges well beyond Naucratis, as well as those in the ancient Naukratis volumes edited by Coulson and Leonard. For further bibliography on Naukratis see the afterword to ch. 10. Kearsley and Boardman review the Al Mina evidence in a recent volume edited by Tsetskhladze. Third, with respect to how inter-regional trade was organised, conditions in the archaic period so closely resembled those of the classical period that we can learn much about the former from the latter. Many of Ste. Croix’s subsidiary points in this talk—for example, his emphasis on a city’s collective, official concern for imports over exports—hinge on the above premise. It is true that the archaic polis was concerned with imports, not exports, but not necessarily by means of the same mechanisms. A large body of post-Finley work deals with the political and legal mechanisms for dealing with trade developed at Athens in the classical period. Suffice it here to mention recent books by Hansen and Millett, collections edited by Cartledge and by Parkins & Smith, and the chapter by Davies in volume V of the CAH2 together with those by Austin and Rhodes in volume VI. These demonstrate how much more we know about classical trade and traders because of their appearance in polisstaged legal proceedings, polis-bestowed honors, or polis-sponsored military and diplomatic initiatives. Archaeologists can accumulate new evidence for archaic exchanges till doomsday without our necessarily

Afterword

369

comprehending its significance, above all because the requisite institutional context for doing so did not then exist. (See, however, Wilson’s article.) Finally, what would Ste. Croix have thought of the recent spate of historicizing treatments of colonisation? Ste. Croix was a historian in the Enlightenment mould, as little inclined as were Sidney and Beatrice Webb or his mentor A. H. M. Jones to reconstruct cultural contexts imaginatively. One can imagine his exuberant objections to the subjectivity of recent works by Dougherty and (to a lesser extent) by Malkin. Nor would he have been reassured by the following—one of the few occasions (p. 89) in Malkin’s Returns of Odysseus where one finds a conjunction of archaic trade and travel: ‘ . . . that Euryalos the Phaiakian [in the Odyssey 8.161–5] taunts Odysseus for seeming not a man ‘skilled in contests’ but a ‘captain of sailors who are traders’ does not, as has been thought, indicate any general derogation of trade.’ Austin, M. M., Greece and Egypt in the Archaic Age (1970) —— , ‘Society and Economy’ CAH2 VI (1994) Boardman, J., ‘Epilogue’ to The Greeks Overseas: Their Early Colonies and Trade4 (1999) —— , ‘The Excavated History of Al Mina’, in Tsetskhladze, G., ed., Ancient Greeks West and East (1999) Cartledge, P. et al., eds., Nomos: Essays in Athenian Law, Politics, and Society (1990) Cawkwell, G. L., ‘Early Colonisation’, CQ 42 (1992) 289–303 Coulson, W. D. E., Ancient Naukratis: The Survey at Naukratis and Environs, II 1 (1996) Davies, J. K., ‘Society and Economy’, CAH2 V (1992) Descoeudres, J.-P., ed., Greek Colonists and Native Populations (1990) Dougherty, C., ‘It’s Murder to Found a Colony’, in Dougherty, C., and Kurke, L., eds., Cultural Poetics in Archaic Greece: Cult, Performance, Politics (1993) —— , The Poetics of Colonization: From City to Text in Archaic Greece (1993) Graham, A. J., Collected Papers on Greek Colonization (2001) —— , ‘The Colonial Expansion of Greece’, CAH2 III 3 (1982) —— , Colony and Mother City in Ancient Greece (1964, repr. 1999) Hansen, M. H., Polis and City-State: An Ancient Concept and Its Modern Equivalent (1998) Kearsley, R. A., ‘Greeks Overseas in the 8th Century B.C.: Euboeans, Al Mina and Assyrian Imperialism’, in Tsetskhladze, G., ed., Ancient Greeks West and East (1999) Leonard, A., Jr., et al., Ancient Naukratis: Excavations at a Greek Emporium in Egypt, I 1 (1997) and 2 (2001) Malkin, I., Religion and Colonization in Ancient Greece (1987) —— , The Returns of Odysseus: Colonization and Ethnicity (1998) Millett, P., Lending and Borrowing in Ancient Athens (1991) Mo¨ller, A., Naukratis: Trade in Archaic Greece (2000)

370

Athenian Democratic Origins

Parkins, H., and Smith, C., eds., Trade, Traders and the Ancient City (1998) Reed, C. M., Maritime Traders in the Ancient Greek World (2003) Rhodes, P. J., ‘The Polis and the Alternatives’, CAH2 VI (1994) Ridgway, D., The First Western Greeks (1992) Tsetskhladze, G., and De Angelis, F., eds., The Archaeology of Greek Colonisation (1994) Wilson, J.-P., ‘The ‘‘illiterate trader’’?’, BICS 42 (1997–8) 29-56

10 But what about Aegina?

This essay, written (it seems) in 1965 or 1966, fulfils the promise made by Ste. Croix in OPW 267 n. 61: ‘I propose to demonstrate in detail elsewhere that even the governing class of Aegina was not at all the ‘‘mercantile aristocracy’’ it is so often assumed to have been, but a small, rich landowning class of archaic type,’ and again (‘as I hope to show shortly’, referring to the status of Naucratis) in a review in JHS 87 (1967) 179. A further promise appears at OPW 265: ‘I hope shortly to deal elsewhere with the civic, social and economic status of Greek merchants in the Classical period.’ Apparently Ste. Croix had intended to include a piece on this subject in his Essays on Greek History, which would no doubt have made use of material from his lecture course on the economic background of Greek politics (see Proc. Brit. Acad. 111 (2001) 460). Indeed, a note attached to the typescript of the present chapter reads: ‘This is a revised first draft, providing material to be used— probably in quite a different form—at the end of my ‘‘Trade and Politics’’ chapter.’ This explains why the opening paragraph of this essay contains phrases that imply a preceding discussion (‘what I have called’, ‘my picture’), and why the reader was instructed to ‘see above’ on various topics*—maritime loans (n. 11), the Athenian coinage decree (n. 74) and Andocides (n. 98)—and, perhaps, why the essay stops so abruptly (n. 110). ‘His picture’ puts Ste. Croix firmly on the ‘primitivist’ side of the longstanding ‘primitivist/ modernist’ debate about the nature of the ancient economy (on which see Finley, AE and BMC, Cartledge, TPR, and Scheidel & von Reden). Finley, M. I., AE ¼ The Ancient Economy (1975; 2nd edn. 1985, reissued by Penguin Books 1991; updated edn. ed. I. Morris 1999) —— , BMC ¼ (ed.), The Buecher-Meyer Controversy (1979) Scheidel, W., and von Reden, S., The Ancient Economy (2002)

* These ‘non-references’ have been adjusted in the present version.

372

Athenian Democratic Origins

One of the leading Manichees of St. Augustine’s day felt that he was scoring a useful point off the Catholics when he asked, in effect, how they could accommodate the scorpion in their scheme of things.1 I have often been reminded of this when those who belong to what I have called the ‘modernising’ school say to me, ‘Ah, but what about Aegina?’. Augustine found no difficulty with the scorpion, and Aegina will fit very nicely into my picture. The almost universal tendency to assume that Aegina, an island of only about 85 square kilometres, ‘must have been’ an ‘industrial and trading city’,2 governed by some kind of mercantile ruling class, a Handelsaristokratie [‘trade aristocracy’], even a Kaufmannsaristokratie [mercantile aristocracy],3 rests upon no evidence whatever: there is not in any source, as far as I know, any basis for believing that the governing class of Aegina at the time of her 1 See Augustine, De Mor. Manich. (VIII) 11 (PL XXXII 1349–50). The Manichee maintained that evil was a substantia, and the essence of his proof, as represented (or misrepresented) by Augustine, was that the scorpion was both evil and a substance. 2 Among many statements by modern scholars which might be quoted, I shall single out a few characteristic examples (with my italics): (a) Meyer, GdA III2 503 (Aegina the most important ‘Handelsstadt’ of the whole Saronic Gulf area in the late Archaic period); Kleine Schriften I2 (1924), 113–14 (Aegina in the sixth century ‘vielleicht das bedeutendste Handelsemporium der griechischen Welt . . . Seine Schiffe befahren alle Meere, seine Kaufleute gewinnen fabelhafte Reichtum, neben der Handelsflotte steht eine starke Kriegsflotte; eine rege Industrie, die grossen Sklavenmassen bescha¨ftigt, entwickelt sich auf der Insel’); (b) Busolt, GS I 164, 168 (Aegina described as ‘die bedeutende Industrie- und Handelsstadt’); (c) Beloch, GG I2 i 279; II2 i 25, 77; BGRW 96; (d) Holm HG I 428. [The author intended to add further examples.] 3 One of the most confident proponents of this view was Eduard Meyer: see GdA IV3 i 399 (Aegina mentioned among the many ‘Handels- und Industrie-staaten’ in which ‘eine neue kaufma¨nnische Aristokratie’ arose, which ‘auf dem Capital und dem Besitz von Schiffen, Fabriken, Sklaven beruht und den alten Adel in sich aufnimmt’); ¨ gina fand die Tyrannis keinen Boden; hier bestand von Anfang III2 583 (‘Nur auf A an die rein merkantile Politik, zu der Korinth nach dem Sturz der Kypseliden gelangte’), 707 (‘In Korinth fu¨hren die Kaufleute ein weises und stabiles Regiment . . . ¨ gina gewesen sein’); Kl. Schr. I2 (1924) Nicht viel anders werden die Zusta¨nde in A 194 (‘Im manchen griechischen Staaten hat die Kaufmannsaristokratie die ¨ gina, in Rhodos, ebenso in Karthago’) Herrschaft gewonnen, so in Ionien, in A (my italics throughout). See also (again with my italics) Adcock, in CAH IV 26 (Aegina ‘the jealous island of merchants’); Walker, ibid. 263–4 (‘In Aegina there was . . . an aristocracy of merchant princes’); Busolt, GS I 191 n. 1 (Vorwiegend oder ausschliesslich Grossha¨ndler und Fabrikanten waren die Kapitalisten in Aigina, Korinthos, Milet und andern Sta¨dten’); Kahrstedt, in Go¨ttingische Gelehrte Anzeigen 1926, at p. 104 (‘Ferner etwa wa¨re Aigina nachzutragen als Beleg dafu¨r, dass der Adel anfa¨ngt Handel zu treiben, man denke an die hochfeudalen Stammba¨ume der aiginetischen Handelsherren bei Pindar’).

But what about Aegina?

373

greatness, in the sixth century and the first half of the fifth, was any different from the other landowning oligarchies of that time. That it was in any sense ‘mercantile’ in composition is a pure supposition a priori. Yet Glotz and Cohen, for example, can say confidently, ‘Son aristocratie ne fut jamais une classe de proprie´taires fonciers; elle se composa toujours d’industriels, de ne´gociants et d’armateurs’ [‘her aristocracy was never a class of landed proprietors; it was always made up of manufacturers, merchants and ship-owners’]; Aegina had ‘une politique obstine´ment mercantile’ [a persistently mercantile policy].4 Other scholars have spoken in much the same vein. Hasebroek of course protested against this picture, but his treatment of Aegina (TPAG 51–3) was sketchy and inaccurate, and the ‘modernisers’ could afford to take little notice of him. And then, in 1938, there appeared two monographs devoted entirely to Aegina: Gabriel Welter, Aigina, a publication of the German Archaeological Institute, and Hans Winterscheidt, Aigina. Eine Untersuchung u¨ber seine Gesellschaft und Wirtschaft, a Cologne dissertation by a pupil of Hasebroek. These two publications were given a long review by E. Kirsten, in Gnomon 18 (1942) 289–311 [henceforward Kirsten, ‘Review’], with several pages5 devoted to an examination of the social and economic problems. Winterscheidt, working in Hasebroek’s own university, evidently felt he could take for granted both Hasebroek’s conception of early Greek society in general and his ‘neues Bild’ [new image] of Aegina in particular, without detailed refutation of rival views. He was mainly content, therefore, to point out that the usual theory of an Aeginetan ‘commercial aristocracy’ is entirely devoid of foundation, and to replace this with a collection of the evidence for a governing class of fairly typical archaic Greek pattern. Denying altogether, with Hasebroek, the existence of large-scale Greek trade (Grosshandel), he believed that the only citizens of Aegina to take part in mercantile activity came from the lower classes. Doubtless he also believed that a great part of the trade of the island was in the hands of men who were not citizens of Aegina at all, but—perhaps by an oversight—he does not seem to state this explicitly. 4

HG I 315; cf. 480 (‘Les oligarchies commerc¸ants de Me´gare, de Chalcis et ´ gine’). d’E 5 Esp. 297–301, cf. 294–5, 302, 303, 304.

374

Athenian Democratic Origins

Kirsten, rejecting Hasebroek’s general viewpoint, was very critical of Winterscheidt’s ‘neues Bild’ of Aegina; but (as will appear presently) the course of reasoning which led him to reject it was essentially a priori, and we shall see that even those of his arguments which are not entirely misconceived contain fallacious presuppositions of a type which is familiar in the works of the ‘modernising’ school. Asking himself the question who carried on the trade of Aegina, nobles or subject demos, or metics, Kirsten (‘Review’, 297 ff.) notes that according to Winterscheidt it was members of the demos (conceived as humble Dorians, not as Ho¨rige [‘subjects’]), and he goes on to say that Welter follows the predominant view, that trade was in the hands of nobles and metics. He himself prefers what he conceives as a solution intermediate between the two. However, this characterisation of Welter’s view6 seems to me not to give a fair impression of what Welter actually says. I can find not a word in Welter’s little book, or in his series of ‘Aeginetica’ in Archa¨ologischer Anzeiger,7 about direct participation in trade on the part of the Aeginetan ruling class. In fact Welter8 sees ‘die herrschende dorische Oberschicht’ [‘the ruling Dorian upper class’] as ‘konservativ, landgebunden’ [‘conservative landowners, strongly attached to their estates’]— rightly, in my opinion. With them he contrasts the lower-class citizens, among whom were to be found not only the sculptors but also ‘die Schiffsbesitzer, die als Kapita¨ne auf eigenem Schiff Handel trieben’ [‘the ship-owners, who practised trade as captains of their own ship’]. He then goes on at once to say quite explicitly, ‘Der Handel lag zum gro¨ssten Teils im Ha¨nden der Meto¨ken’ [‘trade lay for the most part in the hands of the metics’]; and later he speaks of ‘die vordorische und Meto¨ken-Bevo¨lkerung, die bewegliche Tra¨gerin des Handels’ [‘the pre-Dorian and metic population, the active representatives of trade’]. Welter sees the ruling class as concerned with trade in only one way: ‘das Prinzip der a¨ginetischen Handelspolitik’ [the principle of Aiginetan trade 6

See esp. his Review, 297 (‘W[elter] folgt der herrschenden Meinung’) and 300 n. 1, where Kirsten represents Welter as seeing ‘neben dem Grosshandel der aiginetischen Herren noch ein Seedarlehensgescha¨ft’ with metics. I can find nothing in Welter corresponding to the words I have italicised. But see n. 9 below. 7 AA (1938) 1–33 (‘Aeginetica’ I–XII), 480–540 (Id. XIII–XXIV), and (1954) 28–48 (Id. XXV–XXXVI): see esp. (1954) 29–30 (§ XXV). 8 Aig. 96, 99, with 31; cf. AA (1954) 29–30 (§ XXV).

But what about Aegina?

375

politics] can be defined, according to him, as ‘Sicherung des Meto¨ken zur Hebung der Darlehensgescha¨fte’ [‘ensuring the metic’s security in order to encourage loan businesses’]. It was the ‘umfangreichen Schiffsdarlehngescha¨fte des reichen a¨ginetischen Adels’ [‘extensive maritime loan businesses of the wealthy Aiginetan aristocracy’] which brought large numbers of foreign merchants to settle in Aegina as metics. The state profited from the poll tax on the metics and the taxes on imports: these were its principal sources of revenue. On the other hand, Welter (Aig. 30) speaks of a large and growing ‘merchant fleet’ (Handelsflotte) as a possession of ‘die oligarchische Adelsregierung’ [‘the oligarchic regime of the aristocrats’]: it is not clear to me whether he conceives the ships as in public or private ownership, or how he thinks they were manned; but at any rate he clearly does not suppose that the great men had any direct participation in trade.9 In addition to the Kleinhandel [retail trade] revealed by the archaeological evidence, he insists— without of course being able to produce any reasons—upon the existence of a Grosshandel [‘wholesale trade’] (which has admittedly left no trace in the archaeological record) in raw materials (cereals, ore and wine), slaves and Schiffsfracht [‘freight’],10 especially with Asia Minor, Egypt and the Black Sea. Here, he attributes an important role to a factor I have already mentioned: ‘extensive mercantile loans by the wealthy Aeginetan aristocracy’. Kirsten objects to this that Seedarlehensgescha¨ft [‘mercantile loans, bottomry’] is a phenomenon which does not appear until later (‘Review’, 300 n. 1); and here I think he is very probably right, although this cannot actually be proved. As I have said elsewhere, all we know is that bottomry loans were a well-established institution by the end of the fifth century.11 I should be surprised 9

Cf. Aig. 96–9. Perhaps it was his statement on p. 30 (see above) that misled Kirsten? 10 Aig. 30–1. His statement is repeated with approval by W. Kraiker, Aigina. Die Vasen des 10. bis 7. Jahrh. v. Chr. (Deutsches Archa¨ol. Inst., Berlin, 1951) 12. 11 [See Ste. Croix’s ‘Ancient Greek and Roman Maritime Loans’ in Debits, Credits, Finance and Profits (Esssays in honour of W. T. Baxter), ed. H. Edey and B. S. Yamey (1974), 41–59. Claims for earlier examples remain dubious (Cartledge, TPR 182 n.19): the practice cannot be traced back with any confidence earlier than 421 BCE (see Harvey in ZPE 23 (1976) 233). The two recent discussions by P. Millett, ‘Maritime Loans and the Structure of Credit in Fourth-Century Athens’, in P.Garnsey et al. (eds.), Trade in the Ancient Economy (1983), 36–52,

376

Athenian Democratic Origins

myself if they appeared much before the mid-fifth century, but of course it is quite possible that they were invented in the sixth, and I suppose that even a seventh-century origin is just conceivable. Welter’s picture, then, can neither be proved correct nor convincingly disproved. To my mind, Aeginetan Grosshandel [‘wholesale trade’] and the giving of bottomry loans by the nobility of the archaic age are to be rejected; but I would certainly accept the remaining elements in Welter’s picture: the personal participation in foreign trade of poor citizens, and an even greater mercantile activity on the part of metics—and, of course I would add (for the reasons I have explained [in OPW 264–7]), other foreigners. Welter, Winterscheidt and Kirsten between them cite nearly all the relevant evidence and bring forward most of the possible arguments, but with very varying effectiveness, and on several matters they are very far from having said the last word. I think my best method of approach is a piecemeal one: without trying at this stage* to develop any continuous argument, I shall attack our problem from various different directions, one after another, analysing significant pieces of evidence in so far as they exist, but sometimes merely exposing ‘modernist’ presuppositions or criticising arguments that seem to me invalid.

(I) I begin with the one glimpse we have of internal politics at Aegina: the unsuccessful revolution led by Nicodromus (Hdts VI 88–91), just before or after the battle of Marathon,12 when Herodotus tells ’ us (VI 91.1) that Ai’ ginZt
evn oi‘ pax
eeB epanast
antoB swi to &y ’ d
Zmoy a
‘ ma Nikodr
omfiv epekr
atZsan [‘the wealthy Aeginetans prevailed over the demos, when they rose up against them along with Nicodromus’], massacring seven hundred after others had and at pp.188–96 in his Lending and Borrowing in Ancient Athens (1991) (with notes at 305–8), a ‘selective restatement’ (305 n.17) of parts of his earlier article, are, as he says (305 n.17), ‘dependent on the fundamental study by de Ste. Croix’. For further treatment of the subject, see now Charles Reed, Maritime Traders in the Ancient Greek World (2003).] * [This phrase strongly suggests that Ste. Croix originally intended to provide further discussion.] 12 Contrast the chronology of A. Andrewes, in BSA 37 (1936–37) 1–7, with that of N. G. L. Hammond, in Historia 4 (1955) 406–11, approved by L. H. Jeffery in AJP 83 (1962) 44–54. [Further bibliography in the Afterword.]

But what about Aegina?

377

escaped by ship with their leader. Apart from the fact that Nicodromus had conspired with the Athenians to betray Aegina to them, the purpose of the revolution is not explicitly stated by Herodotus, but the revolutionaries must certainly have intended, with Athenian help, to broaden the class of those entitled to exercise political rights, if not to set up a complete democracy on something like the Athenian model. Nicodromus himself was a man of some distinction [VI 88]. No clue is given to the social composition of his followers: doubtless they included well-to-do citizens outside the governing oligarchy as well as many humble men. I find it interesting that the term Herodotus uses for the Aeginetan oligarchy, oi‘ pax
eeB, is a word he applies elsewhere only to the Hippobotai of Chalcis (V 77.2) and the wealthy class of Naxos (V 30.1) and Sicilian Megara (VII 156.2), to whom no one, I think, will wish to attribute a mercantile complexion. Seven hundred is a very large number of victims, out of a citizen population of only a few thousand (see Section (v) below), and it may well be an exaggeration; but it is easy to believe that every potential democratic leader who did not escape with Nicodromus was killed. What probably remains in the minds of most readers of Herodotus is the vivid picture of the unknown man whose severed hands were left grasping the door handles of the sanctuary of Demeter Thesmophoros, at which he had vainly tried to take refuge (VI 91.2)—giving an impression of the ferocity of the rulers of Aegina13 which was no doubt calculated by Herodotus, whether or not he was the dupe of Athenian propaganda. When the island capitulated to Athens c. 457 (Thuc. I 108.4; Diod. XI 78.4) the strength of the Aeginetan aristocracy was broken, probably for ever, and there may well have been some democratisation of the constitution. In 431 the Athenians expelled the Aeginetans altogether and installed colonists of their own; and we are told by Thucydides (II 27.1–2) that while some of the Aeginetans were dispersed throughout Greece, some were settled by the Spartans at Thyrea. The latter were slaughtered by the Athenians when they captured Thyrea in 424 (Thuc. IV 56.2– 57), and although some of the others were brought back to the island by Lysander in 404 (Xen., Hell. II ii 9), when the oligarchy will certainly have been restored, the old aristocracy can never 13

Cf. also IX 78–9, with n. 29 below.

378

Athenian Democratic Origins

have fully regained its old position, and Aegina was never again a power of any real importance.

(II) I come now to some evidence about the Aeginetan ruling class which is unequivocal. Pindar, the great poet of the first half of the fifth century whom we have good reason to connect with ‘the international aristocracy of Greece’ (as Wade-Gery has aptly called it, EGH 246), wrote more than twice as many of his surviving epinician odes for Aeginetan victors in the four great Panhellenic festivals as for men from any other state: not to mention a fragment composed for an unknown victor, there are no less than eleven in all, written for ten different Aeginetans,14 three of them trained by the great Melesias of Athens,15 and most of them belonging to clans16 the names of which are rehearsed in such a way as to show that they are blue-blooded. The great names ring out: the Bassidai, pala
iwatoB gene
a [clan of ancient fame],17 with no less than twentyfive triumphs at the Panhellenic Games, and more crowns won in the boxing ring than any other house in Greece; the Blepsiadai, Chariadai, Euxenidai, Midylidai, Psalychidai, Theandridai.18 14 Ol. VIII (Alcimedon); Pyth. VIII (Aristomenes); Nem. III (Aristocleides); IV (Timasarchus); V (Pytheas); VI (Alcidamas); VII (Sogenes); VIII (Deinis or Deinias); Isth. V and VI (Phylacidas); VIII (Cleandros); IX Snell (victor unknown) ¼ fr. 1 Bowra. [See n. 28 for recent work on Pindar and xenia.] 15 See Ol. VIII 54 ff.; Nem. IV 93 ff.; VI 66 ff. (For Pindar’s Melesias as the father of the Athenian politician Thucydides, see Wade-Gery, EGH 244–6; [T.J. Figueira has more recently devoted a lengthy chapter (8) in his Excursions in Epichoric History (1993) to this shadowy but important figure.]) Another Aeginetan victor, Pytheas, was trained by Menander of Athens: Pind. Nem. V 48–9; Bacchyl. XIII 191–8. 16 They are p
atrai, and p
atra is the Doric word corresponding to g
enoB: see Busolt, GS I 133 n. 6. 17 Nem. VI 31; cf. 32–46, 8–27. H. Knorringa, Emporos (1926) 16–17, absurdly ’ treats the metaphor that follows, in line 32 (’
idia naystol
eonteB epik
vmia [conveying their own praises]), as ‘an allusion to the trading practices of this Aeginian house’. (Contrast Knorringa’s interpretation of the metaphor in Nem. III 19–23.) And Hasebroek (TPAG 21), the German original of whose book was published two years after Knorringa’s, also takes the metaphor literally: he says, ‘The sea voyages which the noble families of Aegina undertook, and from which, according to Pindar, they acquired great renown, were plundering expeditions, and had nothing to do with commerce at all.’ 18 Ol. VIII 74–84, cf. 15–18 (Blepsiadai); Nem. VIII 44–8, cf. 16 (Chariadai); Nem. VII 70, cf. 90–4 (Euxenidai); Pyth. VIII 35–42, cf. 19, 71–2, 78–80 (Mid-

But what about Aegina?

379

Two of the epinician odes of Bacchylides were also written for Aeginetans,19 one of them for the Nemean pancratiast Pytheas son of Lampon, whose victory Pindar also celebrated.20 Several other famous Aeginetan athletes are known to us from other sources:21 among them is Crius, undoubtedly that Crius22 son of Polycritus who bandied words with Cleomenes of Sparta and whose son Polycritus later distinguished himself as a trierarch at Salamis.23 All these men will have belonged, like the family of Crius, to the governing oligarchy.24 Pindar, ‘the most articulate voice of that aristocratic order for which Thucydides [the son of Melesias] stood’ (Wade-Gery, EGH 251), is far louder and more heartfelt than convention demanded in his praise of Aegina. Just as Bacchylides declared that Aegina was guided by Areta and Eukleia and Eunomia (XIII 175–89), so for Pindar she was a e y
’ nomoB p
oliB [law-loving city].25 (It is an iron’ ical commentary on the meaning of eynom
ia in the mouth of a Greek aristocrat that Pindar’s poem,26 and probably that of Bacchylides,27 should have been written within a few years of the ylidai); Isth. VI 57–73, cf. 3–7, 16, with V 17–22, 53–6, and Nem. V 4–5, 41–6 (Psalychidai); Nem. IV 73–96, cf. 13–22 (Theandridai). See Welter, Aig. 130–1. 19

XII and XIII, ed. B. Snell. Nem. V. Two of Pindar’s Isthmians, V and VI, were written for Pytheas’ younger brother, Phylacidas. An Aeginetan named Pytheas was a Naopoios at Delphi in 346 (SIG3 241 B. 79), but of course he need not have belonged to the same family. 21 e.g. Pherias, Praxidamas and Theognetus, of whom there were statues at Olympia (Paus. VI xiv 1; xviii 7; ix 1), also Taurosthenes (Paus. VI ix 3; Ael., VH IX 2). A list is given by Winterscheidt Aig. 48 n. 111. Part of a bronze plaque from the base of Pherias’s dedication has been discovered: SEG XI 1231; and see Jeffery LSAG 112–13 (no. 21). 22 Simonides, fr. 507 (Page). 23 Hdts VI 50.2–3 and 73.2; VIII 92.1 and 93.1[ on this incident see also pp. 427–8, 435 below]. 24 If, as is quite possible, Pindar, Nem. III 67–70 (with Schol.), is speaking of a ue
arion which was the official meeting place or residence of state magistrates called uearo
i (uevro
i), who are known elsewhere (see Busolt, GS I 508), then Aristocleides was presumably a member of that board; but I hardly think this is certain (cf. Winterscheidt, Aig. 26 n. 37). [See Figueira ASP 314–21.] Unfortunately, ‘wir wissen u¨ber die Verfassung von Aigina erst recht nichts’ (Wilamowitz, Pindaros (1922), 277). 25 Isth. V 22. Another city Pindar praises as the home of Eunomia is oligarchic Corinth, where Dika and Eirena also dwell: Ol. XIII 6–7. 26 Isth. V has been dated between late 480 and 476. It must have been written (or at any rate finished) after the battle of Salamis, because of lines 48–50. 27 Various dates, between 489 and 481, have been proposed for Bacchyl. XIII, which is coeval with Pind., Nem. V. Pindar’s complimentary reference to Athens 20

380

Athenian Democratic Origins

slaughter that ended the revolution headed by Nicodromus. Those severed hands, whether or not they troubled the poets, may awaken a certain feeling of uneasiness in the mind of the modern reader who is invited to contemplate the happy internal condition of aristocratic Aegina.) Aegina in Pindar is a p
oliB ueowil
ZB [‘city loved by the gods’] (Isth. VI 65–6), near to the Graces, a‘ dikai
opoliB [‘the city of justice’] (Pyth. VIII 22), with unsullied glory from the first (ibid. 24–5), famous for her men (ibid. 25–32), the sons of Aeacus with their golden chariots (Isth. VI 19), renowned for her ships, naysiklyt oB A’
igina (Isth. IX 1–2 Snell ¼ fr.1 Bowra; Nem. V 9)—and these, of course, are not mere merchant ships: Aegina is dolix
ZretmoB (Ol. VIII 20), long-oared, and Pindar is thinking of the exploits of her warships, above all at Salamis, which he recalls gratefully in the Fifth Isthmian (lines 48–50). Above all, perhaps, one notices in the two poets allusions to Aeginetan hospitality (jen
ia), the friendliness and justice shown by the Aeginetans to j
enoi, a subject dear to the hearts of Pindar and Bacchylides. It is an absurd error to treat this as essentially ’ aeuloB gene
a mere friendliness to traders.28 Lampon, father of a ey
[‘race of fine athletes’], for whose sons Phylacidas and Pytheas Pindar wrote three epinicians,29 is loved, says Pindar (Isth. VI ’ 70), for his j
envn eyerges
iai [‘kindnesses to strangers’]; and (line 49) suggests that a state of war did not exist at that time between Athens and Aegina (cf. n. 12 above). 28 The truth was seen by Winterscheidt, Aig. 29–31, and his conclusion was accepted even by Kirsten, ‘Review’, 298 (but see n. 30 below). Among passages in Bacchylides and Pindar referring to jen
ia and j
enoi which are not mentioned below, see Bacchyl. I 145–50; V 6–14, 49; Pind., Ol. I 103–5; II 5–6; XIII 1–3; Pyth. III 68–71; X 64–6; Nem. I 19–24, with IX 1–3; VII 61; Isth. II 23–4, 47–8; and esp. Ol. ’ ik’ aggel
’ ’ IV 4–5 (je
invn d’ ey
prass
ontvn e
’ sanan ayt
ian poti glyke&ian eslo
i [‘when friends fare well, good men are straightway delighted at the sweet news’]). In Nem. VII 64–5, Pindar records with pride his own projen
ia of ‘the Achaeans’, i.e. apparently the Molossians. [On Pindar and jen
ia see now S. Instone (ed.), Pindar: Selected Odes (1996), 3–5, I.L. Pfeijffer, Three Aiginetan Odes of Pindar (1999), 62–3, 101–3, 111–13 and S. Hornblower, ‘Pindar, Herodotus and Aigina’, a paper delivered at the Institute of Classical Studies, London, on 7 Nov. 2002 (publication forthcoming). Note however the imagery from the traditional aristocratic pursuits of agriculture (lines 9–11) and hunting (14) in Nemean VI.] 29 Nem V; Isth. V and VI. This Lampon is very probably the L
ampvn ’ vni t o‘ Pyu
ev, Ai’ ginZt
evn he a pr vta & [Lampon son of Pytheas, one of the leading men of the Aeginetans], mentioned in Hdts IX 78–9, even if the story there told of him is a malicious Athenian slander. (Cf. How and Wells, CH II 321).

But what about Aegina?

381

Bacchylides, who praises the very same Lampon for his jen
ia wil
aglaoB [‘splendid hospitality’, XIII 224–5], apostrophises the nymph Aegina as d
espoina pagje[
inoy xuon
oB [‘mistress of an all-hospitable land’, XIII 95], and in a fragment from another epinician, addressed to yet another Aeginetan champion wrestler, says that p
otnia N
ika [‘lady Victory’] has sent him to j
enoi in the happy island (XII 4–7). For Pindar, Aegina is not simply t an polyj
enan . . . n &ason A’
iginan [‘the welcoming island of Aegina’, Nem. III 2–3], a divine pillar set up by some ordinance of the Immortals for pantodapo&isin . . . j
enoiB [‘foreigners of all kinds’] ’
(may it be for ever so, Ol. VIII 25–9): she is w
ilan j
envn aroyran [‘a land which is kind to strangers, or a land that strangers love’, Nem. V 8], and more, she is the island where Themis, daughter of Zeus Xenios, is honoured above all (Ol. VIII 21–3), a universal beacon d
ikfia jenark
ei [‘in its justice in protecting foreigners’, Nem. IV 12–13]; which has received not only a naypr
y tanin da
imona ’ [‘ship-ruling fortune’] but t an uem
ijenon aret
an [‘the virtue of ’ e honouring foreigners’, Paean VI 130–1], her men oy’ u
emin oyd ‘ d
ikan je
invn yperba
inonteB [‘transgressing neither divine right nor the justice due to foreigners’, Isth. IX 6]. We know very well what Pindar and Bacchylides mean when they speak like this: Aegina is a city where aristocratic hospitality, ‘guest friendship’, flourishes in abundance, and the j
enoi she receives so nobly are above all the guest-friends of her own aristocracy, coming to her as ’ i jen
ian [‘for hospiCastor and Pollux once came to Pamphae¨s, ep tality’, Nem. X 49], or as the Theandrid Timasarchus, one of ’ vn Melesias’ pupils, came to Pindar’s city, w
iloisi g ar w
iloB elu ’
sty kat
edraken [‘coming as a friend to friends, he beheld j
enion a the welcoming city’, Nem. IV 22–3]. No doubt such a community is hospitable, by Greek standards, to all strangers: some of the jen
ia which a Lampon delights to show men like Bacchylides is very likely to be extended, in a rather different form, even to ’
visiting traders.30 The landed aristocrat, exvn kr
essona plo
y toy m
erimnan [‘who has a concern which goes beyond wealth’] (as Pindar says approvingly—or is it warningly?—of Aristomenes the Midylid, Pyth. VIII 91–2), much as he may despise the foreign trader who brings him what he wants, is likely to cultivate and 30 See Winterscheidt, Aig. 58. I do not see why Kirsten, in his Review, should think this an admission ‘seiner These gegenu¨ber inkonsequent’.

382

Athenian Democratic Origins

protect him. A ruling merchant class, on the other hand, nearly always tries to exclude, or at least restrict and hamper, its foreign competitors. Pindar gives us only a fleeting glimpse, I think, of the mercantile activity which certainly did go on at Aegina: this is when he wishes that his sweet song will spread from the island ’ i p
‘ adoB en ’
t’ ak
’ atfiv [‘on every ship and in every boat’, ep asaB olk
Nem. V 2]. There is nothing here of an ‘Aeginetan merchant fleet’: Pindar is thinking of all the merchants who trade from and with Aegina. It is most interesting and significant that in Isocrates’ Aegineticus (XIX)—the only forensic speech we possess from the Classical Greek period delivered to a non-Athenian audience—there are constant echoes of the emphasis on jen
ia and wil
ia which we find in Pindar. We are in the same world of rich gentlemen (see esp. §§ 7, 13, 36), who even if they are exiled from their homes (§§ 11, 12, 20–7, 31, 38–9) have their j
enoi (§§ 10, 18, 22, cf. 5) and w
iloi31 who will help them. Thus it is, I am sure, that the unwilling Greek concubine of Pharandates, whose father was Hegetoridas the Coan, a man distinguished enough to be the guestfriend of Pausanias the Regent of Sparta, wanted to go to Aegina when she was set free (Hdts IX 76), doubtless because she knew that in Aegina of all places she could rely on help from the j
enoi of her father’s family. It is a proud boast of the speaker of the Aegineticus, a man of the highest society in Siphnos, that he is inferior to none in his friends (peri toyB w
iloyB), and he claims that he had with the deceased (whose property he is claiming, as an adopted son) a wil
ian . . . palai an kai patrik
Zn [‘ancient and ancestral friendship’] (Isocr. XIX 50, cf. 10, 13 etc.). The whole atmosphere is as different as it could be from that of the usual Athenian forensic oration: the speaker is a gentleman who knows he is speaking to gentlemen and delights to stress gentlemanly virtues. The speech is generally agreed to date from the late 390s,32 when the island was still within the Spartan sphere of influence. Ever since the restoration by Lysander of such of the exiles as survived, Aegina will certainly have been under an oligarchic government, doubtless patterned to some extent upon the 31 There are constant references to wil
ia and w
iloi in the speech: see §§ 10, 13, 17, 29, 32, 34, 38, 48, 50. 32 ´ . Bre´mond, in the Bude´ edition of Isocrates [1928], I See G. Mathieu and E 91–2 (‘probably 391 or 390’).

But what about Aegina?

383

old aristocracy, even if most of the old noble families had virtually disappeared in the half-century of Athenian dominance or in the massacre after the capture of Thyrea.33

(III) There is scarcely any trace of Aeginetan industry or manufacture,34 and certainly not a suggestion anywhere that Aegina ever had a large export industry, even of the bronzes which were certainly made on the island. In the period of Aegina’s greatness, in the sixth and early fifth centuries, it seems that no painted pottery was manufactured there. One or two archaeologists, [for example] Weinberg in 1941,35 have made tentative suggestions that Aegina may have been the place of manufacture of certain early decorated ceramic wares which have usually been taken to be Corinthian in origin; but no convincing argument has ever been produced in favour of this, and since Kraiker’s positive and convincing dismissal of such speculations,36 I think they can safely be disregarded. Earlier, in the late eighth and seventh centuries, it is possible that some of the vases we know as Protoattic were made, or at any rate painted, on Aegina,37 the one place apart from Attica 33

See p. 377 above. The only references to industry in Aegina which are worth mentioning are the statements in Pliny (NH XXXIV 8, 10–11, 75) about Aeginetan bronze-work. I know of no authority for saying there was a perfume industry in Aegina: Theophr., De Odor. (VI) 27 (p. 368, ed. F. Wimmer) simply says the best kr
okinon grew ‘in Aegina and Cilicia’; cf. Athen. XV 689D. 35 S. S. Weinberg, in AJA 45 (1941) 30–44, at p. 43. M. Robertson, in BSA 43 (1948) 1 ff., at p. 53, said he was not convinced. 36 Kraiker, op. cit. (in n. 10 above) 11–13, who points out that Weinberg’s two groups of Protocorinthian Geometric are not so distinct, when we take account of the difference between them both and other wares, as to warrant our assigning to them different places of manufacture. He emphasises the significance of the contrast between Aegina, which imported vases from a large number of different states, and real centres of manufacture such as Corinth and Athens, which produced their own wares and imported little. 37 [Ste. Croix alludes to pottery in the so-called Black and White Style. The claims of Aegina have been advanced most fully by S. Morris, The Black and White Style: Athens and Aegina in the Orientalising Period (1984); Boardman, who is apparently not convinced, provides a brief account in his Early Greek Vase Painting (1998) at 89–90, with bibliography at 278.] The evidence consists of two inscriptions painted on sherds found in Aegina: one (Jeffery, LSAG 110, and 112 no. 1) of about the last decade of the eighth century, in a script which might be Aeginetan or Attic, and the other (ibid. no. 2) of the mid-seventh century, in a script recognisably Aeginetan. Miss Jeffery makes the interesting comment, ‘That an Aiginetan should 34

384

Athenian Democratic Origins

where Protoattic has been discovered in some quantity; but of course any such small-scale industry will hardly have continued after about 625. Not that I would place any particular emphasis on the failure of Aegina to develop the manufacture of painted pottery—here we must again remind ourselves of the salutary warning of R. M. Cook (BBKGH, esp. 114, 122–3) not to exaggerate the importance of this industry in the economy of Greek states or judge the extent of their manufactures by it. In this case it is the absence of evidence for a large export industry of any kind that is significant: there is no reason to suppose that even Aeginetan bronzes were manufactured in large quantities, although their artistic quality may have been very high. I of course would rate the volume of industrial production by Greek states in general much lower than some historians: here again I would refer to Cook’s conclusions concerning Athenian fine ceramics (ibid., also 115–16, 118–21). For Aegina we can at least conclude that whatever the sources of wealth of the ruling class, only an insignificant proportion can have come from industry.

(IV) The literary sources preserve the names of just two ‘Aeginetan merchants’, each of them famous in his own way: (a) The first, Sostratus son of Laodamas, was evidently a citizen of Aegina. The one ancient reference to him, Hdts IV 152.3, is often handled in a very unsatisfactory manner. Kirsten, for example, says, ‘Der Reichtum des Sostratos . . . , auf Spanienfahrten erworben, ist gewiss nicht eines Kleinha¨ndlers, sondern des Vermittlers von Erz’ [‘The wealth of Sostratos . . . acquired on journeys to Spain, is certainly not that of a small trader, but of a dealer, in ore’].38 It is of course perfectly possible that Sostratus traded in Spanish silver; but the laconic statement of Herodotus gives no have been employed on Attic pottery in the midst of the e
’ xurZ palai
Z [ancient enmity] is not impossible, for the unbroken series of Attic ware from the eighth century onwards found at various sites on the island shows that, whatever were their feelings of hostility towards each other, Aigina did not cease to have commercial intercourse of some kind with Athens’; and she adds a reference to T. J. Dunbabin’s article, ‘e
’ xurZ palai
Z’, in BSA 37 (1936–7) 83 ff. esp. 84. 38

Kirsten, ‘Review’, 299. The same ideas are to be found in Welter, Aig. 29, 102 (with a date in the second half of the seventh century, which is far from certain [see Afterword]).

But what about Aegina?

385

such details, nor does it provide any support for conceiving Sostratus as a merchant on a grand scale, a member of the imaginary Kaufmannsaristokratie [‘mercantile aristocracy’] of Aegina. Describing how Colaeus of Samos and his crew made an enormous profit, of some 600 talents, out of a single voyage to Tartessus (about 630 B.C.), Herodotus says that it was the greatest profit ’ wort
ivn [‘from merchandise’] made by any Greek ‘after ek Sostratus the son of Laodamas, for with him no one can be compared’. (He gives no indication of Sostratus’ date.) The most natural assumption is surely that Sostratus, like Colaeus, made his vast profit from a single very successful voyage. Whether Sostratus already belonged to the ruling class or whether his great wealth enabled him or his descendants to enter it, and whether he went on trading afterwards or (as I would certainly expect) settled down to live the life of a gentleman, we simply do not know. Such glimpses as we have of the Aeginetan ruling class, from the first half of the fifth century, show us an aristocracy of typical Greek archaic pattern, with not a hint of a mercantile complexion anywhere (see Section (ii) above). Eduard Meyer (GdA III2 496) was entirely unjustified, of course, in stating, solely on the strength of Herodotus’ mention of Sostratus, that since about the end of the seventh century ‘the Aeginetans’ were among the richest ‘Kaufleute’ [merchant peoples] of all Hellas. (b) The only other ‘Aeginetan trader’ we can name is Lampis, famous in the mid-fourth century as the richest of all Greek naukle¯roi—and a metic, to whom, although he had spent a great deal of money on their city and its port, the Aeginetans, according to Demosthenes,39 gave no greater privilege than immunity from their metics’ tax. In Plutarch’s day two anecdotes were still told of Lampis: in one he himself comments on the way his wealth came to him, at first slowly and then quickly;40 and in the other he is the butt of a scornful Spartan witticism.41 39 Dem. XXIII 211. The exact date of the speech is disputed, but I would accept 353/2. 40 Plut., Mor. 787A (¼ An Seni 6); Comm. in Hes. fr. 39; Stob. XXXIX 87. 41 Plut., Mor. 234F (Apophth. Lac. 48); Cic., TD V 40. This Lampis must not be confused, as he is by Obst in RE XII I (1925) 580 and Welter, Aeg. 43, 107 (s.a. 322 B.C.), with the Lampis we encounter some thirty years later in Ps.-Dem. XXXIV, naukleros and moneylender, but a much more insignificant figure, who is described as a slave of Dion and apparently lived at Athens (Ps.-Dem. XXXIV 5–49, esp. 5, 6, 10, 11, 36–7).

386

Athenian Democratic Origins

(V) It has often been maintained that the ruling class of Aegina cannot have been primarily landed, because of the small size and infertility of the island. Against this two lines of argument can be developed: (a) First, the infertility of Aegina has been greatly exaggerated, as the evidence collected by Winterscheidt proves.42 Strabo’s de‘ de x
’ ZB & kat scription of the island is worth quoting: Z vra ayt a ’ ’ ‘ & b
auoyB men ge
vdZB est
i, petr
vdZB d’ epipolZB, kai m
alista Z ’ ‘ pedi
aB
di
oper cil Z p &as
a esti, kriuow
oroB de ikanvB & [‘its land is fertile to some depth, but is rocky on the surface, especially in the plain; for this reason it is completely bare, although it produces a fair amount of barley’, VIII 6.16, p. 375]. (I shall discuss presently the significance of the statement Strabo makes a little later, on the & x
authority of Ephorus, about the lypr
otZB tZB vraB [poverty of the soil] driving the inhabitants to sea trading.) (b) Of course no Grossgrundbesitz [large-scale landed property] in any absolute sense can be attributed to the Aeginetans. But there is an essential fact we must take into account: Aegina throughout its great period was a class oligarchy, and if the aristocrats were few (as they clearly were) and virtually monopolised not only the powers of government but also what good land there was, they could be quite wealthy landowners, according to the very modest standards of ancient Greece. Even in a community the total wealth of which was not above the average,43 such men might be conspicuously rich, simply because they were few as well as all-powerful. Some of the Aeginetans were said to have ‘great fortunes’ in the fifth century (Hdts IX 80.3; cf. V 81.2; and see below); but, Aeginetan society being what it was, their affluent situation is likely to have been balanced by a great deal of poverty among the non-privileged, the great majority of the population. Moreover, rich Aeginetans could afford proudly to display their wealth, to be 42 Aig. 7–8. See esp. E. C. Semple, The Geography of the Mediterranean Region and its Relation to Ancient History (1932), 415, for the way in which the Aeginetans used the underlying mineral fertilisers to increase the agricultural productivity of their stony top soil. 43 The Aeginetans as a whole need not have been more than averagely rich. The very large annual tribute, of 30 talents, paid to Athens by the Aeginetans for some 26 years, down to their expulsion in 431, certainly suggests a prosperous community; but the high assessment may have been a deliberate attempt by Athens to penalise her old and irreconcilable enemy.

But what about Aegina?

387

ostentatiously rich, in a way that would have been dangerous in a democracy like Athens, where a wealthy man who was taken to court would want to be able to plead that he had used his riches not entirely in selfish ways, but for the benefit of the city and his less fortunate fellows. Even in prosperous Athens, with one notorious and misleading exception,44 the largest estates we hear of are of no more than about 70 acres ([c. 28 hectares], the alleged size of the ancestral estate of Alcibiades, Plato, I Alc. 213E), worth perhaps something of the order of four or five talents (see Lys. XIX 29, 42). An Aeginetan noble with no more land than that, who did not have to fear prosecution before an unsympathetic popular court, might make conspicuous display of all the wealth he had, and cut a figure which even richer Athenians might not think it prudent to emulate. Blue blood, and more than sufficient land to live svwr
onvB ’ kai eleyuer
ivB [‘with temperance and liberty’], according to Aristotle’s definition (Pol. II 1265a 28–38, esp. 32–3), with an exceptional concentration upon athletic pursuits, would suffice to content a Greek aristocrat who was a member of a community too small and too much encompassed by larger and stronger neighbours to hope for empire, but who had firm control of the government and did not have to suffer the mortification of living under a democracy. There is an important and neglected piece of evidence which gives strong support to the view I have put forward, of a citizen population divided into a few ‘haves’ and a very much larger number of ‘have-nots’: the fact that Aegina put only the strikingly insignificant number of 500 hoplites into the field in the campaign of Plataea in 47945—compared with 8,000 from Athens, 5,000 from Corinth and 3,000 each from Megara and Sicyon. It certainly looks as if a handful of oligarchs owned most of the land and wealth in the island down to the Athenian conquest in the mid-fifth century, and a high proportion of the citizens (in the broadest sense) were not of hoplite census. These poor citizens would be used as light-armed 44

See my ‘The Estate of Phaenippus (Ps.-Dem. XLII)’, in ASI (Ehrenberg) 109–14. 45 Hdts IX 28.6. With the Persian fleet destroyed, the homeland was in no danger and could not have needed a garrison. Probably there were some Aeginetan hoplites, as well as Aeginetan ships, with the Greek fleet which fought at Mycale (of 110 ships in all, according to Hdts VIII 131.1, with 132.3; 133; IX 90.1 ff.); but we hear nothing about them (cf. Hdts IX 105).

388

Athenian Democratic Origins

and, with slaves,46 as rowers in the fleet, for which the highest recorded number of ships is 70,47 at about the time of Marathon.48 If all the 70 Aeginetan ships in c. 490 were triremes (cf. n. 47), the adult male population of the island, including slaves, must have been well over 14,000, the total of the crews; but this tells us nothing about the total number of hoplites. The whole adult male citizen population, I should say, will certainly have been well under 10,000. Beloch estimated this number at 2,000–2,500,49 and his figures have been accepted by Busolt and others.50 Winterscheidt51 seems to have misunderstood Beloch’s figures as referring only to hoplites (which Beloch in fact believed not to exceed ‘a few hundred’),52 and on this basis he put the adult male citizens at 6,000–7,000; but I myself would prefer a figure rather nearer Beloch’s. The size of the governing oligarchy cannot legitimately be guessed at: it may have been even smaller than the hoplite class. Herodotus (IX 80.3) naively attributes the origin of ‘the great fortunes of the Aeginetans’ to the fact that they bought very cheaply (‘as gold for bronze’) much of the booty obtained from the Persians at Plataea, which was secreted by the Helots and sold off to them. No doubt this sort of thing did happen, whether on a large or small scale; but the notion that Ai’ gin
Ztfi Zsi oi‘ meg
aloi ’ Zn enue ’ ’ enonto [‘the great fortunes of the plo &ytoi arx &yten eg
46 Slaves were evidently used as rowers by many Greek states in the fifth and fourth centuries, e.g. by Corcyra in 433, when nearly four-fifths of the total number on board seem to have been slaves (Thuc. I 55.1: 800 out of 1,050 prisoners); by Chios in 412 (Thuc. VIII 15.2); and by the Peloponnesian fleet based in Aegina in the later stages of the Corinthian war (Xen., Hell. V I 11). On the inflated figure of 470,000 Aeginetan slaves given by Athenaeus, see § 13 below. 47 Hdts VI 92.1. They were defeated by the 70 (ibid. 89) ships of Athens. Winterscheidt, Aeg. 34, thinks they were ‘obviously’ not triremes but penteconters; but I cannot accept this. 48 Cf. n. 12 above. At Artemisium the Aeginetans had only 18 ships (Hdts VIII 1.2), certainly all triremes, and at Salamis only 30 (Hdts VIII 46.1), but we are expressly told by Herodotus that besides these, their best sailers, they had more ships in commission which they kept to guard their island. 49 Beloch, BGRW 122–3, and in Klio 5 (1905), at p. 364 n. 1; cf. 359, and Klio 6 (1906), at pp. 56–7. 50 See e.g. Busolt, GG I2 450 n. 4; GS I 168 [and Afterword, p. 414 below]. 51 Aig. 39 ff.: see esp. 39 n. 78: ‘Die Zahl der waffenfa¨higen ma¨nnlichen Vollbu¨rger scha¨tzt Beloch auf 2000–2500’ (my italics). 52 See the passages cited in n. 49 above, esp. Klio 6 (1906) 56 (‘es kann sicher aber bei der Kleinheit der Insel nur um wenige Hundert handeln’), and the table on p. 57, where the number of 500 is given for hoplites.

But what about Aegina?

389

Aeginetans have their origin in this’], though it is retailed by Herodotus as a fact, without so much as a l
egoysin [‘they say that . . . ’], must be a great exaggeration of what actually happened, a silly slander picked up from the Athenians. There was already considerable wealth in the island by the early fifth century, on Herodotus’ own showing: he speaks of the Aeginetans as being ’ ’ already eydaimon
ifi Z meg
alfi Z eparu
enteB [‘elated by their great ’ ZryktoB [‘unherprosperity’] when they began their p
olemoB ak
alded war’] against Athens (V 81.2). It was certainly for plo &ytoB kai g
enoB [‘wealth and birth’] that the leading families of Aegina were distinguished before ever the Persian wars began: Herodotus (VI 73.2) uses this phrase when describing the hostages taken by Cleomenes in 491, who included Crius the athlete (Section (ii) above) and another, o‘
i per ei
xon m
egiston kr
atoB [‘who possessed the greatest (political) power’].

(VI) Great stress has been laid by writers of the ‘modernising’ school upon a citation of Ephorus in Strabo’s description of Aegina.53 Strabo asserts, on the authority of Ephorus, that Pheidon was the first to strike silver coins, and that he did so in Aegina:54
’ EworoB ’ Ai’ g
infi Z argyron ’
‘ o Fe
idvnoB & d’ en prvton & kopZna
i wZsin yp [‘Ephorus says that silver was first struck into coins in Aegina, by Pheidon’]. Strabo then continues immediately (no doubt still ’ orion g quoting Ephorus), emp
ar gen
esuai [sc. A’
iginan] di a t Zn ’ & lypr
otZta tZB x
vraB tvn & anur
vpvn ualattoyrgo &yntvn ’ ’ ’ emporik vB, & aw’ oyffl t on r‘ vpon & Ai’ gina
ian empol Zn l
egesuai [‘for Aegina became a trading-centre, since because of the poverty of the land her people plied the sea as traders; whence small-wares are called Aeginetan merchandise’]. (We may compare the fragment of Hesiod, quoted by the scholiasts on Pindar,55 crediting the Myrmidons of Aegina with being the first people to build ships and give them sails.) 53

Ephorus FGrH 70 F 176 (cf. 115), ap. Strab. VIII 6.16, p. 376. The run of the sentence shows that this is what Strabo took Ephorus to mean, and not merely that the first coins of Aegina were struck by Pheidon. 55 Hes., fr. 205 (Merkelbach/West), ap. Schol. Pind. Nem. III 21 (III p. 45. 1–8) and Ol. VIII 26d (I p. 242.19–22). [The ‘Myrmidons’ are explained by Strabo VIII 6.16 (p.375).] 54

390

Athenian Democratic Origins

I shall deal with the question of Aegina’s early coinage presently (section (vii)). I need say nothing about the alleged ‘Pheidonian’ coinage of Aegina, since this myth—which first appears (among surviving sources) in this fragment of Ephorus and is not in Herodotus (see VI 127.3)—has been effectively demolished by W. L. Brown.56 It is the latter part of the statement of Strabo (Ephorus) which interests us here. Far from providing evidence in favour of Aeginetan Grosshandel [wholesale trade], conducted by the ruling class, it shows that Ephorus was thinking only of the most petty kind of trade—conducted, therefore, by small men. And the evidence of Ephorus agrees very well with what we find in other sources. The reputation of the Aeginetan merchants was ‘ opvlai, proverbial—but as petty traders, pant
opvlai, rvp
dealers in all sorts of trash. According to Strabo, as we have just ’ seen, Ephorus said t on r‘ vpon & Ai’ gina
ian empol Zn l
egesuai [‘small-wares are called Aeginetan merchandise’], and in the lexicographers it is precisely as hucksters who deal in all kinds of small stuff that the Aeginetans appear.57 It must have been some unknown comic poet who referred to Aegina as xytr
opvliB [‘potseller’, Com. adesp., fr. 350 K/A, ap. Poll. VII 197]. Now although the word x
y trai is sometimes used in a general way for ‘pots’, as xytre
yB for ‘potter’, x
y trai were above all ‘common cooking pots’,58 characteristically unpainted,59 and therefore cheap. We have seen (pp. 383–4 and n. 36 above) that there is no evidence of any fine pottery having been made at Aegina; and in so far as the pots sold by Aeginetan merchants were home-made, they will doubtless have been cheap unpainted wares; but of course the fine ceramics of Corinth and Athens may well have been marketed by Aeginetans. Again, there is Aristotle’s statement in Pol. IV 1291b 17–25. Distinguishing between the demos of Greek cities (in the sense of the lower classes) and ‘those who are called the notables’ (tvn & 56 ‘Pheidon’s Alleged Aeginetan Coinage’, in NC Ser. VI 10 (1950) 177–204 & Pl. XI [and see Afterword, p. 415 below]. 57 Hesych., s.v. Ai’ gina&ia (ed. K. Latte, A 1690); E.M., s.v. Ai’ gina
ia; Steph. Byz., s.v. A’
igina; Schol. Pind., Ol. VIII 29a (I p. 243. 22–4) 58 D. A. Amyx, in Hesp. 27 (1958), at pp. 211–12, with Pl. 48h. See also G. R. Edwards, in Hesp. 18 (1949), at p. 152, with Pl. 16, 15–16 (right). 59 Amyx, op. cit. 212, aptly cites Schol. Ar. Vesp. 279, where the expression x
y tran poik
illeiB [‘You’re decorating a pot’] (like Ar.’s l
iuon e
‘ ceiB [‘You’re boiling a stone’]) is synonymous with useless effort.

But what about Aegina?

391

legom
envn gnvr
imvn), he mentions various categories among the former, one of which is the maritime, [t o to &y d
Zmoy ei
doB] t o peri t Zn u
alattan, of which he proceeds to distinguish four varieties: naval crews (t o polemik
on), merchant seamen (t o xrZmatistik
on), ferrymen (t o porumeytik
on), and fishermen ‘ (t o alieytik
on). In many places, he says, one or other of these varieties is very numerous (pol
yoxlon); and the examples he gives are fishermen at Taras and Byzantium, naval crews (triZrik
on) at ’ Athens, merchant seamen (emporik
on) at Aegina and Chios, and ferrymen at Tenedos. The important thing to notice here is that ’
the emporoi [‘merchants’] of Aegina are members of the lower classes, the demos as opposed to oi‘ leg
omenoi gn
vrimoi [‘those who are called the notables’]. I cannot imagine why Winterscheidt (Aig. 42) should think that Aristotle’s statement is ‘obviously’ derived from Ephorus and therefore relates to the time of Pheidon. This assumption can hardly be proved wrong, but surely it is very much more likely that Aristotle was thinking of his own day and the recent past. The positive evidence of Aristotle, then, will apply directly to the fourth century rather than to the period of Aegina’s greatness. But except that the ruling class, before it was broken by Athenian intervention, was presumably much richer than in Aristotle’s day, and could afford to maintain a navy, what reason is there to suppose that the situation was any different in the seventh, sixth and early fifth centuries?

(VII) According to an Arcadian tradition reported by Pausanias (VIII v 8), Aeginetans, as early as the ninth century, were trading with the Arcadians, via Cyllene on the north-west coast of the Peloponnese—not an activity out of which any great profits were likely to be made. In archaic times it seems very likely that a good deal of Aeginetan wealth was derived from piracy, directed no doubt by the nobility: there is no evidence at all, but the geographical situation of the island, commanding the Saronic Gulf—and thus making Aegina, from the Athenian point of view, ‘the eyesore of the Piraeus’ (Arist., Rhet. III 10.7, 1411a 15–16; Plut., Per. 8.7)— would surely have made recourse to piracy almost inevitable, and the nobles could not have failed to profit most from it. Of the

392

Athenian Democratic Origins

transition we must then assume from piracy to peaceful trade there is equally no evidence. According to some scholars, a decisive step forward in the development of Aegina as a ‘trading city’ was made when coins were struck. In Kirsten’s opinion, for instance, it was now that Aegina took the lead of all the cities of old Greece, and in this we may see the foundation of Aeginetan Handelsmacht [‘trade power’] and its Handelsaristokratie [‘trade aristocracy’]60—in his eyes, ‘mit einem Schlage ist damals aus der Bauern- die Handelsaristokratie geworden’ [at a stroke the landed aristocracy became the trading aristocracy] (‘Review’, 300 n. 1). Aegina’s coinage, Kirsten claims (as if the fact were self-evident), proves that the bearers of her trade were citizens: even if we had no other evidence, he says, the setting up of her mint would be sufficient to establish Aegina’s character as a Handelsstaat [trading state], and not merely a city with metic traders.61 It is true that Aegina may well have been the first state in European Greece to strike coins. But even this is not absolutely certain; and even if it is true, there is no reason to suppose that Aegina began to coin more than a few years before Corinth and Athens,62 whose coinages are now believed to have begun only c. 575 at the very earliest, that of Athens in particular perhaps distinctly later.63 And if Aegina did precede the other cities of old Greece in issuing coins, let us not jump from this to unjustifiable conclusions. The motives which induced a Greek state to coin are obscure to us, and the whole subject needs to be handled very cautiously 60 See his Review, 298–9 (‘Der entscheidende Zeitpunkt . . . ist der Beginn der Mu¨nzpra¨gung. Aigina geht damit allen anderen Staaten des Griechischen Mutterlandes vorauf. Das ist die Grundlage der Vorstellung von der aiginetischen Handelsmacht . . . ’), 300 (‘Das durch die staatliche Mu¨nzpra¨gung bezeugte Interesse der Herren am Seehandel’). 61 ‘Review’, 299: ‘Vor allem aber: das der Staat der Aigineten Mu¨nzen pra¨gt, la¨sst sich nur verstehen, wenn Tra¨ger des Handels nicht die als Demos ausserhalb von ihm stehenden Ha¨ndler waren . . . Wenn wir keine anderen Zeugnisse ha¨tten— die Gru¨ndung der aiginetischen Mu¨nzsta¨tte allein erwiese schon Aiginas Charakter als Handelsstaat.’ 62 This is the opinion of C. M. Kraay, upon whose judgement in such matters I have found that I can rely. [See also Afterword, pp. 415–16 below.] 63 For Corinth, see W. L. Brown, op. cit. (in n. 56 above) 187–8. For Athens, see C. M. Kraay, ‘The Archaic Owls of Athens: Classification and Chronology’, in NC Ser. VI 16 (1956) 43–68 & Pl. XIII; and ‘The Early Coinage of Athens: a Reply’, in NC Ser. VII 2 (1962) 417–23. W. P. Wallace, ibid. 23–42, would date the earliest Athenian coins a good deal later still, c. 545.

But what about Aegina?

393

indeed. A generation ago Keynes gave a salutary warning: ‘I do not think that the act of coinage effected so significant a change as is commonly attributed to it . . . Coinage is not one of the three vital innovations in the evolution of Money . . . It is by no means essential to . . . the designation of the standard by the State, that the State should mint the standard’.64 And in recent years R. M. Cook,65 M. I. Finley66 and above all C. M. Kraay67 have provided a useful correction to simple-minded ‘modernist’ views about the supposed necessary connection between early Greek coinage and trade. I propose to summarise some of the conclusions of an important article by Kraay, which he was kind enough to show me before publication, and in which he has more fully worked out some ideas he sketched in 1960.68 The negative side of Kraay’s thesis, based on the evidence of circulation of Greek coins (mainly provided by hoards) and of the incidence of fractional issues, is that coinage was not ‘originally designed to serve the needs of either local or foreign trade’. As Kraay rightly points out, ‘It is unlikely that in the seventh and sixth centuries B.C. traders were anywhere either so influential or so organised as to be able to secure the public and official adoption of a device designed primarily to serve their interests.’ Against the familiar thesis that coinage was invented to promote foreign trade, Kraay has been able to prove, mainly from 64 J. M. Keynes, A Treatise on Money (1930), I 11–12. Keynes went on to suggest that ‘When the Kings of Lydia first struck coins, it may have been as a convenient certificate of fineness and weight, or a mere act of ostentation appropriate to the offspring of Croesus and the neighbours of Midas. The stamping of pieces of metal with a trade mark was just a piece of local vanity, patriotism or advertisement with no far-reaching importance. It is a practice which has never caught on in some important commercial areas.’ There may be something in Keynes’s emphasis on the psychological element. It is true that Greek coins were at first entirely anepigraphic, and that some of the early types (as Kraay has pointed out: HSCOC 89) are ‘so inexplicit that they defy attribution’: one thinks of some of the earliest coins of Ionia, for example, or the Athenian Wappenmu¨nzen. But, as Kraay insists, ‘wherever it can be identified, the design placed on a coin is the badge of the political authority which issued it’. 65 ‘Speculations on the Origin of Coinage’, in Historia 7 (1958) 257–62, esp. 259–60. 66 ‘Classical Greece’, in id. ed., Second International Conference of Economic History, I: Trade and Politics in the Ancient World (1965), 11–35. 67 In an article of considerable general interest, modestly entitled ‘Caulonia and South Italian Problems’, in NC Ser. VI 20 (1960) 53–82 and Pl. IV, at pp. 78–81 (esp.79); and now in HSCOC. [But see Afterword, p. 415.] 68 HSCOC, esp. the concluding section, pp. 88–91. For the article of 1960, see the preceding note.

394

Athenian Democratic Origins

hoard evidence, that ‘most Greek coinages tended to stay within the areas in which they were minted; to this general statement there were two exceptions, the Thraco-Macedonian mints and Athens, both silver producers, who had surplus metal to dispose of. Neither of these exceptions operated at the very beginning of coinage . . . It would therefore seem that, since most coinages were not exported, and since those that were exported were not among the earliest coinages, the original intention in striking coins was not to facilitate foreign trade, or to provide merchants with a means of purchasing goods or materials not available locally.’ The use of coinage to meet the needs of foreign trade was ‘a secondary development’. The arguments adduced by Kraay against the supposition that coinage was intended to assist internal or local trade are equally cogent. First, in the early days ‘few of even the most important Greek states possessed a regular supply of small denominations’ adequate for local transactions. ‘Second, very many places, especially in the sixth century, had no coinage of any sort. And third, coinage originated, not among the silver-using states of mainland Greece and the West, but with the electrum issues of Asia Minor; these, even in their smallest fractions, must certainly have represented values much higher than those required for retail trade. The use of coinage in retail trade which, even at the end of the fifth century, was confined to a few of the more economically advanced states, cannot be regarded as its original purpose.’ On the positive side, Kraay has developed the theory he produced in 1960. He suggests that city ‘governments’ found it convenient to issue— probably at some profit to themselves—coins the use of which could be made obligatory in official transactions such as the discharge of taxes and fines, and (in the reverse direction) the division of surpluses among citizens, the payment of mercenaries69 and soldiers, also of salaries given to experts such as doctors, and expenditure on public works. A state which issued coins ‘could insist on payment in units which it had itself created, and of which the quality was therefore known’. As for Aegina herself, Kraay points out (HSCOC 78–9) that although her coins, especially the sixth-century ‘turtles’, are found over a wide area, ‘yet over most of this area turtles provide only a very small portion of any find in which they occur’: this is 69

This has been emphasised by R. M. Cook: see n. 65 above.

But what about Aegina?

395

true even of Egypt, where one might have expected the interest of Aeginetan traders in Naucratis (see section (ix) below) to be reflected in higher proportions of Aeginetan coins in Delta hoards. Only in the more immediate neighbourhood of Aegina—roughly the triangle formed by Corinth, Rhodes and Crete—do we find hoards in which half or more of the contents are Aeginetan. I might add that if Aegina’s port was indeed a busy one at the beginning of the sixth century, and she had a good revenue from customs duties, Aegina was likely, even on the explanation of the origin of coinage which I have adopted, to begin coining early, if only to facilitate the payment of duties. From the fact that Aegina very probably began to coin a few years before Corinth and Athens, then, we cannot justifiably draw inferences about the character and outlook of the Aeginetan ruling class, or about the civic status of the merchants who traded from Aegina. The Phoenicians certainly traded on a considerable scale, yet the Phoenician towns almost certainly did not issue any coins before the third quarter of the fifth century. At Carthage there seem to have been merchants even among the governing class [Ar. Pols. V 1316b5–6, cf. II 1273a 33–5], yet there is certainly no substantial coinage of Carthage proper before about 350, when large gold issues appear, although Carthaginian settlements in Sicily, such as Motya and Panormus, struck coins in the fifth century, and the Carthaginian ‘Siculo-Punic’ issues probably go back to the Carthaginian campaigns in Sicily in the 390s and an isolated gold issue of Carthage itself to about the 380s.70 Again, among the Greek cities in the West, Syracuse is likely always to have been of greater ‘commercial’ importance (in the sense of having a greater volume of external trade) than Himera, Selinus and Naxos, yet she seems to have begun to coin only c. 510, whereas they (with Zancle) had started between two and three decades earlier, and at least three of the Greek cities in south Italy, namely Sybaris, Croton and Metapontum, whose ‘commercial’ importance must have been very small, had been minting coins earlier still, c. 550.71 70 See G. K. Jenkins and R. B. Lewis, Carthaginian Gold and Electrum Coins (1963), esp. 18. 71 Several statements in this paragraph are made mainly on the basis of information kindly provided by Kraay, and with his approval [see now Kraay ACGC, 233–5 (Carthage, and Siculo-Punic mints); 204–9 (Sicilian cities); 163–8 (South Italy); Phoenicia: D. B. Harden, The Phoenicians (1962), 166–8)].

396

Athenian Democratic Origins

When Kirsten (‘Review’, 299) says that ‘die Gru¨ndung der aiginetischen Mu¨nzsta¨tte allein erwiese schon Aiginas Charakter als Handelsstaat, nicht nur als Polis mit Ha¨ndlern aus dem Metoikenstande’ [‘the foundation of the Aeginetan mint is in itself a sufficient indication of the character of Aegina as a trading state, not merely a polis with traders from among the metic class’], his statement has not the least justification. Similarly, the attempt of Sutherland72 to show ‘that possession of a reputable coinage was among the first necessities for a vigorous and progressive Greek state, and that a state, once it was happily possessed of a good coinage, engaged immediately in what was actually, if not openly, a national commercial policy designed to supply her with the essentials of life’, fails entirely (quite apart from the fact that his chronology for the coins of Aegina, Athens and Corinth is too high), as Finley (op. cit. in n. 66), and by implication Kraay (HSCOC: see above), have now demonstrated. Statements in Sutherland’s paper, such as the assertion that in the sixth century ‘it seems that Athenian pottery may have been marketed abroad by Corinth’ (op. cit. 142; my italics), show a failure to understand the basic facts about the mechanism of Greek trade, in which states as such took no part.73

(VIII) Nor is the existence of standards which are nowadays associated especially with Aegina, one for weights and coins, and another (unrelated, I believe, as elsewhere: see Ch. 8) for measures of capacity, a fact which has any relevance to this controversy. In the first place, what we tend to conceive as a ‘coin standard’, and can very rarely identify in relation to a given state as anything else, will normally have been employed before the days of coining, as a weight standard. The ‘Aeginetan’ standard, although it appears widely throughout the Peloponnese and central Greece, as far north as Thessaly, and also in the islands of the Aegean, and in a few of the cities of Asia Minor (see Head, HN 2 xlv), made no progress farther afield and was not even adopted by Corinth, 72

‘Corn and Coin’, in AJP 64 (1943) 129–47, at p. 131. As I have said already, there is no sign of ‘merchant fleets’, and the ‘nationality’ of the merchant was ordinarily of little or no consequence, except perhaps in time of war. 73

But what about Aegina?

397

Athens or the cities of Euboea. But the significance to be attached to these facts is anything but clear, for the dispersion of the standards in question as weight standards will have been established well before the appearance of coins; and what we have to consider, therefore, is not so much the spread of an ‘Aeginetan coin standard’ in the sixth century as of a weight standard common to Aegina and many other states, at an earlier period—although it is quite possible, of course, that some states had no official weight standard before they began to issue coins. I am not even sure there is any good ancient evidence that there existed in antiquity any conception corresponding to ours of an ‘Aeginetan standard’. At Athens itself the Aeginetan drachma is said to have been called paxe&ia draxm
Z [the heavy drachma]. Pollux, who preserves this information (IX 76; cf. Hesych., s.v. & paxe
ifia draxmfi Z), attributes the fact to Athenian ‘hatred of Aegina’, but I would take leave to doubt this. Pollux is speaking ‘ Ai’ gina
ia draxm
specifically of the Aeginetan drachma (Z Z), not of drachmae coined on an ‘Aeginetan standard’; but the Aeginetan, although doubtless the most common of such drachmae at Athens, will not have been the only ‘heavy’ one to appear there, and the Athenians perhaps used the expression paxe&ia draxm
Z to cover all drachmae of what we should call ‘Aeginetan standard’. The treaty between Athens, Argos, Mantinea and Elis of 420 B.C., which speaks of ‘Aeginetan obols’ and an ‘Aeginetan drachma’ (Thuc. V 47.6; IG I 3 83 ¼ Tod I 2 72, lines 23–4 [not included in M/L]), may conceivably be referring to actual coins of Aegina: we know from the lexicographers that in ancient times the Aeginetan ‘tortoises’ were ‘the coinage of the Peloponnese’ (see Poll. IX 74; Hesych., s.v. xel
vnZ). By 420, however, the Aeginetan mint had entirely ceased to function for some years—eleven at least, and probably nearer thirty than twenty, according to the date we give to the Athenian coinage decree.74 But would anyone accept, as an Aeginetan drachma, an old and underweight Aeginetan coin of, say, the sixth century?—coins which evidently remained in actual use until well into the fourth century (see Kraay, HSCOC 78 n. 12). I agree with the suggestion made to me by Kraay that by 74 [Now IG I3 1453 ¼ M/L 45; and see SEG XLVIII (1998). The date remains contentious: D. M. Lewis, in I. Carradice (ed.) Coinage and administration in the Athenian and Persian Empires (1987) 53–63, repr. in his Selected papers in Greek and Near Eastern history (1997), 116–30, gives a lucid account of the problems.]

398

Athenian Democratic Origins

the expression ‘an Aeginetan drachma’ we should understand something like ‘a quantity of silver (of any sort) amounting in weight to the drachma issued by Aegina’. This, I suggest, is the nearest the ancient world came to formulating the conception of an ‘Aeginetan standard’; and such an expression would be used only in areas where the actual coins of Aegina circulated in quantity— elsewhere, men would speak of ‘an Athenian drachma’, ‘a Cyzicene stater’, and so forth. In 412 Tissaphernes promised pay to the Spartan fleet at the rate of one Attic drachma per man per day ’ draxm (eB Zn ’Attik
Zn, Thuc. VIII 29.1; contrast 28.4). What this meant in practice may be seen from the unique tetradrachm of Attic weight, bearing on the obverse a superb Persian head (‘one of the earliest, and incidentally one of the finest, portraits on any coin’, as E. S. G. Robinson has said),75 almost certainly that of the satrap himself, and on the reverse the usual owl and olive spray, with the first three letters of the king’s title, BAS, replacing the familiar AQE. I see not the least reason to suppose that what we call ‘the Aeginetan standard’, for weights and coins, either originated in Aegina or owed its fairly wide diffusion (principally on the Greek mainland and in the Aegean) to the ‘Aeginetan trade’. Instead of adopting these hypotheses, which are so often taken for granted nowadays, G. F. Hill observed over half a century ago that the Aeginetan ‘was the weight standard in use all over the Greek mainland as far north as Thessaly from very early times. When the Aeginetan mint was started, it would naturally not create a new standard, but rather adopt one which was likely to favour the widest possible currency for its coins’ (Historical Coins of the Greeks (1906), 5). I would conclude that Aegina’s coin standard was the weight standard she herself already used, in common with the states of the Peloponnese and many of those in central and northern Greece; and that they adopted it, when they began to coin, for the same reason she did: that it was already their native weight-standard. In some cases we can be sure that the ‘Aeginetan’ coin-standard was not chosen because of any direct ‘Aeginetan influence’: the Thessalians, for example, began to coin [on that standard] only after 480 when Athens had outstripped Aegina in 75 ‘Some Problems in the Later Fifth Century Coinage of Athens’, in Amer. Numism. Soc. Mus. Notes 9 (1960) 1–15 & Pl. I–II, at p. 4 (with Pl. I 7).

But what about Aegina?

399

both naval and commercial importance, the Locrians only in the fourth century, long after the collapse of Aegina.76 Some of the outlying cities which adopted the same coin-standard as Aegina may possibly have taken it directly from her; but that is the most we can say.

(IX) The relation of Aegina to Naucratis has been a favourite argument of those who conceive Aegina as a ‘trading city’ above all. Virtually the only piece of evidence, apart from that provided by archaeology,77 is the passage in which Herodotus (II 178–9) describes the foundation of Naucratis. I shall concentrate on those parts of Herodotus’ statement which are of immediate interest to us. The largest, most famous and most frequented sanctuary (t
emenoB) at Naucratis, he says, namely the Hellenion, was founded jointly by Chios, Teos, Phocaea, Clazomenae, Rhodes, Cnidus, Halicarnassus, Phaselis and Mytilene (178.2). These nine cities and these ’ alone provided prost
atai to &y empor
ioy [officials of the emporion] and any other cities which claimed a share had no right to do so. & Separate sanctuaries were also founded: Ai’ ginZtai . . . Di
oB, kai ’
llo S
a amioi
‘ HrZB kai Mil
Zsioi ’Ap
ollvnoB [‘Aeginetans . . . a sanctuary of Zeus, Samians another of Hera, and Milesians one of Apollo’] (178.3). Several different theories have been developed to explain this brief statement, most of them giving a far-fetched interpretation of it or even directly contradicting it. I think the work which has done most to lead scholars astray is that of H. Prinz, Funde aus Naukratis (Klio, Beiheft 7), published in 1908, a monograph—admirable in some ways, considering the date at which it was written—that took for granted the ‘modernising’ view of the Greek economy previously developed by Meyer and Beloch (see esp. pp. 146–7). Prinz believed that in addition to an emporion run by the city of Naucratis itself, there ‘must have 76

[For Thessaly (which had previously coined on the Persian standard) see Kraay ACGC 115; for Locris, ibid. 122–3.] 77 A convenient bibliography up to 1937 will be found in R. M. Cook, ‘Amasis and the Greeks in Egypt’, in JHS 57 (1937) 227–37, esp. 227 n. 6. The latest work I know of any importance is that of J. Boardman, ‘Chian and Naucratite’, in BSA 51 (1956) 55–62, which refers to all the more recent publications. [See Afterword, pp. 416–20 below, for more recent work on Naucratis.]

400

Athenian Democratic Origins

been’ a separate one corresponding to each of the four sanctuaries (op. cit. pp. 5–6, cf. 115–16), so that there were in all no less than five emporia! In support of this curious theory Prinz produced nothing in the way of evidence or even argument. Others, Glotz for instance, reduced the five emporia to four by discarding the emporion (and sanctuary) peculiar to the state of Naucratis; but Glotz’s picture,78 in which ‘special temples and quays were reserved for the Milesians . . . ,79 the Samians, and the Aeginetans’, and there were ‘four warehouses dominated by temples’ (for ‘at Naucratis every emporium had its guardian deity’), is otherwise very much that of Prinz. In the first volume of the[ir] Histoire grecque, Glotz and Cohen could even say, ‘Sous la protection de leurs divinite´s, les ‘‘nations’’ de Naucratis eurent chacune leurs magistrats et leurs tribunaux, avec un droit d’appel a` la justice de la me´tropole. Naucratis, ville internationale, fut comme le prototype d’Alexandrie, une sorte de Shang-Haı¨ antique’ [‘Under the protection of their gods, the ‘nations’ of Naucratis would each have had their own magistrates and tribunals, with a right of appeal to the justice of the mother-city. Naucratis, the international city, could be seen as the prototype for Alexandria, a sort of ancient Shanghai’].80 The groundless idea that the founding of a sanctuary at Naucratis by citizens of a particular state somehow implies the possession of a ‘factory’ at Naucratis belonging to that state constantly appears, as in the remark of Seltman,81 ‘Only one state of Greece proper secured the privilege of a factory at Naucratis— Aegina.’ A more general statement, which however departs even further from Herodotus, is that of Andrewes, that ‘the management [of Naucratis] was vested in nine East Greek cities who shared a common sanctuary, and three greater states who built separate temples’ (GT 118; my italics). Hasebroek (TPAG 64–5) produced a picture much more in accordance with Herodotus, in that he confined the control of the emporion of Naucratis to citizens of the nine states participating in the foundation of the Hellenion; but G. Glotz, TGA 130, 145, 138 ¼ AGW 106–7, 119, 113 (my italics). Glotz adds, ‘who enjoyed undisputed pre-eminence’. It is not clear to me what authority he believed himself to have for this statement. 80 HG I 205. The only authorities given for the statement made in the first sentence quoted above are Hdts II 178 and Hermeias (FHG II 80–1) ap. Athen. IV 149D–F. 81 Seltman, GC2 82, referring also to a ‘Milesian factory at Naucratis’. 78 79

But what about Aegina?

401

even he followed Prinz to the extent of treating the members of those nine states, and of the other three, as forming communities ‘very distinct from the full citizens of Naucratis’, for of course he regarded it as a confirmation of his general theory about the relation between trade and politics in the Greek world that (as he saw it) ‘even in Naucratis trade was in the hands of non-citizens and foreigners’. Against all this, Roebuck (ON 215; cf. ITC 134–5) has rightly pointed out that ‘Herodotus does not state or imply that there were emporia attached to the separate temene for purposes of trade’, and that from the excavations ‘the sanctuaries appear to have been only religious establishments with small temples, altars, and open precincts . . . There was probably a single dock and warehouse area along the river bank’, with ‘a common regulating authority’. So far, so good. Roebuck believes, however, that Herodotus’ ’ prost
atai to &y empor
ioy [officials of the emporion] were ‘the chief magistrates or leaders of the whole community of Naucratis’, which ‘had developed into a unified Greek state at an early date’ and ‘acted and was recognised by other Greek cities as a normal city state’ (ON 216). The decisive step, he thinks, was ‘the creation of Naukratis as a unified political community through and around the Hellenion. When Herodotus speaks of its founding states as furnishing the magistrates of Naukratis, he is speaking inexactly; it is rather the citizens originally from those states and their descendants, who were still aware of their origin through the continuance of their cults, who furnished the magistrates.’ The nine states participating in the Hellenion were ‘the political founders of Naukratis and their descendants formed the main element of its population’ (ON 218). It seems to me that this reconstruction not merely goes far beyond our meagre evidence, but strains unbearably the meaning ’ ‘ of Herodotus’ prost
ataB to &y empor
ioy aytai ai‘ p
oli
eB ei’ si ai‘ par
exoysai [‘these are the cities which provided the officials of the emporion’], and moreover depends upon one assumption which cannot be proved and is in my opinion almost certainly false: that Naucratis had become a proper polis by the time of Herodotus.82 82 And see Cook, op. cit. (in n. 77 above), 233 n. 29: ‘Naucratis was not a Greek colony, properly founded, so that there may have been few data about its origins. Two centuries later Apollonius Rhodius wrote a Naykr
atevB kt
isiB (Athen. vii. 283): but how far he is likely to have used historical evidence I do not know. Strabo’s

402

Athenian Democratic Origins

The scanty evidence very strongly suggests that Naucratis was still not a polis but a mere emporion at the very end of the fifth century and did not become a polis until some time in the first half of the fourth. As far as I know, the earliest references to Naykrat&itai [‘citizens of Naucratis’] are from about the mid-fourth century.83 That Herodotus refers to Naucratis as a polis in II 178.1 is not significant, for he can use the word quite untechnically, as when he calls Babylon a polis (Hdts I 178.1). The main pieces of positive evidence that Naucratis was not regarded as a proper polis by the Greeks before the fourth century are as follows: (a) In a decree of the new state of Rhodes, to be dated probably 411–408, just before the completion of the synoecism, a man who is appointed Rhodian proxenos is described as Ai’ g[ : . . . . . . t]on ’ Naykr
ek atj[ioB].84 The ethnic is usually restored Ai’ g[in
: atan], y ption] has also been suggested. In either event the but Ai’ g[
: fact remains that the man who is evidently to hold the Rhodian proxenia at Naucratis is described not as a citizen of Naucratis but as an Aeginetan (or Egyptian) from Naucratis. I find it hard to believe that the Rhodians would have failed to choose a citizen of Naucratis as their proxenos there had Naucratis been a proper polis at the time. (Incidentally, the Rhodians are among the states mentioned by Herodotus II 178.2 as participating in the appoint’ ment of prost
atai to &y empor
ioy. It is interesting to find them feeling the need to appoint in addition a member of some other state dwelling in Naucratis as their proxenos.) (b) The most decisive evidence is a decree of Lindus, contained in an inscription discovered at Naucratis85 and passed probably a little earlier than the Rhodian decree mentioned above. In appointing a Lindian proxenos at Naucratis it describes the ’ Ai’ g
yptvi oi’ k
eonta man concerned as Dam
ojenon
‘ ErmvnoB en [Damoxenos son of Hermon, living in Egypt] and provides for a ’ Ai’ g
ypt[vi en ’ tvi copy of the decree to be set up en & ‘E]llan
ivi date for the foundation of the MilZs
ivn te&ixoB (xvii. 801) could fit the archaeological conclusions about Naucratis’ [see also Afterword, p. 417 below]. 83

See SIG3 239 B 37 (363 B.C.). And in IG II2 206 (of 349/8) a man named Theogenes, appointed Athenian proxenos, is described as Qeog
enZB o‘ Naykrat
itZB (lines 7–8, 19–20). Dem. XXIV 11 is not informative. 84 SIG3 110, the most recent edition of which is by Chr. Blinkenberg, Lindos II. Inscriptions (1941), i cols. 210–14, no. 16, lines 3–6. 85 See SIG3 110 n. 4, and for a better text Blinkenberg, op. cit. col. 212–14, App. to no. 16, lines 4–5, 16–18 [and see Afterword, p. 418 below].

But what about Aegina?

403

[in Egypt in the Hellanion]. Here, the existence of a polis Naucratis could hardly be more specifically denied: the proxenos is an ‘Egyptian’ and the Hellenion is ‘in Egypt’. ’ Ai’ g
yptfiv oi’ k
eonteB (c) Herodotus II 180.2 speaks of oi‘ en
‘ EllZneB [the Greeks living in Egypt] who are not called Naykrat&itai [citizens of Naucratis] although they must have been mainly the inhabitants of that place.86 (d) Naucratis seems not to have issued coins until the fourth century.87 (I would not, of course, claim that this fact by itself is of any great significance.) I do not see how we can give a confident explanation of Herodotus’ rather cryptic statement about the prost
atai to &y ’ empor
ioy [officials of the emporion], but the one which seems the most probable to me, and certainly gives the most natural interpretation of Herodotus’ words, is that each of the nine cities concerned chose from among its own citizens one prostate¯s of the Naucratite emporion, a man who would ‘represent’ the interests of its citizens there, as we should expect of one bearing the title prostate¯s [lit., ‘one who stands before].88 Elections may have been made at regular intervals, or (more likely) each state may have chosen its prostate¯s for an indefinite period, perhaps so long as he continued to reside at Naucratis—speculation on such points is futile. From the very fact that Herodotus mentions these officials in his brief description of Naucratis, it is evident that they were important. Whether the ‘emporion’ in which they functioned was an area within Naucratis (the port and its quays and warehouses), 86 [Ste. Croix has added: ‘Abusir Milesian c. 500’, a reference to the GrecoEgyptian grave-stele illustrated in Boardman GO 136 fig. 159, presumably as an example of a Greek living in Egypt but not at Naucratis; but the ‘Milesian’ is now believed to be a Carian (Boardman, op. cit. 137).] 87 See Head, HN 2 845; and E. T. Newell, Miscellanea Numismatica (NNM 82, 1938), 60 ff. (Naucr. IVc AR coin: NAU, obol., gr. 0.64, imitation of Athenian type). 88 The choice of epidemiourgoi [magistrates] for Potidaea by Corinth down to 432 (Thuc. I 56.2) provides a partial analogy. If each state simply provided its own prostate¯s, there is no reason to object, as Roebuck does (ON 216), that the arrangement ‘would necessitate a most unusual degree of co-operation between states which were, on different occasions, at variance elsewhere in the Aegean’. Nor can I follow Roebuck’s other objection (loc. cit.), that ‘some of them, namely Phocaea, Teos, and Miletus, ceased to exist as important and independent communities in the archaic period’: all our nine states continued to exist as poleis beyond the fifth century, if with different degrees of independence and importance.

404

Athenian Democratic Origins

or whether Naucratis was the emporion89 and whether there existed in the sixth and fifth centuries, either beside or above the prostatai, more important officials, for example the timoxenoi who are attested much later,90 we have no means of telling. Herodotus’ statement about the foundation of sanctuaries by Aeginetans, Samians and Milesians is made after he has already told us not only about the Hellenion but also that no states other than the nine which founded it had any claim to participate in the provision of prostatai; and there is no suggestion in Herodotus that Aeginetans, Samians or Milesians had been given any special privileges at Naucratis except the right to found sanctuaries. Clearly citizens of all three states were trading at Naucratis. Attic and Corinthian wares were both used extensively in Aegina, and it is very likely to have been Aeginetan traders who carried the vases made at Athens and Corinth which have been found in quantity at Naucratis.91 There have also been found in Aegina scarabs and faience figurines which are said to be of Graeco-Egyptian origin (probably from Naucratis) and painted pottery of the style commonly known as ‘Naucratite’, which is thought by some archaeologists to have been manufactured in Chios, but by others92 to be partly the product of Chian potters working at Naucratis itself. The Aeginetan sanctuary to Zeus at Naucratis has unfortunately not been identified, as have the Hellenion, the sanctuaries of Apollo and Hera (undoubtedly those of the Milesians and Samians), and two others, one evidently dedicated to Aphrodite (and probably founded by Chians),93 and the other to the Dioscuri. Although the stratification is far from certain,94 it seems that the 89

In Hdts II 179 Naucratis itself is certainly described as an emporion. And see Roebuck, ON 215 and 219 n. 22. 90 Hermeias, as cited in n. 80 above. Cf. Roebuck, ON 218 n. 4. 91 See esp. J. D. Beazley and H. G. G. Payne, ‘Attic Black-figured Fragments from Naucratis’, in JHS 49 (1929) 253–72; R. M. Cook, in BSA 44 (1949) 154–61. 92 Including Boardman: it will be sufficient to refer to his article cited in n. 77 above, where full references are given to earlier work. I find Boardman’s arguments very plausible. [His theory, briefly restated in The Greeks Overseas4 (1999), 123–4 (with notes at 274) and Early Greek Vase Painting (1998), 144–5 (bibliography on 279), has aroused considerable controversy; the arguments have recently been summarised by A. Mo¨ller in her Naukratis (2000), 136–40, who concludes (140) that it is ‘quite conceivable’.] 93 See Roebuck, GTGE 241–2; ON 217; Boardman, op. cit. in n. 77, 61–2. 94 See esp. R. M. Cook, in JHS 57 (1937), at pp. 227–8; and in BSA 34 (1933–4), at p. 86 n. 2.

But what about Aegina?

405

sanctuaries of Apollo and Aphrodite probably belong to the earliest occupation levels, while the Hellenion is later, presumably of the time of Amasis (c. 569–525).95 Milesians, as the ancient literary evidence suggests,96 may well have been the original founders of the settlement, although the evidence provided by Herodotus and the excavations suggests that even if they were there first their influence did not remain predominant.97 Although the Aeginetans were certainly the only people of old Greece to have a sanctuary to themselves at Naucratis, it is clear in the light of the foregoing that there is no reason to think the polis of Aegina had any say in the administration of Naucratis or its port. And with all due diffidence I am inclined to go further and suggest that Herodotus may be making a technical point when he refers specifically to the provision of prostatai of the emporion by the nine p
olieB he names, but speaks of the three separate sanctuaries & as being founded by Ai’ ginZtai, S
amioi, and Mil
Zsioi, without using the article. It is true that Herodotus quite often omits the article when speaking of individual poleis, even when they are represented as acting in their official capacity; but I suggest that in this particular case he may well mean just what he says: that the three separate sanctuaries were founded by ‘Aeginetans, Samians and Milesians’—individual members of the three cities concerned who were trading at Naucratis and, wishing to make a prolonged stay there, introduced the cults of the gods they mainly worshipped at home, with or without the official sanction of their own states. At any rate, the inscribed dedications recovered from the various precincts show, as Roebuck has well demonstrated (GTGE 241–2; cf. ON 212, 217), that neither the Hellenion nor the individual sanctuaries which have been excavated at Naucratis were restricted 95

This opinion has been expressed to me by Boardman. Cf. C. C. Edgar, in JHS 25 (1905), at p. 136. 96 See Strabo XVII i 18, p. 801, and the other sources cited by Ure, OT 103–5. [The value of this evidence is disputed: see Afterword, p. 417 below.] 97 R. M. Cook (see n. 94 above) argues for a foundation date of c. 615–10. Others prefer a slightly earlier date: see Roebuck, GTGE 244 n. 7. Boardman tells me that among the sherds found at Naucratis are a piece of early Attic BF now in Toronto, of c. 615, a scrap of Corinthian of c. 625 or rather earlier, and East Greek pottery of the second half of the seventh century which we cannot yet date with any precision: he would see no objection to dating the foundation of Naucratis some time in the third quarter of the seventh century (still within the reign of Psammetichus I).

406

Athenian Democratic Origins

to the citizens of the states which had founded them. Although there is likely to have been ‘a tendency among visiting traders and residents of the first and second generations to frequent the sanctuaries of their own gods, . . . there does not appear to have been any tendency to maintain the sanctuaries as the separate property of a far away state over a long period of time’ (GTGE 242).

(X) I come now to the well-known passage in which Herodotus (VII 147.2) tells how Xerxes in 480 saw the corn ships at the Hellespont sailing by ‘to Aegina and the Peloponnese’. This story, whether true or false, is likely to come from the early fifth century (or Athens would surely have replaced Aegina as the destination of the corn ships), and I personally would accept it as evidence that at that time Aegina was a main entrepoˆt of the Pontic corn trade with Greece. It is quite wrong to infer, however, as Kirsten does (‘Review’, 299), that those who carried this trade ‘can only have been’ large-scale merchants (grosse Unternehmer) and must therefore have been the very aristocracy of Aegina, for there is a very obvious analogy which entirely destroys the argument. No one will wish to dispute that in the later fifth century and in the fourth Athens played at least as great a role in the corn trade between the Pontic area and Greece as ever Aegina could have done earlier, and probably indeed a much greater one. Yet, as we have seen, there is not a single scrap of evidence to connect this trade, or indeed any other branch of ‘Athenian’ mercantile activity (in the sense of the trade flowing into and out of Attica) with those Athenians who were the wealthiest and the most politically important and influential: a large part of ‘Athenian’ trade was in the hands of metics and other foreigners, and not a single one of the citizen merchants we hear of was a man known to have taken any part in politics.98 All that we are entitled to conclude from Herodotus’ story, therefore, is that the island of Aegina—not ‘the Aeginetans’— played a significant part in the early fifth-century corn trade between the Pontus and the Peloponnese,99 just as the geographical position of the island would lead one to expect. When Percy 98 Andocides is the exception who proves the rule. [See OPW 265–7, and the introductory note to this chapter.] 99 By 427 (see Thuc. III 86.4) there seems to have been a significant import of Sicilian corn into the Peloponnese; but this trade may have grown up to some extent

But what about Aegina?

407

Gardner100 took Herodotus’ statement as evidence that ‘after the fall of Miletus there came a time when the trade of the Black Sea fell partly into the hands of the Aeginetans’ (my italics), he was drawing an unwarrantable inference. Whether the corn was brought to Aegina, or taken from it, by merchants who were Aeginetan citizens, we do not know. The Aeginetan state must certainly have derived profits from customs duties on the corn, probably twice over, on entry and exit; but we have no information who the merchants were who conducted the traffic. Some of them are very likely to have been Aeginetans, but the majority, I imagine, will as usual have been metics and foreigners of all kinds.101

(XI) The ‘medism’ of Aegina in 491 (Hdts VI 49(-73)) is always explained nowadays by the fact that the island was ‘dependent on its eastern trade’. This again is a pure assumption, and (I believe) a false one. Who were the other principal medisers in 491–79? The most prominent were the landowning Aleuads of Thessaly (Hdts VII 6.2; 130.3; 172.1; IX 1) and the Theban aristocracy (Thuc. III 62.3–4; Hdts VII 205.3; IX 15.4; 38.2; 40; 67; 86–8; cf. Plut., Aristid. 18.7; Paus. IX vi 2). Of the latter, who were anything but a ‘commercial aristocracy’, Thucydides makes their fellow citizens say in 427 that they hoped with Persian help to keep themselves the more firmly in power (III 62.4). Is not a precisely similar motive, a political one,102 likely to have inspired the medism of the Aeginetan oligarchs? Why should we look for another? I would go further, indeed, and suggest not merely that the conception of Aegina’s ‘dependence on her eastern trade’ is an unwarrantable assumption a priori but that it is positively wrong. Once more as a replacement of an earlier Pontic supply which had been reduced or even entirely cut off by Athens during the ‘First Peloponnesian war’ of c. 460 ff. and the war of 431 ff. My own feeling is that most Peloponnesian cities, except perhaps Corinth and Megara, fed themselves mainly on home-grown corn but would need to import whenever the local crop failed to give a good yield—that is to say, perhaps fairly often. 100

History of Ancient Coinage, 171. Cf. Lionel Casson, The Ancient Mariners (1959), 112. 101 [A note by Ste. Croix indicates that he intended to revise this paragraph.] 102 The only writer I have come across who gives a political explanation is Welter, in AA (1954) 30–1 (§ XXVI). Sixteen years earlier, in his Aig. 34, he seems to have accepted the usual explanation.

408

Athenian Democratic Origins

historians have allowed themselves to forget that the trade of a Greek state was not normally carried on by the citizens of that state, and that the trade between Aegina and the East would not have been regarded by the Greeks—or, for that matter, by the & Persians—as the trade of oi‘ Ai’ ginZtai. Moreover, trade evidently went on unhindered throughout the fifth century between mainland Greece and the Greek settlement at Al Mina near the mouth of the Orontes; and indeed the painted pottery which went to Al Mina was entirely Athenian,103 the Persians evidently making no attempt to interfere with this trade, although had they wished to do so the situation of Al Mina was such as to leave it entirely at the mercy of the Phoenician navy. In any event, the only evidence we have about the nature of ‘Aegina’s Eastern trade’ in the early fifth century, namely the statement of Herodotus discussed in section (X) above, relates to the corn trade with the Pontus, conducted through waters in which the Persians showed not the slightest interest and into which (as far as I know) there is no reason to think they ever sent a navy except in support of large-scale military operations in Europe.

(XII) Some of the believers in an Aeginetan commercial and manufacturing ruling class have seized upon any scrap of evidence which can be twisted to fit their conception. According to Herodotus (V 88.2; cf. Athen. XI 502C), after the defeat of the Athenian attempt to get possession of the wooden statues of Damia and Auxesia, the Argives and Aeginetans made a rule to bring nothing Athenian, pottery or anything else, into the temple (of Damia and Auxesia), ’ xytr
idvn epixvri
’ but to drink there in future ek evn [‘from locallymade pots’]. On this Macan104 commented, ‘The exclusion of Attic ware from the cult . . . may possibly be an understatement and 103 See C. L. Woolley, in JHS 58 (1938), at pp. 21–3; S. Smith, in Antiq. Jnl 22 (1942) at pp. 109–11. 104 Macan, Hdts IV–VI I 232n. Cf. How and Wells, CH II 49; Glotz-Cohen, HG I 315 n. 104 (where the embargo seems to be conceived as a general one ‘in time of war’); Bo¨ckh, SA I3 73–4 (‘In Aegina und Argos scheinen sogar fru¨hzeitig Attische Fabrikate verboten worden zu sein, wiewohl aus einem angeblich religio¨sen Grunde, und zuna¨chst fu¨r den heiligen Gebrauch’—my italics); Ehrenberg, PA2 325.

But what about Aegina?

409

pseudo-explanation of a commercial measure or custom for the protection of native wares from Attic competition . . . The ritualistic facts are probably correct: the reasons given therefore highly suspicious.’ Taking up Macan’s suggestion with avidity, Ure (OT 167–8, 314–20 (esp. 319–20)) discussed the supposed commercial embargo at great length. All this is pure fantasy.

(XIII) I do not think anyone nowadays will wish to defend the fantastic figure of 470,000 slaves105 which Aegina is said to have once possessed, according to Athenaeus (VI 272D) and a scholiast on Pindar (Schol. Ol. VIII 30d (I p. 244.21–3), cf. 30l (ibid. 345.24– 5)), both purporting to cite Aristotle’s lost Constitution of Aegina (fr. 472 Rose106). The total population of Aegina in the census of [1951] was [8,859].107 Welter produced the theory that there was a great slave market on Aegina, fed in particular by enslaved debtors of the Athenian and Megarian upper classes (in the seventh century).108 This may be true, but I know of no evidence. I do not see how we can make a confident estimate of the population of Aegina at any time [in antiquity] (cf. section (v) above).

(XIV) The evidence and the arguments I have set out above strongly suggest that Aegina in her great days had a small, wealthy, landed ruling class, which retained right down to the Athenian conquest 105 Beloch, BGRW 84–5, 95–6, wished to reduce it to 47,000, but even this may be much too large a number. I do not see how we can profitably guess what Aristotle said, or what the maximum number really was. 106 This is the only surviving fragment of Arist.’s Ai’ ginZt vn & polite
ia. 107 In 1928 it was 8,832: see Welter, Aig. 133. [Now 12,430, according to the 1995 edition of the Blue Guide.] 108 Welter, Aig. 30, accepted by Kirsten, ‘Review’, 299. In AA (1954) 29–30 Welter added the groundless speculation that even earlier Pheidon sold his warcaptives to Aegina in exchange for the silver which he coined; and he also sought to account for ‘der unauslo¨schliche Hass der gesamten griechischen Welt gegen die Aigineten’ as ‘gewo¨hnlicher Handelsneid’, similar to that incurred later by Venice, the Turks, the Knights of Malta and the French corsairs of the seventeenth century! He cites no evidence for this general hatred of the Aeginetans, and I see no reason to accept its existence.

410

Athenian Democratic Origins

in the mid-fifth century an outlook characteristic of the aristocracies of the archaic age, and a much greater number of politically unprivileged citizens, many of whom went in for trade—and, presumably, small-scale manufacture. Doubtless metics and other foreigners also played a part in the commerce of Aegina, as of other Greek cities; but perhaps in this case their share was smaller, owing to the number of citizens who were driven to devote themselves to trade. By the mid-fourth century, when Lampis the metic naukle¯ros was prominent (see section (iv) (b) above), the land of Aegina may perhaps not have been so extensively monopolised by a small class, and a higher proportion of the citizens may have become landowners; but the evidence of Aristotle (section (vi) above) shows that in the second half of the fourth century a significant proportion of them were still traders. I have already discussed [in the opening section of this chapter] Welter’s theory that the old ruling class financed mercantile activity on a large scale by bottomry loans. It cannot be disproved, but to me it seems very implausible. An unacceptable alternative theory has been produced by Kirsten (‘Review’, 299–301; cf. n. 5 above), which is devoid of evidence in its support and so entirely dependent upon ‘must-have-beens’ that it scarcely deserves discussion. Appealing both to the very inappropriate analogy of ‘die Ho¨rigen der Dorier, Heloten und Klaroten’, in Sparta and Crete (who were of course agricultural in character),109 and to that of freedmen in later times, he sees the greater part of the demos of Aegina as being in a dependent position, carrying on trade under the instructions and for the benefit of the aristocracy, who (he is obliged to admit) rarely went in person on trading voyages. Closer study of the not inconsiderable evidence for the mechanism of Greek trade might have convinced Kirsten that a situation so entirely different from that which we find elsewhere in the Greek world cannot be seriously considered without some positive evidence for it; and of this there is none. However, Kirsten seems to accept in many respects Winterscheidt’s picture of the Aeginetan ruling class, apart from the fanciful addition I have just outlined (on the strength of which he calls the nobility a Handelsaristokratie: see esp. the first few lines of p. 301 of his Review), and an emphasis 109 There would be all the difference between keeping dependent Greeks who were agricultural workers in a serf-like position and setting them up in trade.

Afterword

411

upon the ‘Dorian’ character of the aristocracy which some may find exaggerated. Seeing Aegina above all ‘recht als dorischen Staat . . . , als Vorposten des Doriertums’ [‘truly as a Dorian state . . . , as an outpost of Dorianism’], he speaks of ‘der unvera¨ndert dorische Charakter (vgl. den DvrieyB kvmoB & bei Pindar P. 8,20) der Lebensformen der adligen Herren’ [‘the unchanged Dorian character (cf. the DvrieyB kvmoB & of Pindar Pyth. 8,20) of the life-style of its aristocracy’], and he accepts Winterscheidt’s demonstration ‘der dorischen Lebensformen, der agrarischen Grundstruktur auch des aiginetischen Staates’ [‘of the Dorian way of life, and the basic agrarian structure of the Aeginetan state’], asserting at the same time an ‘Einfluss der Mischbevo¨lkerung des Demos auf den Geist des Staates’ [‘influence of the mixed population of the demos on the nature of the state’], which set Aegina apart from Sparta and Athens (‘Review’, 300–1). I am far less clear than Kirsten what political and economic conditions ‘Doriertum’ [‘Dorian character’] may imply—there is very considerable variation in these respects among Dorian states: Sparta, Argos, Corinth, Megara, Sicyon, Crete, Rhodes and the rest.110

AFTERWORD The view against which Ste. Croix brings his battering-ram in this essay is still firmly entrenched. The Princeton Encyclopaedia of Classical Sites tells us that ‘Aigina had a stable and developed mercantile aristocracy which spread the fame of its products throughout the Mediterranean basin’, adding ‘The spread of Aiginetan money shows clearly her absolute supremacy’ (19–20).a More recently, and much less wildly, Astrid Mo¨ller writes in her book on Naukratis (76): ‘The Aeginetan elite . . . differ from the elites of other Greek poleis, in that, unable to earn a substantial income as big landowners from agriculture . . . they had to rely on revenue from seafaring.’ There is a brief and lucid account of Aeginetan history, on conventional lines, by L. H. Jeffery in her AG 150–1, and again in the second edition of the CAH. 110 [The piece stops, rather than concludes, perhaps because Ste. Croix regarded it as ‘material to be used’, rather than as an independent, free-standing essay; see p. 371 above.] a We are further informed that ‘Solon passed special laws to limit the spread of Aeginetan commerce, thereby causing the island to ally itself first with Sparta, then with Thebes, and finally with Persia to oppose the rising Athenian power’.

412

Athenian Democratic Origins

The three books by T. J. Figueira, ASP, AA and EEH, constitute by far the most substantial recent contribution to Aeginetan studies. ASP, originally part of his doctoral thesis, ‘is not a history of Aegina, but rather a historical work that has Aegina as its subject’. Figueira does not simply assume that Aegina was governed by a commercial aristocracy (that view lurks unstated behind Morris’s chapter on ‘History and the role of Aegina’ in her BWS 92–103): he argues that it was so (ASP, chs. 4 & 5, esp. pp. 280–350). He had some conversations with Ste. Croix in Oxford in 1976, but did not see the present essay until after ASP had been published; the anticipations of Ste. Croix’s position in his book were thus based on those conversations. Besides general arguments from probability, Figueira stresses (322–30) the emphasis on ships and xenia in Pindar’s odes for Aeginetan victors. These arguments would not, we feel, have convinced Ste. Croix. We should not lose sight of the simple fact that if a man who lives on an island wants to meet a man from another state, still more if he is to establish ties of xenia with him (for what this involved see Herman, ch. 5, esp. 128–30), one or other of them has to travel on board a ship. Figueira’s AA is concerned with the period beginning in 457/6, i.e. after the archaic period with which Ste. Croix’s essay is concerned. His EEH very conveniently gathers together revised versions of no fewer than eleven articles previously published in various journals, and adds three new studies. Loomis’s review provides brief summaries of all of them: those most relevant to the present context are AASC, CCAA, and TMA. Conticello, B., article ‘Aigina’, in R. Stillwell (ed.) The Princeton Encyclopaedia of Classical Sites (1976), 19–21 Figueira, T. J., ASP ¼ Aegina: Society and Politics (1981, title-page) ¼ Aegina: Society and Economy (1981, p.iii) —— , AA ¼ Athens and Aegina in the Age of Imperial Colonization (1991) —— , AASC ¼ ‘Athenians, Aiginetans, and the Solonian Crisis’, ch. 3 (pp. 31–84) in his EEH (below) (not previously published) —— , CCAA ¼ ‘The Chronology of the Conflict between Athens and Aegina’, in QUCC 28 (1988) 49–90; revised as ch. 5 (pp. 113–49) of his EEH. —— , EEH ¼ Excursions in Epichoric History: Aeginetan Essays (1993) —— , FNAE ¼ ‘Four Notes on the Aeginetans in Exile’, in Athenaeum 66 (1988) 523–51, at 543–9, repr. as ch. 11 of his EEH 293–323, at 316–22 —— , TMA ¼ ‘Thoukydides, Melesias, and the Aeginetans’, ch. 8 (pp. 197–230) in his EEH (not previously published) Herman, G., Ritualised Friendship and the Greek City (1987) Jeffery, L. H. AG ¼ Archaic Greece (1976) —— , CAH ¼ ‘Aegina’, in The Cambridge Ancient History2 IV (1998), 364–7 Loomis, W. T., review of Figueira, EEH, in Bryn Mawr Class. Review 5.7 (1994) 575–8 Mo¨ller, A., Naukratis. Trade in Archaic Greece (2000) Morris, S. P., BWS ¼ The Black and White Style: Athens and Aegina in the Orientalising Period (1984)

Afterword

413

But in this essay Ste. Croix pronounces on a number of issues other than the nature of the ruling class of Aegina, and it will, we hope, be helpful to review recent scholarship on these too. Section (i). Aegina’s early fifth-century wars with Athens The chronological problems of the wars between Aegina and Athens remain unresolved. Hammond has reiterated his views in the second edition of CAH IV. Podlecki throws the second phase of the war (Hdt. 6. 87–93) back to c. 506, but has convinced few scholars—not even himself (see his pp. 402–3). Figueira, CCAA, after a thorough examination of the evidence, very reasonably concludes that Herodotus was unable to reconcile in his narrative the conflicting chronologies implied by the stories that he gathered from his various local sources (Sparta, Aegina, Athens)—if indeed he was aware of them; he (Figueira) prefers to place the second phase of hostilities after Marathon. Figueira, CCAA: as above Hammond, N. G. L., in The Cambridge Ancient History2 IV (1988), 501–2, 784 Podlecki, A. J., ‘Athens and Aegina, 510–480 B.C.’, in Historia 25 (1975) 396–413

Section (iv). Sostratos Since Ste. Croix wrote, an inscribed anchor dedicated to Aeginetan Apollo has been discovered at Gravisca, the port of the Etruscan city Tarquinia, bearing the name of Sostratos and dated (on the basis of its letter-forms) to the late sixth or early fifth century. This is either the Herodotean Sostratos (in which case Welter’s seventh-century date (n. 38) must clearly be rejected), or a later member of the same family. Boardman, J., The Greeks Overseas4 (1999), 206, 295 Figueira, ASP (as above), 241–50 (notes at 290–1) Gianfrotta, P. A., ‘Le ancore votive di Sostrato di Egina e di Faillo di Crotone’, in La Parola del Passato 30 (1975) 311–18 (overlooked by Harvey) Harvey, F. D., ‘Sostratos of Aegina’, ibid. 31 (1976) 206–14 Johnston, A. W., ‘The Rehabilitation of Sostratos’, ibid. 27 (1972) 416–23 —— , ‘Aeginetans Abroad’, in Horos 7 (1989) 131–5, at 133–5 Roebuck, C., in The Cambridge Ancient History2 IV (1988), 456–8 (over-confident identification?) Torelli, M., ‘Il santuario de Hera a Gravisca’, in La Parola del Passato 26 (1971) 44–67, at 55–66

414

Athenian Democratic Origins Section (v). Population statistics

Opinions remain sharply divided on the size of the population of Aegina. The innocent enquirer is informed by the Oxford Classical Dictionary3 (Hornblower) that it was c. 40,000 and by Der Kleine Pauly (Kaletsch) that it was a maximum of 13,000 to 20,000; neither indicates that other views are held. Figueira, ASP 37–8, in the course of a long and careful discussion (29–52, with notes at 56–64) puts it between 35,000 and 45,000 or c. 42,000,b and his lowest figure (35,000) is accepted by Horden & Purcell: ‘By conventional calculations its [Aegina’s] own resources can support at minimal nutritional levels only some 5,000 people. Incidental evidence from around 500 B.C. makes it certain that the population was then at least seven times as large, and by no means existing precariously. The conclusion that Aegina was heavily dependent on a complex, reliable and large-scale trade in staples seems inescapable’ (119; our italics). But it is difficult to accept that the population of the island in the late archaic period was five times greater than it was in the 1960s. Beloch had an uncanny knack of being right on such matters; whether the ships of the Aeginetan navy were triremes or pentekonters (Ste. Croix n. 47; Figueira, ASP 30), there is no need to assume that an Aeginetan ship of the early fifth century required as many rowers as the classical Athenian trireme; and Figueira’s arguments against the use of hired rowers (33–5) and slaves (35–7), though strong, are not conclusive. It seems wisest to remain agnostic, as Ste. Croix does at the end of section (xiii). Figueira, ASP: as above Horden, P., and Purcell, N., The Corrupting Sea: a study in Mediterranean history (2000)

Section (vi). Coinage: chronology and purpose Ste. Croix’s view on the dates of the earliest Greek coinage (‘Corinth and Athens, whose coinages are now believed to have begun only c. 575 at the very earliest, that of Athens in particular perhaps distinctly later’) is now generally accepted. Further precision, at least for the Aeginetan coins, is unattainable (Price, in CAH2 IV plates 238; cf. the generous agnosticism of Howgego, 4–6). Kroll & Waggoner give clear and full arguments for putting the first issues of the earliest Athenian coins, the didrachms with heraldic designs known as Wappenmu¨nzen, in the midsixth century (c. 550); this now commands general assent. What is b 70 ships ¼ 14,000 rowers;  3 for women, children and males over military age ¼ 42,000.

Afterword

415

uncertain is how much earlier the coinage of Aegina begins. Kroll & Waggoner, like Holloway, DFGC and AHCAC, put it as early as c. 580/ 570. Price & Waggoner, AGC, however, argue for a date a generation later, c. 550, which is adopted also in the handbooks by Carradice & Price and by Jenkins, AGC2 (revised downwards from the first edition). Kraay repeatedly adopted a middle view (AHSCC, ACGC 41–3, CAH IV2 437–8), putting it c. 560 (or 570–50). Clearly, fine tuning in this world of proto-tortoises is hardly possible: see the careful discussion in Figueira, ASP 88–107, 155–8. Kagan’s attempt in PAC to defend the ancient tradition that Pheidon of Corinth (whose own chronology is wobbly) introduced coinage into Greece at a much earlier date has found little favour. Similarly, Kagan’s second attempt (in DEC) is dismissed by Figueira, EEH 63 n. 5 as ‘not much help’. For attempts to make sense of this tradition, see (preferably) Kraay, ACGC 313–14 and in CAH IV2 432–3; Figueira, ASP 65–80.c Kraay’s argument that coinage was at first issued only in large denominations, and that it could not have been used for everyday purchases (first argued in HSCOC, quoted at some length here by Ste. Croix, and reiterated by Kraay in ACGC 317–28 and CAH IV2 441–5) has been standard doctrine for more than a generation now: see Rutter, Price, FEGBC, TBC and in CAH IV2. Recent work, however, especially by H. Kim (GFSC, CMSC, ACEUM) has made it clear that much more fractional coinage was issued in the sixth century than had previously been realised (Be´rend; Carradice & Price, 27, Howgego, WASSC esp. 3, 22 and AHC 6–7; cf. tentatively Jenkins, AGC1 28 ¼2 14), which suggests that this ‘standard doctrine’ should be modified if not abandoned; if Kim is right, early coins might after all have been intended for use in everyday retail transactions. This is hardly the place to discuss the wide-ranging speculations of Seaford (220–32 and elsewhere) and von Reden, EAG (esp.177–8, 184–7) and MLE (esp.156–7). Be´rend, D., ‘Re´flexions sur les fractions grecques’, in A. Houghton et al. (ed.) Festschrift Leo Mildenberg (1984) 7–30, with Pls. 1–2 Carradice, I., and Price, M. J., Coinage in the Greek World (1988), 27, 35–8, with plates 2–3; bibliography at 137–8 Holloway, R. R., AHCAC ¼ ‘An Archaic Hoard from Crete and the Early Aeginetan Coinage’, in Amer. Numism. Soc. Museum Notes 17 (1971) 1–21, with Pls. I–VIII —— , DFGC ¼ ‘The Date of the First Greek Coins’, in Revue Belge de Numismatique 130 (1984) 5–18 c Further bibliography: Kroll & Waggoner, 325; Price, in Pls. to CAH2 IV 238; Figueira EEH 63 n.5.

416

Athenian Democratic Origins

Howgego, C. J., WASSC ¼ ‘Why Did Ancient States Strike Coins?’, in Numism. Chron. 150 (1990)1–25 —— , AHC ¼ Ancient History from Coins (1995) 4–9, with figs.16–20 Jenkins, G. K., AGC1 ¼ Ancient Greek Coins1 (1972) 27–51; Jenkins, AGC2 ¼ revised edn. (1990) 13–26 Kagan, D., PAC ¼ ‘Pheidon’s Aeginetan Coinage’, in TAPA 91 (1960) 121–36 —— , DEC ¼ ‘The Dates of the Earliest Coins’ in AJA 86 (1982) 343–60 Kim, H., GFSC ¼ Greek Fractional Silver Coinage, unpublished M.Phil. thesis, (Oxford, 1994) (non vidi) —— , CMSC ¼ ‘Coinage, Money and Small Change in Archaic Greece’, lecture delivered at the conference on money in the ancient world (Univ. of Exeter, July 1999; unpublished, but full summary circulated) —— , ACEUM ¼ ‘Archaic Coinage as Evidence for the Use of Money’ in Money & its Uses in the Ancient Greek World ed. A. Meadows & K. Shipton (2001), 7–21 with Pl. I.1 Kraay, C. M., AHSCC ¼ ‘The Asyut Hoard: Some Comments on Chronology’ in Numism. Chron. 137 (n.s.17) (1977) 189–99 —— in CAH IV2 ¼ ‘Coinage’, ch. 7d in The Cambridge Ancient History2 vol. IV (1988), 431–45 (bibliography at 861–3) Kroll, J. H., and Waggoner, N. M., ‘Dating the Earliest Coins of Athens, Corinth and Aegina’, in AJA 88 (1984) 325–40 Price, M. J., FEGBC ¼ ‘The Function of Early Greek Bronze [sic] Coinage’, in supplement to Annali dell’Istituto Italiano di Numismatica 25 (1979) 351–65 —— , TBC ¼ ‘Thoughts on the Beginning of Coinage’ in C.N.L. Brooke et al. (eds.) Studies in numismatic method presented to Philip Grierson (1983), 1–10 —— in plates to CAH2 ¼ ‘Coinage,’ ch. 15 in The Cambridge Ancient History2 (1988), plates to vol. IV 237–43 —— and Waggoner, N. M., Archaic Greek Coinage: the Asyut Hoard (1975) 61–76, 122–3, 131–2 with Pls. XIX–XXI Rutter, N. K., ‘Early Greek Coinage and the Influence of the Athenian State’, in B. Cunliffe (ed.), Coinage and Society in Britain and Gaul [sic]: some current problems (1981), 1–9 Seaford, R. A. S., Reciprocity and Ritual (1994) von Reden, S., EAG ¼ Exchange in Ancient Greece (1995) —— , MLE ¼ ‘Money, Law and Exchange: Coinage in the Greek Polis’, in JHS 117 (1997) 154–76

Section (ix). Naucratis The disproportionate length of this section is primarily due to the number and complexity of the issues involved, but the fact that Ste. Croix did his wartime service in Egypt, which gave him a renewed interest in the ancient world (Parker, 452–3), may not perhaps be altogether irrelevant. The best general account of the archaeology and history of Naucratis is still that of Boardman in GO, originally published as a Pelican book in 1964. His picture of the nature of the community is close to that of Ste.

Afterword

417

Croix, who was writing at about the same time; Ste. Croix might have objected to Boardman’s judgement that ‘Naucratis attracted the getrich-quick merchants of East Greece, and their Aeginetan colleagues who ran the business with central Greece’, but would have been pleased that he does not identify these Aeginetans with members of the ruling aristocracy. Boardman (GO 130) emphasises, as indeed do most scholars who have written about Naucratis in recent years (Austin, 22, 27–9, 44–5; Braun, 40–1; Figueira, ASP 253, 262; Bresson, RHSN 293 [16–17], 297 [22–3]; Bowden, 29; Mo¨ller, 204, 207–8 etc), that Naucratis was dependent on the favour of the Pharaoh, a point that Ste. Croix does not mention but would not have denied. Indeed, it strengthens his case, since it could hardly have constituted an autonomous polis if it was under Pharaonic control (Bowden, 29; contra, Hansen, 93). Boardman differs from Ste. Croix, however, in maintaining that the prostatai tou emporiou were elected locally at Naucratis, not by their home cities (130–1). Austin, too, in his excellent monograph rejects Herodotus’ testimony on this question: ‘How could a whole series of Greek states . . . , often quarrelling with one another and . . . incapable of any coherent plan of action, actually agree year after year on the appointment of officials . . . to take charge of affairs in a distant port in Egypt?’, he asks (31–2), and many others (e.g. Figueira, ASP 261–3; Mo¨ller, 195) take a similar line. Some, however, have found parallels for such an arrangement, especially among amphictionies (Bowden, 32–3; Bresson, RHSN 311–12 [42–4], id. RN 74). But Herodotus says nothing about meetings, joint action or annual magistracies: these difficulties are entirely illusory, and there is no need either to reject the evidence or to search for justificatory parallels if we understand the passage as Ste. Croix does. (Braun, 42, also accepts Herodotus without further ado.) Austin agrees with Ste. Croix that the Greek settlement at Naucratis may well have been established—a less misleading word than ‘founded’—informally by traders rather than formally by states (Austin, 23, 59; so too Bresson, 315–16 [¼ 51–2]; Mo¨ller, 183; but denied with reference to the Hellenion by Bowden, 32). But unlike Ste. Croix, he rejects the possibility that there might be some truth in Strabo’s account, according to which the settlement was first established by Miletus only, on the grounds that it is incompatible with Herodotus (Austin, 23, 59; also Bresson, RHSN 315–16 [¼ 51–2]; Drijvers; Figueira seems in two minds [‘obviously erroneous’, ASP 253;‘should not be dismissed out of hand’, 254, cf. 257]; accepted by Braun 37–8.)d Boardman’s pottery chronology (now GO 121), followed here by Ste. Croix (n.97), is still d Who (unlike the Loeb translator) does not imagine that Inaros was a city, pace Mo¨ller 186 n. 34.

418

Athenian Democratic Origins

generally accepted (Austin 23; Figueira ASP 255–6; Mo¨ller 91; but note the dissent in Bowden 26–8). Austin also stresses (30, pace Mo¨ller, 197 n.110) Herodotus’ distinction betwen the residents of Naucratis and the visiting traders (a point that seems to have escaped Ste. Croix: see Bresson, RHSN passim, esp. 291–9 ¼ 13–26; Bowden, 28–9; contra, Mo¨ller, 188–9, 197–8, 214, who argues for a double aspect, not a physical division). The former eventually established a polis. When this happened remains contentious. Austin gives no date for its establishment (31), but seems to imply one considerably earlier than Ste. Croix would have countenanced. Hansen, 92–3, argues that it was already a polis (in the sense of a city-state) in the fifth century, on the grounds that Herodotus says so, and that his usage is consistent; this is rightly disputed by Bresson, RN 74–9. Bresson himself (RHSN, 316–17 [52–5], etc.; followed by Bowden, 29–30) at first dated the change in status to some time after the arrival of Alexander: he is still inclined to that opinion, but would now prefer to leave the question open (RN 78–9; further discussion forthcoming). Mo¨ller, 184–91, is in accord with Ste. Croix’s fourth-century date. The two Rhodian proxeny decrees referred to by Ste. Croix (pp. 402–3) form the chief focus of Bresson’s long and complex RHSN (main conclusions summarised in RN 74); in SIG3 110 he restores Ai’ g[ypt
: ion rather than Ai’ g[in
: atan: the strongest argument against this is that Pytheas, the proxenos’ father, bears a name frequently found on Aegina (Figueira, FNAE 321; cf. Hornblower’s paper cited in n. 28). Bresson, RN 66–73, also presents and discusses interesting new evidence from the Egyptian side. Bowden, after giving an account of the 19th-c. excavations, suggests that Hogarth’s identification of the Hellenion, generally accepted (for reasons outlined by Lloyd, 224), was incorrect (21–4), and that various chronological problems might be resolved if the chronology of archaic pottery were revised, bringing it down some forty years later (24–8). There has been no rush to follow this proposal. He maintains that Naucratis was not established informally by traders, but by Greek poleis and an Egyptian pharaoh or pharaohs, who aimed to control and profit from its trade. Braun provides a vivid and lively account of the society of Naucratis and the goods exchanged there, but says little about the topics that interested Ste. Croix. The latest general account of Naucratis is the book by Astrid Mo¨ller. Pride of place (and much space) is given to considering whether Naucratis qualifies as a ‘port of trade’ as defined by Polanyi; there is also a (necessarily) selective catalogue of finds. But there is no sign of Herodotus 2.178–9, the text from which any historical discussion must

Afterword

419

start, before p. 183, and even then it appears in translation, not in Greek. As Osborne says in his somewhat adverse review, the book is more valuable on archaeology (including individual finds) than on economic history. It is no secret that the original excavations of Naucratis, together with the recording of the finds, left much to be desired, even by the standards of the time (Boardman, GO 118–21, with photos; Bowden, 19–22; Mo¨ller, 92–4, 111–13, 116–19 etc.). Hogarth in 1899 was not even able to find the Hellenion that Petrie claimed to have discovered in 1884 (Coulson & Leonard, INE 364–6; Mo¨ller, 111–12). A lake, caused by the rising water-table, now covers the archaic settlement, and the whole area has been disturbed by local farmers digging for fertiliser (sebakh). It is hardly surprising that more recent investigations have thrown more light on Ptolemaic and Roman Naukratis than on the archaic period (the non-specialist will find Coulson & Leonard, INE, or Coulson, NS, quite adequate; fuller accounts in their CDN, AN II and in Leonard & Berlin, 28–30). But, as Charles Reed points out elsewhere in this volume, even if archaeologists were to excavate in ideal conditions until they were blue in the face, they would not come up with answers to the kinds of question that interested Ste. Croix. Austin, M. M., Greece and Egypt in the Archaic Age (Proc. Camb. Philol. Soc. suppl. 2 [1970]), esp. 22–33, 37–45 (with notes at 58–75) Boardman EGVP : as above —— GO ¼ The Greeks Overseas4 (1999), 117–33 (notes at 290–1) Bowden, H., ‘The Greek Settlement and Sanctuaries at Naukratis: Herodotus and Archaeology’ in M. H. Hansen and K. Raaflaub (eds.), More Studies in the Greek Polis (Historia Einzelschrift 108 [1996]), 17–37 Braun, T. F. R. G., ‘Naucratis’, in The Cambridge Ancient History2 III 3 (1982), 37–43 Bresson, A., RHSN ¼ ‘Rhodes, l’Helle´nion et les statut de Naukratis (VIe–IVe sie`cle a.C.)’ in Dialogues d’HistAnc 6 (1980) 291–349, repr. with slight revisions as ch. 1 (pp. 13–63) of his La cite´ marchande (2000) —— , RN ¼ ‘Retour a` Naucratis’, ch. 2 (pp. 65–84) of his La cite´ marchande Coulson, W. D. E., NS ¼ ‘The Naukratis Survey’ in E. C. M. van den Brink (ed.), The Archaeology of the Nile Delta: problems and priorities (1988), 259–63 —— , AN II ¼ Coulson, et al., Ancient Naukratis ii: The Survey at Naukratis and Environs pt. I: The Survey at Naukratis 1997 Coulson, W. D. E., and Leonard, A., jr., CDN ¼ Cities of the Delta i: Naukratis (1981) —— , INE ¼ ‘Investigations at Naukratis and Environs’ in AJA 86 (1982) 361–80 at 361–75 (the rest of the article reports on the nearby site of Kom Firin) Drijvers, J. W., ‘Strabo 17.1.18 (801c), Inaros, the Milesians and Naucratis’, in Mnem. 52 (1999) 16–22 Figueira ASP: as above

420

Athenian Democratic Origins

Figueira FNAE: as above Hansen, M. H., ‘Emporion: A Study of the Use and Meaning of the Term in the Archaic and Classical Periods’ in T. H. Nielsen (ed.), Yet More Studies in the Ancient Greek Polis (Historia Einzelschrift 117 [1997]), 83–105 at 91–4 Leonard, A., and Berlin, A., Ancient Naukratis: excavations at a Greek emporion in Egypt II ¼ AASOR 54 (1997)—the entire volume. Lloyd, A. B., Herodotus Book II: Commentary 99–182 (1988) 221–31 (biblio. at 231) Mo¨ller: as above Osborne, R., review of Mo¨ller, in CR 52 (2002) 96–7

11 Herodotus and King Cleomenes I of Sparta

Ste. Croix wrote in 1972 that he knew of no satisfactory discussion of Spartan foreign policy in the reign of Cleomenes (OPW 167 n. 1). In this lecture he sets out to fill the gap himself. It was first given at a Sixth Form Conference at Radley College, Oxfordshire, on 5 October 1976, and subsequently at Sheffield University (19 October 1977) and at a Sixth Form Conference at Bristol (8 March 1979). We have not modified the informal but vigorous style: this is not an article, but a script prepared for oral delivery. References, originally listed on a handout, and a number of substantial manuscript addenda have been incorporated into the text or offloaded into footnotes.

Since I believe that Herodotus’ picture of Cleomenes, overall, is gravely inadequate and in parts misleading, I think I had better begin by making it clear that I greatly admire Herodotus and agree with Collingwood’s remark about him, that he was ‘one of the great innovating geniuses of the fifth century’. He was certainly ‘the Father of History’—although when I find my colleagues being uncritically reverent towards Herodotus (as I think many of them are), I like to say, ‘Yes, Herodotus was the father of history—in the sense that history began in the next generation, with Thucydides.’ But I don’t mean that seriously. My own favourite remark about Herodotus is Gibbon’s, that he ‘sometimes writes for children and sometimes for philosophers’. Herodotus is quite frank about his historiographic method. In VII 152.3 (cf. II 123.1; IV 195.2) he says that he conceives it to be his duty to l
egein t a leg
omena, to repeat what his informants told him; and he goes on, ‘But that doesn’t mean that I necessarily believe it’; and he adds, ‘And let that go for the whole of my History.’ Well, of course it’s very nice to know from Herodotus about all sorts of things that were believed in his day which weren’t

422

Athenian Democratic Origins

true; but I must admit that I do sometimes wish Herodotus had taken more trouble to find out how much of the information he received was true. Herodotus often did that, but by no means always. This explains, as I believe, his unnecessarily unfavourable account of Cleomenes: he relied on the Spartan tradition, which was evidently very hostile. Herodotus has two passages and two only which are actually favourable to Cleomenes. In III 148.2 he praises Cleomenes for refusing to be bribed by Maeandrius the Samian (this will be in 517 or 516 BC); and in VI 61.1 he says that Cleomenes in Aegina in 491 was ‘working for the common good of Hellas’ in his attempt to obtain hostages as security against Aeginetan medising—a subject I shall spend some time on later. Elsewhere, Herodotus is often very hostile. He even says that Cleomenes ‘reigned for no long time’ (V 48), whereas we know for certain that the reign lasted for some thirty years. I propose to give an outline narrative of the main events of the reign of Cleomenes first, and to leave explanations mainly to a later stage. *** For reasons I needn’t go into, we can be reasonably sure that Cleomenes was born somewhere around 540—between, say, 545 and 540 or perhaps 535. His mother, it has been suggested, was a member of the family of Chilon, the great ephor (perhaps of 556/5): the relevant passages in Herodotus are V 41.3 with VI 65.2. Cleomenes probably came to the throne about 520, and the very first action of his that we know of is in 519/18: that date rests on the authority of the manuscripts of Thucydides, which do sometimes make mistakes over numerals; but I see no good reason for rejecting the date I’ve mentioned (Hdts VI 108 with Thuc. III 68). Thebes, at that time, was trying to coerce little Plataea, seven or eight miles to the south of her, into a Boeotian league through which Thebes hoped to gain a dominant position in Boeotia. According to Herodotus, Cleomenes and the Spartans happened to be present at Plataea, presumably on some unrecorded expedition. The Plataeans ‘tried to give themselves to Cleomenes and the Spartans’; but Cleomenes said ‘No, you go to Athens.’ (Actually, Herodotus speaks of ‘the Spartans’ as giving this advice, but I would have no hesitation in ascribing the reply to Cleomenes himself.) The Plataeans did go to Athens, and with Athenian

Herodotus and King Cleomenes I of Sparta

423

help they remained independent of Thebes, and indeed Athens defeated Thebes in a battle and obtained a frontier settlement (with Corinth as arbiter) very favourable to Plataea. Herodotus’ comment is interesting: almost everyone just repeats it, but it obviously won’t do: he says that the purpose of the advice was to create trouble for Athens by making her the enemy of Thebes (VI 108.3). But Athens didn’t have to accept the Plataean alliance: she was now to be given the choice whether to help Plataea or to do Thebes a good turn by remaining neutral. The people who were bound to suffer if Athens did help Plataea were the Thebans; and it was Thebes rather than Athens for which Cleomenes was creating trouble. This incident was indeed the beginning of a long-lasting enmity between Athens and Thebes, such as existed only too often between Greek states which were neighbours. And we must not fail to notice what a clear benefit this was for Sparta: by encouraging Plataea to ally with Athens she set the two greatest states north of the Isthmus of Corinth at one another’s throats. It was a master-stroke of policy. I shall move on now to 510, passing over some minor incidents in the career of Cleomenes which I shall mention later. In 510 Cleomenes and a Spartan army marched to Athens and threw out the tyrant Hippias, son of Peisistratus (Hdts V 64–5; Arist., Ath. Pol. 19 etc.)—a fact which the Athenians naturally preferred to forget: they liked to pretend that it was their own heroes Harmodius and Aristogeiton who overthrew the tyrant, although in fact all they did was to assassinate the tyrant’s brother Hipparchus and get themselves executed. Some of you will know the delightful Athenian drinking-songs in praise of Harmodius and Aristogeiton (PMG 893, 895): these skolia say that they killed the tyrant (which of course they didn’t) and that they made Athens isonomos (equal before the law), which of course again they couldn’t: isonomia at Athens was the creation of Cleisthenes & Co. in 508/7, a couple of years after the expulsion of Hippias. In 508/7 Cleomenes tried to stop the new democratic constitution of Cleisthenes from going through: at the invitation of the Athenian aristocratic leader Isagoras he came to Athens, presumably with just a small bodyguard, expecting that his great personal prestige as king of Sparta and liberator of Athens from the tyrant would overawe any opposition. It didn’t: there was a great rising of

424

Athenian Democratic Origins

the Athenians against him; and he was driven out, and the constitution of Cleisthenes went through.1 Round about 506, Cleomenes organised a great expedition against Athens, consisting of the Spartans under both their kings (Cleomenes himself and Demaratus), their Peloponnesian allies, and also the Boeotians (Thebes and her allies) and the Chalcidians from Euboea. This turned out to be an appalling fiasco, the one and only Spartan disaster in the second half of the sixth century, and it must have dealt a serious blow to the reputation of Sparta in general and Cleomenes in particular. The expedition began to break up when the Corinthians deserted; then Demaratus quarrelled with Cleomenes and went off home, the other Peloponnesians followed suit, and this left the Chalcidians and Boeotians isolated against Athens, which inflicted crushing defeats on them both (Hdts V 74–7). At some time in the last four years or so of the sixth century (I don’t think we can date it more closely than that) the Spartans decided to restore the tyrant Hippias to Athens; but they realised that they couldn’t do this without their allies, and when the allies absolutely refused to consider it, the Spartans abandoned the idea (Hdts V 90–3). For some reason I’ve never been able to understand, many modern historians have stated it as a fact that this was Cleomenes’ idea. Well, of course I can’t prove it wasn’t his; but I would emphasise that Herodotus, who throughout Cleomenes’ reign almost always speaks of Spartan foreign policy as if it were decided by Cleomenes, on this one occasion doesn’t mention Cleomenes at all; and I would suggest that the probabilities are strongly against his having had a hand in the proposal to recall Hippias. To reinstate a hated tyrant in another city was something that most Greeks would have strongly disapproved of, and I don’t see how Cleomenes could have been left out of Herodotus’ story if the policy really was his, since the Spartan tradition that Herodotus received was in general so hostile to Cleomenes, and it would have been very convenient to lay the blame on him. I suggest that after the great fiasco of c. 506, the other Spartans very naturally said to themselves, ‘Cleomenes’ policy of throwing out our old friend Hippias has been a disaster for us, so let’s go back to where Cleomenes came in, and reinstate Hippias.’ 1

For the whole story, see Hdts V 70, 72; Arist., Ath. Pol. 20.2–3, etc.

Herodotus and King Cleomenes I of Sparta

425

The next event that we hear of at Sparta is in 499, when there came to Sparta Aristagoras the tyrant of Miletus, who was preparing to organise a great revolt of the Ionian Greeks of Asia Minor and the islands, against the Persians, who had conquered them in the previous generation. He is represented by Herodotus as going in the first place to Cleomenes; and I think we can take it from this that Cleomenes had by now recovered his prestige as the most powerful figure at Sparta—Aristagoras obviously went to him because if you had Cleomenes on your side you could expect to be able to swing the Spartan Assembly. (By the way, the Spartan Assembly was certainly not called ‘the Apella’, as practically all the books say; it was almost certainly called ‘the ekklesia’, as in most other Greek cities [OPW 346–7].) I have time to mention only two features of Herodotus’ story about Aristagoras’ visit to Sparta (Hdts V 49–51). First, Cleomenes’ daughter, Gorgo, who on this occasion plays a very admirable role. Herodotus recounts two incidents, in each of which dear little Gorgo (who, he says, was only eight or nine at the time of Aristagoras’ visit) plays an important part: in each case she is wise far beyond her years (V 51, VII 239). And on this occasion she may have been the only other person present with daddy and Aristagoras, and therefore the probable source of the story. Gorgo later on married King Leonidas of Sparta, the hero of Thermopylae, and her son Pleistarchus ruled as king at Sparta for over twenty years. I like to think that when the young Pleistarchus said to Gorgo, ‘Oh Mummy, do tell me about some of the things you did when you were a little girl’, Gorgo brought out these two stories. The other thing I want to mention is the motive Herodotus attributes to Cleomenes for refusing to help Aristagoras. Cleomenes was not hostile from the first to the idea of helping the Ionians revolt. He originally listened to Aristagoras and said he wanted two days to think it over. It was only when Aristagoras seemed to be expecting the Spartans to march with him right into the interior of Asia, as far as Susa, and revealed that Susa was a three-months’ journey from the sea, that Cleomenes turned him down, and indeed told him to clear out of Sparta before sunset. Well, the Ionian Revolt took place, but after five years, in 494, it ended disastrously, with a naval defeat at Lade and the capture and sack of Miletus.

426

Athenian Democratic Origins

The next thing we hear of Cleomenes is connected with the disaster to Miletus: it is the famous battle of Sepeia, a Spartan victory over Argos, which was remarkable in involving the greatest slaughter of hoplites known to me in any war between two Greek states: according to Herodotus, six thousand Argives were killed, partly in the actual battle and partly afterwards.2 Most people nowadays date this battle in 494, because of the so-called ‘double oracle’ (Hdts VI 18–19, esp. 19.2, with 77.2) given by Delphi to the Argives ‘and to the Milesians although they were not present’—a unique document. As with so many recorded Delphic oracles, its genuineness is disputed. Since it evidently prophesies something nasty to Argos, and in a side-swipe at Miletus in its second half seems to say that Miletus is going to be done down by the Persians, it is easy to conclude that the oracle was manufactured after 494 to show Delphi’s prescience in associating the disaster to Miletus, which certainly took place in 494, with the disaster to Argos at Sepeia, which must then have been more or less contemporaneous. Hence the date commonly given to Sepeia, of 494. But . . . what if the oracle is genuine?—as I think it is. Unless we are prepared to attribute divine inspiration to Delphi, how would the oracle know in advance that the fall of Miletus and the battle of Sepeia would take place at about the same time? Well, I accept the view (originally suggested a long time ago by Bury)3 that the oracle was delivered in 499. Aristagoras, who after Sparta went to Athens, and therefore will have passed close to Argos, asked Argos also for help; and Argos, not wanting to get involved, consulted the Delphic Oracle (which at this time was medising thoroughly), because she knew the Oracle would say ‘Keep out’, as in effect it did. The reply to Argos was just generally nasty (Delphi seldom cared to be too specific in its prophecies), and there was a sidekick at Miletus, saying almost in so many words, ‘The Milesians haven’t a hope against Persia.’ All that this gives us for Sepeia is a date after 499; but for other reasons I’d be quite happy to say ‘about 494’. I don’t want to give any details about the Sepeia campaign just yet, but I do want to say that it shows very remarkable ability on 2

See Herodotus VI 75.3–82 with VII 148.2 for the whole story. [J. B. Bury, ‘The epicene oracle concerning Argos and Miletus’, Klio 2 (1902), 14–25.] 3

Herodotus and King Cleomenes I of Sparta

427

Cleomenes’ part, which people too often miss, as well as extreme ruthlessness. The cleverness with which the campaign was conducted does not fully emerge from Herodotus’ account of the actual events; but with the aid of other evidence, mostly supplied by Herodotus himself, we can begin (as we shall see presently) to appreciate Cleomenes’ achievement. I should perhaps add that I am saying nothing about the Telesilla story, given by Pausanias (II 20.8–9), because I believe it to be a fiction. I would now like to give a brief account of the last year or so of Cleomenes’ life, a story which Herodotus presents very well in part, before (as I believe) he loses his grip, in his reliance on what the Spartans told him. Well before the end of the sixth century the Persian King Darius had already shown his aggressive intentions in Europe, by conquering the whole North Aegean seaboard as far as Macedon. And not long after the collapse of the Ionian Revolt, in 492 or 491, he sent ambassadors to Greece to demand from the cities earth and water, the symbols of submission to Persia by land and sea. It was obvious that he would punish Athens and Eretria, the two cities which had sent help to the Ionians in their revolt against Persia a few years earlier. A few Greek cities, including Athens and Sparta, rejected the demand for earth and water, but many submitted, including the important island of Aegina, immediately opposite Athens’ port of Peiraeus and at this time a bitter enemy of Athens.4 (Later, Pericles, we are told, called Aegina ‘the eyesore of the Peiraeus’!) If a Persian fleet arrived and was able to use Aegina as a base, Athens’ chances of survival would be greatly reduced. So Athens appealed to Sparta to coerce Aegina, which was a member of the Peloponnesian League of which Sparta was hegemon and as such was obliged (as I believe) to follow Sparta in her foreign policy (OPW 333–5, section B). The events which followed were very interesting indeed. Herodotus gives a characteristically delightful narrative.5 Cleomenes went off on his own to Aegina and demanded hostages (who of course might be killed if Aegina persisted in her medising); but the Aeginetans refused to obey, under the leadership of one of the principal medisers, Crius son of Polycritus (both, by the way, famous athletes, like so many of the Aeginetan governing class). Crius admitted that if both the Spartan kings 4

[See ch. 10 of this volume.]

5

VI 48–51.1; 61.1; 64–67.1; 73.

428

Athenian Democratic Origins

had come, Aegina would have had to give the hostages, because the presence of both the kings would have been proof that they were acting in pursuance of an official decision by the Spartans; but as it was (said Crius), Cleomenes was acting on his own, because he had been bribed by the Athenians. And Herodotus adds that Crius acted in this way on the advice of Demaratus, the other Spartan king, who was an enemy of Cleomenes: he had been so ever since the great fiasco of c. 506, when the expedition against Athens had broken up (you will remember) partly because of a quarrel between the two kings. Demaratus was now telling the Aeginetans not to give in to Cleomenes. Whether their pretext for refusing was based on a genuine constitutional point, or whether it was just a bluff, is too complicated a question to go into now (see OPW 149–51). Anyway, Cleomenes returned to Sparta empty-handed. And he then did something very extraordinary: by working the Delphic Oracle, through a leading man there, Cobon, who got at the Pythia, Cleomenes managed to get Demaratus deposed from his kingship, on the ground that he was really a bastard. (Demaratus later fled to the Persian sphere, where he was royally treated, and given a vast estate in the Troad.)6 Demaratus was replaced as king by Leotychidas, who was evidently very much under Cleomenes’ influence. Back went Cleomenes to Aegina, this time with his new co-king, and the Aeginetans now dared not refuse to hand over the hostages—who were ten of the leading men, including the bold Crius. Now even Herodotus, speaking of Cleomenes’ first visit to Aegina, says he was ‘working for the common good of Hellas’ (p. 422 above). I think Cleomenes’ action in preventing Aegina from medising may have been very important in saving Athens in the next year, 490, when a Persian fleet came to Greece, sacked Eretria, and landed in Attica, but sailed away again after the battle of Marathon. As I have said, I doubt if Athens could have survived, had the Persians been able to use Aegina as a base. Here I would also like to emphasise how remarkably selfless Cleomenes’ action over Aegina was. He was coercing a Dorian aristocracy which is never recorded to have done anything detrimental to Sparta in general or to Cleomenes in particular, for the benefit (in the first place) of a state (Athens) which had been recalcitrant to Sparta and had twice brought Cleomenes himself 6

Hdts VI 70:2; cf. Xen., Hell. III i 6; Anab. III i 3; VII viii 17.

Herodotus and King Cleomenes I of Sparta

429

to disaster, in 508/7 (the Cleisthenes affair) and in c. 506 (the big expedition that broke up). We now come to the final episode in the career of Cleomenes (Hdts VI 74–5). His machinations at Delphi were discovered, and he had to flee from Sparta. He went first to Thessaly (more of this later), and from there to Arcadia, where, according to Herodotus, he tried to put into effect some revolutionary designs (ne
vtera e’
prZsse pr
Zgmata), banding together ‘the Arcadians’ against ’ i tfi Z& Sp
Sparta (synist aB toyB ’Ark
adaB ep artfi Z). He wanted the Arcadians to swear great oaths to him by the River Styx that they would ‘follow him whithersoever he should lead them’. (It isn’t quite clear to me from Herodotus’ words whether the oaths were actually sworn or not.) The Spartans were so alarmed by this that they invited Cleomenes to return as king again. Cleomenes, according to Herodotus, had always been ‘a bit cracked’ ‘ (ypomarg
oteron), and he now went raving mad and very conveniently committed suicide by gradually cutting himself into little strips, from his legs to his belly. Herodotus wasn’t surprised at this, and, characteristically, he gives various explanations offered by the Greeks in general, the Athenians, the Argives, and (rather later) the Spartans (VI 75.3; 84.1–3). The Spartan theory is too delightful to omit: according to Herodotus, they said that an embassy once came to Sparta from the Scythians (this, if genuine, would be 514 or 513), to propose joint action against Persia, and that the Scythian envoys got Cleomenes into the nasty habit of drinking unmixed wine. According to the Spartans, it was directly due to this that Cleomenes eventually went mad. But Herodotus shows here that he wasn’t at all happy with this explanation; and certainly the Demon Rum doesn’t show much sign of influencing Cleomenes’ behaviour for a good long time, if at all. Herodotus adds his own opinion as to why Cleomenes died as he did: that he was ‘paying the penalty to Demaratus’, ‘making retribution for what he did to Demaratus’. Herodotus uses the word t
isin, which he employs elsewhere several times for a punishment which overtakes someone by the will of the gods or of fate. For example, the flying serpents of Arabia, according to Herodotus (III 109, esp. section 2), kill the male in the very act of mating; but their offspring then eat their way out of their mother—and in this manner the female ‘makes ’ retribution to the male’: the very same words, t
isin apot
inei. ***

430

Athenian Democratic Origins

Well, those are the main events of Cleomenes’ reign (I’ve omitted one or two minor ones). I think they are not at all well explained by Herodotus. I now want to suggest that, partly with the aid of information embedded in other parts of Herodotus’ narrative and partly by taking into account other evidence, we can construct a very probable picture of a really great (if very ruthless) man, who stands out among the Spartans of his day perhaps as much as Themistocles among the Athenians—another character, incidentally, who receives very inadequate treatment from Herodotus, but of whom (unlike Cleomenes) there is a tremendous and welldeserved panegyric in Thucydides. I am going to suggest that Cleomenes, like few other Spartans of his day, had a very real understanding of the grave threat to Greece from Persia, and that most of his actions, right down to and even including his alleged treachery with the Arcadians, are best understood in the light of this overriding awareness. First, let me just mention the existence of certain passages in Herodotus which, when taken together, show that as early as the 540s, when Persia threatened to conquer Croesus’ kingdom of Lydia and take over the Asiatic Greeks, there were Spartans who were prepared to see Sparta as the natural leader of all Greeks, and (in response to an Ionian-Aeolian embassy) actually warned Cyrus off conquering Greeks anywhere.7 Of course, Sparta wasn’t a naval power, and there really wasn’t anything she could do to stop Persian expansion in Asia. But, in 525–4, when Polycrates the powerful tyrant of Samos medised and openly submitted to Persia (III 44), Sparta launched an expedition against him, which ended in total failure (III 45.2–47, 54–6). Rather later, around 516 (when Polycrates was dead), Sparta refused to respond to an appeal for help against Persia from the ex-tyrant of Samos, Maeandrius (III 148); but Persia now had a very firm grip on Samos and many of the other offshore islands of the Eastern Aegean and there was no conceivable chance of effective help coming from Sparta. In about 515, however, Sparta may have put down the tyrant Lygdamis of Naxos, who may have been medising; and Sparta may have hoped that Naxos would take the place once occupied by Samos as the principal naval power in the Aegean. I can’t give the arguments now, as they’re very complicated, and I must admit 7

Hdts I 152.3; cf. l41.4, 152.1–2.

Herodotus and King Cleomenes I of Sparta

431

that many people wouldn’t accept the view I’ve just mentioned, or alternatively would put Sparta’s intervention in Naxos as early as 525–4.8 I believe that the Spartan ejection of Hippias from Athens in 510 is best explained by the fact that Hippias had just publicly advertised his medism. The evidence for Hippias’ medism, and for dating it securely to 514–13 or the year or two immediately following, is provided by a passage in Thucydides, as partly elucidated by Wade-Gery,9 to whose account I have something essential to add. Wade-Gery shows how Hippias, frightened by the murder of his brother Hipparchus in 514, decided to insure himself by marrying his daughter Archedice to Aiantides, the son of the tyrant of Lampsacus, Hippoclus, a prominent pro-Persian. Thucydides gives the charming epitaph of Archedice, in two elegiac couplets: in it she is made to boast of being the ‘daughter, wife, sister and mother of tyrants’—the word tyr
annvn is actually used. Certainly the murder of Hipparchus, in 514, must have been one of the things that drove Hippias to medise, for his own protection, as indeed Thucydides’ narrative implies; but I suggest there was something else, perhaps even more decisive. After his transDanubian expedition of (probably) 514 or 513, Darius left a large army in Europe under Megabazus (Herodotus no doubt exaggerates when he puts it at 80,000 men), which conquered virtually the whole Aegean seaboard as far as Macedon10 and thus endangered one of Hippias’ most vital possessions: the valuable gold mines in the area of Mount Pangaeum on the Thracian coast, east of the River Strymon, from which Hippias no doubt drew a large income, as his father Peisistratus had done. Hippias was now likely to lose control of the mines unless he submitted to Persia. When he married his daughter to a particular pet of Darius, any intelligent 8 G. L. Huxley, Early Sparta (1962), 74–5, with 139 nn. 524–7, is just one recent brief treatment of this subject. The alleged deposition of Lygdamis by the Spartans may have something to do with the fact that the so-called ‘Thalassocracy List’ taken by Eusebius from Diodorus seems to allow for a two-year Spartan supremacy of perhaps 517–15, followed by a ten-year Naxian supremacy. But for my own part I would prefer not to take any serious account of the List. W. G. Forrest’s article, ‘Two chronographic notes’ in CQ n.s.19 (1969) 95–106, is the latest treatment of it that I happen to have seen: I would accept very little of this myself, and I mention the article merely for its bibliographical value. For the List, see Eusebius, Chron. (ed. A. Schoene), I 225. 9 Thuc. VI 59, esp. section 3: see Wade-Gery, EGH 159–61. 10 Hdts IV 143.l, 3; 144.3; V 1–2, 14–17; cf. 11, 23–24.1.

432

Athenian Democratic Origins

Greek could see that Hippias had joined the ranks of the medising tyrants. It is on these lines, as I see it, that we should explain the Spartan expedition led by Cleomenes in 510 to expel Hippias— preceded, incidentally, by an unsuccessful naval expedition under Anchimolius, probably in 511. The approximate chronology is quite clear: murder of Hipparchus and Persian threat to the Pangaean mines in 514–13; in (probably) 513 or 512 the marriage of Archedice; and in 511 and 510 Sparta’s action against Hippias. You may say, ‘Oh, but in intervening at Athens Cleomenes need have had no other motive than to increase Sparta’s sphere of hegemony.’ I agree that one can’t actually prove anti-Persian motives here. But why should Sparta expel the Peisistratids at this precise moment, when they had always been their very good friends in the past? Since the friendship between Sparta and the Peisistratids has recently been denied by Professor Forrest, let me emphasize that Herodotus speaks of it on no fewer than three separate occasions and I see not the least reason to doubt it.11 The next episode we need to notice is Cleomenes’ personal intervention at Athens in 508–7, to stop the reforms of Cleisthenes. This may well have been due not simply to a fear of democracy as such (although I dare say that was present) but also to a fear of the instability which drastic reforms of that nature might be expected to produce in a city that was now Sparta’s ally—although of course the very reverse actually happened, and Athens became united and stable. As a reaction to Cleomenes’ intervention, the Athenians 11 [Forrest, HS 81–2, cf. 86, 89.] The only argument against accepting the emphatic and repeated statements of Herodotus (V 63.2; 90.1; 91.2), who is followed by Aristotle (Ath. Pol. 19.4), is that the Peisistratids were friendly with Argos. But it isn’t as if there was a military alliance between Peisistratid Athens and Argos, which might have alarmed the Spartans. The Peisistratids took care to develop friendly relations with many other states, and I see no reason to suppose that their good relationship with Argos need have aroused the least anxiety at Sparta. So the sudden change of policy by Sparta towards the Peisistratids, precisely in 511–10, does require an explanation. Of course I would not entirely discount the standard explanation of the Spartan ejection of Hippias: that it was due to the Athenian Alcmaeonids working the Delphic Oracle (Hdts V 62–63.1, etc.). Incidentally, the text of Herodotus V 63.1, as we have it, attributes this story to ‘the Athenians’, oi‘ ’AuZna&ioi (Schweigha¨user and a few others have wanted to substitute here oi‘ Lakedaim
onioi, but I see no reason for this). At any rate, in support of my argument that Hippias’ medism played an essential part in making the Spartans decide to get rid of him, I would link these events with the expedition against Polycrates in 525–6: you will remember that it was precisely when Polycrates openly medised (p. 430 above) that Sparta launched an expedition against him. The two cases are parallel.

Herodotus and King Cleomenes I of Sparta

433

actually made submission to Persia (Hdts V 73, esp. section 3). Herodotus makes a very ambiguous comment, that when the Athenian ambassadors who gave earth and water at Sardis returned home, ‘they were much blamed’; but he doesn’t say that the submission was actually repudiated; and if you look at a later passage, V 96.2, referring to another Athenian embassy sent to Sardis in the very last years of the sixth century, you will see that up till then the submission couldn’t have been repudiated, since the Athenians evidently felt they had a right to protest at the Persians harbouring Hippias. (Many historians have paid insufficient attention to this.) It was only when the Persian satrap Artaphernes now ordered the Athenians to take Hippias back that they repudiated the submission. This, to my mind, helps to explain the great expedition against Athens organised by Cleomenes in c. 506, which was such a disaster. Let me make myself clear: I’m not presenting Cleomenes as in any sense a panhellenic figure. No Spartan I’ve ever come across was that: every Spartan would naturally think solely in terms of Sparta’s interest—the indoctrination of the Spartan youth was terrific. But if Persia once took over Athens, then Sparta would be in a very dangerous position, because her helots might well revolt (as they did on several occasions), and Argos, her great rival for Peloponnesian hegemony, was very likely to be tempted by a Persian offer to make her the dominant city in the Peloponnese, in place of Sparta, and medise. Argos never in fact lifted a finger to help the other Greeks against Persia, and there were ugly rumours that she had actually promised help to Persia, though she never in fact gave any.12 (It’s interesting, by the way, to find Herodotus saying, in VII 149.3, that the Argives themselves declared they would rather have submitted to Persia than yield to the Spartans in any respect.) And that brings me to the campaign of Sepeia, conducted by Cleomenes and his Spartan army against the Argives—a much more remarkable series of events than most people (I think) realise. The horrible slaughter of some six thousand Argives recorded by Herodotus (VII 148.2) must have wiped out virtually the whole of the Argive hoplite army. As I said earlier, I don’t myself know of any greater loss in war by any Greek state at any time; as far as I 12

Hdts VII 148–52 (esp. 149.3); IX 12.

434

Athenian Democratic Origins

know, one has to go nearly another four centuries, to the battle of Sellasia in 222, before one finds a comparable disaster—this time to Sparta. But the mass killing of Argives is not the only noteworthy feature of Cleomenes’ activities. If you read Herodotus’ account carefully, you will see that Cleomenes, after making what one might call the orthodox approach towards Argos, from the south-west, towards the River Erasinos, suddenly turned back, crossed the Argolic Gulf from west to east, and made his approach from the south-east (VI 76). Why? And why afterwards is he found at the Argive Heraeum, their great Temple of Hera, some five miles to the north of Argos?—and incidentally committing a characteristically brutal act, having the priest flogged for objecting to his sacrificing there (VI 81). The explanation of these episodes comes from later passages in Herodotus and the great inscription known as the ‘Serpent Column’ recording the victories of the Greeks over the Persians in 480–79.13 Two neighbours of Argos which had once been very important, namely Mycenae and Tiryns, had been conquered and taken over by Argos; but Cleomenes must have reconstituted them as independent cities, for they sent contingents to the Greek army that fought the Persians at Plataea in 479, whereas of course Argos sat tight and did nothing, and was suspected of wanting to help the Persians (p. 433 with n. 12 above). Hence, surely, the crossing of the Argolic Gulf, which would bring Cleomenes and his army into the immediate neighbourhood of Tiryns; Sepeia is close to Tiryns; and the Argive Heraeum is close to Mycenae. So Cleomenes both destroyed virtually the whole Argive hoplite army, and set up two independent states on her borders—one, Tiryns, between her and the best harbour on the Argolic Gulf, at Nauplia (an important harbour today), which would make it much more difficult for her to receive a Persian force landing in the Gulf. All this would effectively knock Argos out as a military power for a long time; and I suggest that it was a much better policy, from Sparta’s point of view, than merely forcing the Argives to submit to Sparta (which they would never willingly do), or turning them into Helots, which seems to me out of the question—Sparta already had enough trouble, keeping down the Helots of Laconia and especially Messenia. 13 Hdts IX 28.4; 31.3; cf. VI 83; Serpent Column: M/L 27 ¼ Fornara 59, coils 6–7; and see Paus. V 23.2.

Herodotus and King Cleomenes I of Sparta

435

I now want to stress two features of Cleomenes’ actions in taking the hostages from Aegina, which I have already mentioned (pp. 427–8 with n. 5 above). First, he had been at enmity with his co-king Demaratus for some fifteen years, ever since the great fiasco of c. 506; but it was only now that he took extraordinary steps to have him removed—steps which, if they were discovered (as they were) would certainly lead to his own ruin and possibly to his execution, for the Spartans believed deeply in the Delphic Oracle, and they were bound to regard tampering with it as a capital crime. And what benefit did Cleomenes obtain personally from the whole episode? He got the hostages from Aegina. But here I come to my second point. What proves conclusively, to my mind, that he was not just feathering his own nest here, but acting in a purely selfless and patriotic way, for Sparta’s benefit, is the fact that when he got the hostages he didn’t take them back to Sparta, but handed them over to Athens, thus putting them entirely out of his own control—surely because he realised that if the truth about his fixing the Oracle came out, the hostages would very likely be returned; and he was determined to prevent that. (Actually, there was great trouble later on, because the Athenians refused to give back the hostages even after the Persians had returned home in 490: see VI 85–6.) But what about the final episode in Cleomenes’ life, you will say. What was he doing, uniting the Arcadians against Sparta (p. 429 above)? Well, that was what the Spartans told Herodotus. But what most people overlook is the significant fact that the people Cleomenes is dealing with here are oi‘ ’Ark
adeB, the Arcadians. This fact is significant because Sparta always refused to have anything to do with a united Arcadia. On the principle of divide et impera, she would only deal with the individual Arcadian cities: Tegea, Mantinea and the rest. It was only when Sparta was at her most prostrate, after her disastrous defeat by Thebes at Leuctra in 371, that an Arcadian League (calling itself oi‘ ’Ark
adeB) was formally organised—and even then it would hardly have come about without the help of the Thebans, who invaded the Peloponnese in force in 370–69 under Epaminondas. The very fact that Cleomenes in 490 is dealing with oi‘ ’Ark
adeB is highly significant—especially as we happen to know, mainly from some numismatic evidence,14 14 W. P. Wallace, ‘Kleomenes, Marathon, the Helots, and Arkadia’, JHS 74 (1954) 32–5; see esp. Hdts VI 74.1. [But see Afterword, p. 439 below.]

436

Athenian Democratic Origins

that at this very time there was a real movement in Arcadia towards uniting in a League—a movement which, of course, as I’ve said, did not come to fruition until 369, after the collapse of Spartan power. Now put yourself in the position of a highly intelligent Spartan in the late 490s. The Persian king Darius is known to be organising an expedition against Athens and Eretria, the two cities of Old Greece which had given assistance to the Ionians in their revolt against Persia in 499 and the years following. And it’s hardly likely (since earlier he had reduced virtually the whole of the North Aegean seaboard) that Darius will be content with merely punishing Eretria and Athens. Now, the Arcadians want political unity: they want to be oi‘ ’Ark
adeB. And they are your principal allies. If Persia subdues Athens and then moves against the Peloponnese, the loyalty of the Arcadians will be vital. Why not give them what they want, even if this involves a much greater degree of independence than they have at present? Now Cleomenes was a man in exile: he could demand nothing from the Arcadians. If they were being asked, as oi‘ ’Ark
adeB, to swear a great oath to follow him whithersoever he might lead them, then he must have promised to give them what they wanted: that Sparta, when he was king again, would recognise oi‘ ’Ark
adeB as a united people. And he may well have said to them, ‘Of course, you may have to help me, in effect, to become tyrant in order to get this’! Cleomenes may even have been making promises to the Helots. There is a tradition (rather a late tradition, first found in Plato) that it was a Helot revolt in 490 which made the Spartans arrive late at Marathon after the battle was over (see Wallace’s article, cited in n. 14). There is also some possible epigraphic evidence, and some numismatic evidence from Zancle-Messana (see Wallace again). I’ve never been able to make up my own mind about this: on the existence of the revolt I’m 50 : 50. But again, an intelligent Spartan might well realise that if Persia moved against the Peloponnese the Helots were likely to rise; and to prevent that, he might be willing to hold out to them hopes of some improvement in their situation. A little later, Pausanias, the victor of Plataea, is said by Thucydides (I 132.4) to have promised the Helots freedom and citizenship if they would help him to achieve his designs—which of course, according to the information available to Thucydides (whether true or false), were revolutionary, and no doubt included helping to make Pausanias tyrant. If Cleomenes acted in a similar way,

Herodotus and King Cleomenes I of Sparta

437

then, when he died in mysterious circumstances at Sparta, the Helots might easily revolt. I would leave that open. But my own belief, of course, is that the Spartans decoyed Cleomenes back to Sparta, and then simply murdered him. I realise that I have left several loose ends, in particular why Cleomenes, when he fled from Sparta in 490, went first to Thessaly. I know of absolutely no relevant evidence, and I don’t want to guess; but surely it’s very likely to have had something to do with the question of Persia. The Thessalians were deeply divided here: the Aleuads of Larisa medised thoroughly, but other aristocratic families wanted to resist in 480.15 Perhaps I should mention that some have inferred an alliance between Sparta and the Thessalians, or some Thessalians, from Pindar’s ’ ia Tenth Pythian, of 498, which opens with the words, olb
Lakeda
imvn, m
akaira Qessal
ia [‘Lakedaimon is rich, Thessaly blessed’].16 But I see no certainty of any kind here. The only other matter I feel obliged to mention is the famous reply of Cleomenes to the Athenian priestess who forbade him to enter the shrine of Athena on the ground that he was a Dorian. He ’ is said by Herodotus to have replied: v
g
ynai, all’ oy’ Dvrie
yB ’ ’ ’Axai
ei’ mi all oB (‘Madam, I am not a Dorian but an Achaean) [V 72.3]—and here we may remember that Cleomenes’ younger brother had actually been named Dorieus! All I would say here is that to me Cleomenes stands in the line of succession going back to the mid-sixth century (perhaps to Chilon), which claimed for Sparta a hegemony which went right back not merely to the return of the Heraclids after the Trojan War, but to Agamemnon the Achaean himself. This attitude, new in the mid-sixth century, was associated with the bringing to Sparta of the bones of Orestes (and later of his son Tisamenus), making Agamemnon himself king of Sparta, and locating his tomb at Spartan Amyclae. This policy received support from Delphi, and from the poets Stesichorus and Pindar, and it was a great success.17 15 [The evidence is discussed by H. D. Westlake, ‘The Medism of Thessaly’, JHS 56 (1936) 12–24 and (briefly) in A. R. Burn, Persia and the Greeks (London 1962), 269, 341.] 16 [For example C. M. Bowra, Pindar (Oxford, 1964), 104–5.] 17 I have [discussed this, and given references to the modern literature,] in OPW 96–7. [The main source is Hdts I 65–8; cf. Paus. VII 1.8. For the poets, see Stesichorus, fr. 39 PMG (¼ scholiast on Eur. Orestes 46) and Pindar, Pythian XI 16, 31–2, Nemean XI 33–4.]

438

Athenian Democratic Origins

I think I have said enough about Cleomenes. His great achievements were belittled by his many enemies; his undoubted ruthlessness was much emphasised; and his patriotic motives were thoroughly obscured. He is anything but an attractive figure. But he seems to me to be the one Spartan of the historical period, before Lysander (or perhaps Archidamus), who has a real claim to greatness.

AFTERWORD Readers interested in Herodotus can now consult the commentaries on individual books published by the Fondazione Lorenzo Valla. Most of what he tells us about Cleomenes is to be found in books V and VI, edited in this series by Giuseppe Nenci, with generous bibliographies. An English translation is in preparation. The most influential view of Cleomenes in the 1970s was probably that of Ste. Croix’s younger colleague and friend George Forrest. Forrest (HS 80–94) saw Cleomenes as a meddler and a failure who weakened the power of Sparta, a view sharply opposed here by Ste. Croix. The three most important subsequent treatments of Cleomenes are those of Carlier, Cartledge, and Cawkwell. Pierre Carlier focuses on the question of who was responsible for the initiation and conduct of foreign policy during this period—a restricted theme, one might think, but one that leads to a wide-ranging discussion. Paul Cartledge gives the best account of his reign, succinct but nuanced and containing much detail. George Cawkwell argues for a Cleomenes who believed that Sparta should be the leader (hegemon) of the mainland Greeks against the Persian threat. His opponents, whose distorted account led Herodotus astray, feared that his policy would eventually corrupt Spartan values. Frustrated in Sparta, at the end of his life he was seeking support in Arcadia and Thessaly. Further discussion in Thommen, 85–98; we have not seen Klein’s dissertation. Alan Griffiths’s thought-provoking article differs from the mainstream in starting out ‘not from historical but from literary considerations’. He characterises much that Herodotus tells us about Cleomenes as folktale, ‘free-floating stories of the malicious anti-tyrant type’; in particular, he finds strikingly similar elements in his tales about Cambyses. Georges Devereux’s ‘ethnopsychoanalytic study’ of the king, already signalled in Forrest, has been assembled posthumously from his notes. Heavily Freudian and largely uncritical, it diagnoses paranoid schizophre-

Afterword

439

nia (but see Cawkwell, ‘Cleomenes’, 508–9). One would like to have heard Ste. Croix’s comments on this. There have been numerous studies of specific problems. The majority of scholars, like Ste. Croix (n. 8) are sceptical of the value of the Thalassocracy list: see e.g. Jeffery, AG 252–4. Athens: Robert Parker’s inaugural lecture, Cleomenes on the Acropolis (Oxford, 1998), is a series of variations and finale on the theme of Cleomenes’ confrontation with the priestess of Athena (Hdts V. 72). Recent studies of Aegina (for bibliography see the Endnote to Ch. 10) and of Argos (bibliography in Cartledge SL 159 ¼ 2nd edn. 137) have not thrown any new light on Cleomenes’ activities. Hendriks 42–57 (notes at 189–97), however, has argued in favour of dating the battle of Sepeia in 494. The last stages of the intriguing king’s career remain controversial. Thessaly: many would accept Daphne Hereward’s emendation of Thessalia in Herodotus’ account to Sellasia, which provides a more rational itinerary and eases chronological compression. Ste. Croix, it seems, was not convinced. Arcadia: T. H. Nielsen has brought powerful arguments against Wallace’s influential numismatic hypothesis (n. 14 on p. 435 above). He accepts Kraay’s view (ACGC 95–8) that the ARKADIKON coinage to which Ste. Croix refers on p. 435 may have begun, not in 490, but some ten or fifteen years later, and believes it may have been issued for a festival, not a confederacy. As Ste. Croix says, if there was ‘a movement in Arcadia towards uniting in a league’ in c. 490 ‘it did not come to fruition until 369’. There is no agreement on whether or not there was a Helot revolt in 490. The relevant inscription, now M/L 22 (¼ Fornara 38), does not mention the Messenians, and Lionel Pearson (PMA n. 56 on pp. 421–2; cf. 401) suggests that Pausanias’ informants were mistaken in connecting it with them: see Ducat, 142–4. A seventh-century date for this inscription has recently been upheld by Dillon; objections to it are registered in SEG 45 (1995) no. 406. Cawkwell (‘Cleomenes’, 511–12) believes that there was a revolt, but dissociates it from Cleomenes. Cleomenes’ death has been discussed by Harvey, but ‘speculation is here fruitless’ (Cawkwell, ‘Cleomenes’, 508 n. 6). Devereux’s speculations (suicide) must now be added to the diverse studies listed by Harvey 253 n. 1. Carlier, P., ‘La vie politique a` Sparte sous le re`gne de Cle´ome`ne Ier: essai d’interpre´tation’, in Ktema 2 (1977) 65–84 Cartledge, P. A., SL ¼ Sparta and Lakonia: A Regional History 1300–362BC (1979), 143–54 (bibliography 159; and see index, p. 396); 2nd edn. (2002), 123–33 (bibliography 135–7; index 345; addenda 315–17)

440

Athenian Democratic Origins

Cawkwell, G. L., ‘Cleomenes’, in Mnemosyne 46 (1993) 506–27 Devereux, G. (ed. J. Chemouni), Cle´ome`ne le roi fou (1995) Dillon, M. P. J., ‘The Lacedaimonian Dedication to Olympian Zeus: the Date of Meiggs & Lewis 22 (SEG 11, 1203A)’, in ZPE 107 (1995) 60–8 Ducat, J., Les Hilotes (¼ BCH supplement 20) (1990) Forrest, W. G., HS ¼ A History of Sparta 950–192 BC (1968; 2nd edn. 1980) Griffiths, A., ‘Was Cleomenes Mad?’ in A. Powell (ed.) Classical Sparta: Techniques behind her Success (1989), 51–78 Harvey, F. D., ‘Leonidas the Regicide?’, in Arktouros: Hellenic Studies presented to Bernard M. W. Knox, ed. G. W. Bowersock, W. Burkert & M. C. J. Putnam (1979), 253–60 Hendriks, J. H. M., De interpolitieke en internationale Betrekkingen van Argos in de Vijfde Eeuw v. Chr. (Groningen, 1982) Hereward, D., ‘Herodotus vi.74’, in CQ n.s.1 (1951) 146 Jeffery, L. H., Archaic Greece: The City-States c.700–500 BC (1976), 252–4 Klein, S. C., Cleomenes: A Study in Early Spartan Imperialism (diss. Kansas, 1973) Kraay, C. M., ACGC ¼ Archaic and Classical Greek Coins (1976), 95–8 Nenci, G. (ed.), Erodoto: Le Storie libro V (Milan,1994); libro VI (1998) Nielsen, T. H., ‘Was There an Arkadian Confederacy in the Fifth Century BC?’ in More Studies in the Ancient Greek Polis (¼ Historia Einzelschrift 108), ed. M. H. Hansen & K. Raaflaub (1996), 39–61 Parker, R., Cleomenes on the Acropolis (inaugural lecture, Oxford 1998) Pearson, L., PMA ¼ ‘The Pseudo-History of Messenia and its Authors’, in Historia 11 (1962) 397–426 Thommen, L., Lakedaimonion Politeia: die Entstehung der spartanischen Verfassung (¼ Historia Einzelschrift 103) (1995)

Abbreviations and Bibliography

Abbreviations not listed here may be found in the Oxford Classical Dictionary3 (1996), pp. xxix–liv. ASI (Ehrenberg) ¼ Ancient Society and Institutions. Studies Presented to Victor Ehrenberg on his 75th Birthday, ed. E. Badian (1966) ATL ¼ B. D. Meritt, H. T. Wade-Gery and M. F. McGregor, The Athenian Tribute Lists, 4 vols.: I (1939), II (1949), III (1950), IV (1953) CAH ¼ Cambridge Ancient History. References are to the first edition (1923–39) unless otherwise stated. Crux ¼ Crux. Essays presented to G. E. M. de Ste. Croix on his 75th Birthday, ed. P. Cartledge and F. D. Harvey (1985) DAA ¼ Dedications from the Athenian Acropolis, ed. A. E. Raubitschek, with the collaboration of L. H. Jeffery (1949) FD ¼ Fouilles de Delphes FGrH ¼ F. Jacoby, Die Fragmente der griechischen Historiker, I and II (1923–30 and repr., containing the fragments of historians numbered 1–261), III (1940–58 and repr., containing the fragments of historians numbered 262–856). FGrH III b (Suppl.) i [Text] and ii [Notes, etc.] (1954) contains the Commentary (in English) on the fragments of the Atthidographers, nos. 323a–334 in vol. III B (1950). These two volumes are abbreviated as Jac., i and ii. Distinguish FGrH III b (2 vols., Text and Noten, 1955), which contains the Commentary (in German) on the fragments of the remaining historians in vol. III B: nos. 297–322 and 335–607. FHG ¼ Fragmenta Historicorum Graecorum, 5 vols. (1841–72) FIRA2 ¼ Fontes Iuris Romani Antejustiniani, 3 vols., 2nd edn., ed. S. Riccobono etc. (1940–3) Fornara ¼ C. W. Fornara, Archaic Times to the End of the Peloponnesian War. Translated Documents of Greece and Rome, vol. I (1983) IC ¼ Inscriptiones Creticae, ed. M. Guarducci, 4 vols. (1935–50) Jac. i, ii ¼ see under FGrH K/A ¼ R. Kassel and C. Austin, Poetae Comici Graeci (1983– ) L/P ¼ E. Lobel and D. Page, Poetarum Lesbiorum Fragmenta (1955) M/L ¼ R. Meiggs and D. M. Lewis, A Selection of Greek Historical Inscriptions to the End of the Fifth Century B.C. (1969) PMG ¼ Poetae Melici Graeci, ed. D. Page (1962) SGDI ¼ Sammlung der griechischen Dialekt-Inschriften, ed. H. Collitz etc. (1884–1915)

442

Abbreviations and Bibliography

Tod I & II2 ¼ M. N. Tod, A Selection of Greek Historical Inscriptions, 2 vols. (1933–47) Andrewes, A., GT ¼ The Greek Tyrants (1956) —— , PP ¼ ‘Philochoros on Phratries’, in JHS 81 (1961) 1–15 Beloch, K. J., BGRW ¼ Die Bevo¨lkerung der griechisch-ro¨mischen Welt (Historische Beitra¨ge zur Bevo¨lkerungslehre I, 1886) —— , GG ¼ Griechische Geschichte (2nd edn.), I2 i (1924) and ii (1926), II2 i (1927) and ii (1931), III2 i (1922) and ii (1923) Bengtson, H., GG2 ¼ Griechische Geschichte, von den Anfa¨ngen bis in die ro¨mische Kaiserzeit (2nd edn., 1960: 4th edn., 1969) Bloch, H., SHLFC ¼ ‘Studies in Historical Literature of the Fourth Century B.C.’, in Athenian Studies presented to W. S. Ferguson, HSCP Suppl. vol. 1 (1940) 303–76 Bo¨ckh, A., SA I3 ¼ Die Staatshaushaltung der Athener I (3rd edn., 1886, ed. M. Fra¨nkel) Bonner, R. J., and Smith, G., AJHA ¼ The Administration of Justice from Homer to Aristotle, 2 vols., I (1930) and II (1938) Bradeen, D. W., TCR ¼ ‘The Trittyes in Cleisthenes’ Reforms’, in TAPA 86 (1955) 22–30 Busolt, G., GG ¼ Griechische Geschichte, I2 (2nd edn., 1893), II2 (2nd edn., 1895) —— (or Busolt-Swoboda), GS ¼ Griechische Staatskunde, 2 vols, I (1920), and II (1926, ed. H. Swoboda, whose contribution begins at p. 881) Cadoux, T. J., AAKH ¼ ‘The Athenian Archons from Kreon to Hypsichides’, in JHS 68 (1948) 70–123 Calderini, A., O ¼ l’Ostracismo (1945) Carcopino, J., OA2 ¼ L’Ostracisme athe´nien (2nd edn., 1935) Cartledge, P., TPR ¼ ‘ ‘‘Trade and Politics’’ Revisited: Archaic Greece,’ in Garnsey, P. et al., eds., Trade in the Ancient Economy (1983) Cook, R. M., BBKGH ¼ ‘Die Bedeutung der bemalten Keramik fu¨r den griechischen Handel’, in Jahrbuch des deutschen archa¨ologischen Instituts 74 (1959) 114–23 Day, J. H., and Chambers, M. H., AHAD ¼ Aristotle’s History of Athenian Democracy (U. Calif. Pub. Hist. lxxiii, 1962) Dover, K. J., AO ¼ ‘Androtion on Ostracism’, in CR 77 ¼ n.s. 13 (1963) 256–7 Ehrenberg, V., GS ¼ The Greek State (1959) —— , PA2 ¼ The People of Aristophanes (2nd edn., 1951) Eliot, C. W. J., CDA ¼ Coastal Demes of Attica (1962) Finley, M. I., AD ¼ ‘Athenian Demagogues’, in Past and Present 21 (1962) 3–24 Forrest, W. G., EGD ¼ The Emergence of Greek Democracy (1966) —— , HS ¼ A History of Sparta, 950–192 B.C. (1968)

Abbreviations and Bibliography

443

Freeman, K., WLS ¼ The Work and Life of Solon (1926) Fritz, K. von, and Kapp, E., ACA ¼ Aristotle’s Constitution of Athens and Related Texts (1950) Fuks, A., AC ¼ The Ancestral Constitution (1953) Gilbert, G., CASA ¼ The Constitutional Antiquities of Sparta and Athens (1895: Eng. trans., by E. J. Brooks and T. Nicklin, of HGS ¼ Handbuch der griechischen Staatsalterthu¨mer I2, 1893) Glotz, G., AGW ¼ Ancient Greece at Work. An Economic History of Greece from the Homeric Period to the Roman Conquest (1926: Eng. trans., by M. R. Dobie, of TGA ¼ Le travail dans la Gre`ce ancienne, 1920) —— , GC ¼ The Greek City and its Institutions (1929: Eng. trans., by N. Mallinson, of La Cite´ grecque. Le developpement des institutions, 1920) Glotz, G., and Cohen, R., HG ¼ Histoire ge´ne´rale: histoire grecque, I. Des origins aux guerres me´diques (4th edn., 1948), II. La Gre`ce au Ve sie`cle (4th edn., 1948), III. La Gre`ce au IVe sie`cle (1936, repr. 1941) Gomme, A. W., EGHL ¼ Essays in Greek History and Literature (1937) —— , HCT ¼ A Historical Commentary on Thucydides (1945–56), I (1945, on Book I), II (on Books II–III) and III (on Books IV and V 1–24, 1956) —— , MEGHL ¼ More Essays in Greek History and Literature (1962) Grote, G., HG ¼ History of Greece (new edn. in 10 vols., 1888) Hammond, N. G. L., HG ¼ A History of Greece to 322 B.C. (1959) Hansen, M. H., AD ¼ The Athenian Democracy in the Age of Demosthenes (1991) Harding, P., ATA ¼ Androtion and the Atthis (1994) Harrison, A. R. W., LMAEFC ¼ ‘Law-making at Athens at the End of the Fifth Century B.C.’, in JHS 75 (1955) 26–35 Hasebroek, J., TPAG ¼ Trade and Politics in Ancient Greece (1933: Eng. trans., by L. M. Fraser and D. C. Macgregor, of Staat und Handel im alten Griechenland, 1928) Head, B. V., HN2 ¼ Historia Numorum: A Manual of Greek Numismatics (2nd edn., 1911) Headlam, J. W., ELA2 ¼ Election by Lot at Athens (Cambridge Historical Essays, iv, 2nd edn., 1933) Heichelheim, F. M., AEH ¼ An Ancient Economic History: from the Palaeolithic Age to the Migrations of the Germanic, Slavic and Arabic Nations, 3 vols. (1958–70: Eng. trans., by J. Stevens, of WA ¼ Wirtschaftsgeschichte des Altertums: vom pala¨olithikum bis zur vo¨lkerwanderung der Germanen, Slaven, und Araber, 1938) Hignett, C., HAC ¼ A History of the Athenian Constitution to the End of the Fifth Century B.C. (1952) Holm, A., HG ¼ History of Greece, 4 vols. (1894–8) Hopper, R. J., BAD ¼ The Basis of the Athenian Democracy (Inaugural Lecture, Sheffield U., 1957)

444

Abbreviations and Bibliography

How, W. W., and Wells, J., CH ¼ A Commentary on Herodotus, 2 vols. (1912; corrected reprint, with additions, 1936) Jacoby, F., A ¼ Atthis: The Local Chronicles of Ancient Athens (1949) —— , GH ¼ Griechische Historiker (1956), reprinting Jacoby’s articles in RE on various Greek historians Jarde´, A., CAG ¼ Les ce´re´ales dans l’antiquite´ grecque, I. La production (Bibliothe`que des e´coles franc¸aises d’Athe`nes et de Rome, fasc. 130, 1925) Jeffery, L. H., LSAG ¼ The Local Scripts of Archaic Greece (1961) Jones, A. H. M., AD ¼ Athenian Democracy (1957) Kahrstedt, U., BGRF ¼ ‘Untersuchungen zu athenischen Beho¨rden, IV. Bemerkungen zur Geschichte des Rats der Fu¨nfhundert’, in Klio 33 (1940) 1–12 —— , NLA ¼ ‘Untersuchungen zu athenischen Beho¨rden, II. Die Nomotheten und die Legislative in Athen’, in Klio 31 (1938), 1–32 —— , SSA ¼ Studien zum o¨ffentlichen Recht Athens, Teil I. Staatsgebiet und Staatsangeho¨rige in Athen (1934) —— , UMA ¼ Studien zum o¨ffentlichen Recht Athens, Teil II. Untersuchungen zur Magistratur in Athen (1936) Keil, B., SVAVA ¼ Die solonische Verfassung in Aristoteles Verfassungsgeschichte Athens (1892) Kirchner, J., PA ¼ Prosopographia Attica, 2 vols. (1901, 1903) Kirsten, ‘Review’ ¼ E. Kirsten, review of S. Welter, Aigina (1938), and H. Winterscheidt, Aigina. Eine Untersuchung u¨ber seine Gesellschaft and Wirtschaft (1938), in Gnomon 18 (1942) 289–311 Kraay, C. M., ACGC ¼ Archaic and Classical Greek Coins (1976) —— , HSCOC ¼ ‘Hoards, Small Change and the Origins of Coinage’, in JHS 84 (1964) 76–91 Larsen, J. A. O., RGGRH ¼ Representative Government in Greek and Roman History (Sather Classical Lectures 28, 1955) Ledl, A., SAAV ¼ Studien zur a¨lteren athenischen Verfassungsgeschichte (1914) Le´veˆque, P., and Vidal-Naquet, P., CA ¼ Clisthe`ne l’Athe´nien: essai sur la representation de l’espace et du temps dans la pense´e politique grecque de la fin du VIe sie`cle a` la mort de Platon (1964) Lewis, D. M., CA ¼ ‘Cleisthenes and Attica’, in Historia 12 (1963) 22–40 Lipsius, J. H., ARR ¼ Das Attische Recht und Rechtsverfahren (1905–15) Macan, R. W., Hdts IV–VI ¼ Herodotus: The Fourth, Fifth, and Sixth Books, 2 vols. (1895) —— , Hdts VII–IX ¼ Herodotus: The Seventh, Eighth, and Ninth Books, 3 vols. (1908) Meritt, B. D., AFD ¼ Athenian Financial Documents (1932) —— , AY ¼ The Athenian Year (1961)

Abbreviations and Bibliography

445

Meyer, E., FAG II ¼ Forschungen zur alten Geschichte, II. Zur Geschichte des fu¨nften Jahrhunderts v. Chr. (1899) —— , GdA ¼ Geschichte des Altertums Mitchell/Rhodes, DPAG ¼ Mitchell, L. G., and Rhodes, P. J., eds., The Development of the Polis in Archaic Greece (1997) Murray, O., EG ¼ Early Greece2 (1993) Newman, W. L., PA ¼ The Politics of Aristotle, 4 vols., I & II (1887), III & IV (1902) Osborne, R., GM ¼ Greece in the Making (1996) Osborne/Hornblower, RFP ¼ Osborne, R., and Hornblower, S., eds., Ritual, Finance, Politics (1994) Parker, R. C. T., AR ¼ Athenian Religion: A History (1995) Raubitschek, A. E., OO ¼ ‘The Origin of Ostracism’, in AJArch 55 (1951) 221–9 —— , TO ¼ ‘Theophrastus on Ostracism’, in Classica et Mediaevalia 19 (1958) 73–109 Rhodes, P. J., AB ¼ The Athenian Boule (1972) —— , CAAP ¼ A Commentary on the Aristotelian Athenaion Politeia (1981, reissued with addenda 1993) Roebuck, C., GTGE ¼ ‘The Grain Trade Between Greece and Egypt’, in Classical Philology 45 (1950) 236–47, repr. in his Economy and Society in the Early Greek World (1979), 30–41 —— , ITC ¼ Ionian Trade and Colonisation (1959) —— , ON ¼ ‘The Organization of Naukratis’, in Classical Philology 46 (1951) 212–220, repr. in his Economy and Society in the Early Greek World (1979), 44–52 Ross, W. D., DAT ¼ ‘The Development of Aristotle’s Thought’, in PBA 43 (1957) 63–78 Ruschenbusch, E., PP ¼ ‘PATPIOS POLITEIA’, in Historia 7 (1958) 398–424 Ste. Croix, G. E. M. de, CAE ¼ ‘The Character of the Athenian Empire’, in Historia 3 (1954/5) 1–41 —— , CFT ¼ ‘The Constitution of the Five Thousand’, in Historia 5 (1956) 1–23 —— , CSAGW ¼ The Class Struggle in the Ancient Greek World (1981) —— , DTAEFC ¼ ‘Demosthenes’ timema and the Athenian eisphora in the Fourth Century B.C.’, in Classica et Mediaevalia 14 (1953) 30–70 —— , NJAE ¼ ‘Notes on Jurisdiction in the Athenian Empire’, in CQ 55 ¼ n.s. 11 (1961) 94–112 (I), 268–80 (II) —— , OPW ¼ The Origins of the Peloponnesian War (1972) ’ iB: Storia della reppublica Ateniese dalle origini Sanctis, G. de, A2 ¼ Aty
alla eta` di Pericle (2nd edn., 1912)

446

Abbreviations and Bibliography

Sandys, J. E., ACA2 ¼ Aristotle’s Constitution of Athens (2nd edn., 1912) Seltman, C. T., GC2 ¼ Greek Coins: A History of Metallic Currency and Coinage down to the Fall of the Hellenistic Kingdoms (2nd edn., 1955) Stier, H., WAW I ¼ Westermanns Atlas zur Weltgeschichte, I. Vorzeit und Altertum (1956) Sumner, G. V., AF6 ¼ ‘Androtion F6 and Ath. Pol. 22’, in BICS 11 (1964) 79–86 Thompson, W. E., TTAH ¼ ‘Three Thousand Acharnian Hoplites’, in Historia 13 (1964) 400–13 Ure, P. N., OT ¼ The Origin of Tyranny (1922) Usener, H., OWA ¼ ‘Organisation der wissenschaftlichen Arbeit’, in Vortra¨ge und Aufsa¨tze (1907), 67–102 Wade-Gery, H. T., EGH ¼ Essays in Greek History (1958) Weil, R., AH ¼ Aristote et l’histoire: essai sur la Politique (1960) Welwei, K.-W., Athen ¼ Athen. Vom neolithischen Siedlungsplatz zur archaischen Grosspolis (1992) Whitehead, D., DA ¼ The Demes of Attica (1986) Wilamowitz-Moellendorff, U. von, AuA ¼ Aristoteles und Athen, 2 vols. (1893) Wyse, W., SI ¼ The Speeches of Isaeus (1904)

Index

Achaean 437 Adcock, F. E. 233, 297, 300, 301 Aegina 356, ch. 10 (371–420), 435 aristocracy/oligarchy 371–3, 377, 407, 409, 410 and see ruling class Athenian colonists 377, 409–10 athletes 378–9; 427 clans 378; Cleomenes and 422, 427–8 coin standards 396–9 coinage 37–8, 389, 392–6, 411, 414–15 and corn trade 406–7 drachma 397 and East 407–9 eunomia 379 exports 384 ‘eyesore’ 391, 427 ‘factory’ at Naucratis 400 fortunes see wealth hoplites 387–8 ‘industry’ 383–4 infertility 386, 389 justice 379, 381 land-owners 374, 386 medism 407–8, 427 ‘mercantile aristocracy’ 371–6, 385, 392, 410–11, 417 merchants 384–5, 390, 391 ‘merchant fleet’ 375, 382, 412 metics 374–5, 392, 410 and Naucratis 399–406, 416–19 petty trade 390 Pindar on 378–82, 412 piracy 391

population 387–8, 409, 414 revolution led by Nicodromus 376–8 ruling class, in poetry 378–82 ruling class, wealth of 386–9 ships 380, 389 size 386 slaves 409, 410 trade ch. 10 passim traders, who were they 374 as ‘trading/industrial city’ 372, 392 underprivileged classes 386–8 victors in Panhellenic games 378–9, 399, 412 wars with Athens 413 warships 388, 414 wealth 386–7; origin of 388–9 weight standards 396–9 and xenia 380–2, 412 Aelian 181 Aeschylus 18 agricultural land 349–52 agroikoi 81 Aithiops 360–1 Al Mina 363, 364, 368, 408 Alcibiades 134, 203, 205 Alcmeonid family 132–3, 134, 151, 152, 158, 163, 190, 237 Aleuads 407, 437 aliens 234–5, 241–2, 245–7, 252 amphictionies 417 Andrewes, A. ch. 3 (109–28), 326, 400 Andron 283, 293, 305 Androtion 110, 120

448

Index

Androtion (cont.) and Aristotle 187–8, 288, 293 cites Solon’s laws 320 exile of 284, 304–5 on Kolakretai 289, 290 on ostracism 182–3, 184–8, 197, 229 Philochorus, connexion with 291–3, 326 research by 304–5, 307 as source for Athenaion Politeia 277–8, 283–301, 304–7 as principal source thereof 284–6, 326–7 Anchimolius 432 Anthemion, dedication of 70–1 Antipater 62–3, 261 Antiphon 13, 274–5, 314 apodektai 289 Apollodorus 10 Arabia, flying snakes of 429 Arcadia 391, 429; Arcadian league (?) 435–6 archaeological evidence 366–7, 368–9 Archedice 247–8, 431, 432 Archias 352, 361 Archilochus 360–1 archives 281, 303–9 archon basileus 55 archons and archonship 7, 9, 56, 132–3, 152, 230 appointment of 89–104, 166, 215–23, 226–8 Aristotle on 218–19, 228 demotion of 221–3, 230 eligibility 217 eponymous 55, 56, 82, 100, 132, 133, 221, 223 eupatrid monopoly of 63–4 Isagoras 132, 208–9 list 322, 327

qualifications for 8, 9, 27–8 replacement as supreme magistrates by Strategoi 215–28, 230–2; date 215; democratic? 216–20; real purpose 220–4 Areopagus 82–3, 86–7, 88, 141 Argos: 204, 213, 238, 426; Sepeia, battle of 426–7, 433–4 Aristagoras 425, 426 Aristeides, ostracism of 198, 200, 204–5, 206, 207 aristocrats, aristocracy 352, 361; and see Aegina, Athens Aristogeiton 423 Aristophanes 13, 25, 235, 238, 239 Aristophon 18, 239, 241, 251 Aristotle 6, 76–7, 224, 237, 258, 324–5 and archives/documents 303–22 on Areopagus 82 and Atthidographers 187, 286–307 and axones 317–18, 322 on citizenship 145, 168, 169, 243–4, 246 on Cleisthenes 129–30, 132, 181 collaboration with pupils 256–7, 270 on Council of Four Hundred 83, 87–8 on debt 110, 119–21 on demes 163 on demos 390–1 documents see archives errors in 321 on foreign trade 354 Hektemoroi in 110, 113, 116, 120, 121, 122, 123, 124–5 and Herodotus 270 on Hippeis 25–6

Index on laws and constitutions 265–6, 271 leaves Athens 261–2 as literary source for constitution 137–8, 141 and magistrates, selection of 89–90, 215, 218–19, 228 on metrology 328–32, 336–7, 338, 341 on naukraroi 147–8 Nicomachean Ethics 265–6, 328–41 nobility, definition of 140 on ostracism 181, 182, 183, 184, 186, 192–3, 193–7, 204–6, 213 Politeiai 256, 265–6, 268, 271, 302, 324–5 on politics and law-making 265–6 pupils, collaboration with 256–7, 270 reliability of 30–2, 84, 110, 316–17 research by 266, 268, 270–2, 278, 302, 303, 305–6, 307–22 Rhetoric 217, 265, 270 Rhetorica ad Alexandrum (falsely ascribed to Ar.) 94, 223 on Seisachtheia 110, 114, 119–20 on Solonian laws 78, 310–12, 315–16, 320–1 on tele¯, qualifications for 52–3 on trade 354–5 on tribal reforms 174–6 on trittys system 150, 156 other works 27, 265, 266, 268, 270 works, conflict between 103–4 see also Athenaion Politeia; Politics artaba, artabus 333, 348 Artaphernes 433 Asclepiades 320

449

Assembly 85, 86, 141, 142, 161, 171 Aspasia 240 Aston, T. H. 118 Athenaeus 409 Athenaion Politeia (Aristotle) 6, ch. 7 (254–327) on archons 100–1, 215–6 authorship of 254, 256–7, 326–7 on cavalry 16 census class membership 8–10 on citizenship 168–9, 237, 243–5, 246 on clans, phratries and priesthoods 161 on Cleisthenes 129–30, 137–8, 141, 181 on constitutional reforms 134, 215–6 date of 260–73 on debt 120–1 documentary sources for 278–9, 307–22, 327 Draco, spurious constitution of 7, 27, 321 on eupatrids 81–2 Four Hundred, revolution of 273–4, 275 Hektemoroi in 110, 116, 120, 122 literary sources for 277–86, 289–90, 293–301, 325, 326–7 London papyrus 217, 219, 254 magistrates, selection of 89–91 on ostracism 181–8, 193–4, 197 on Peisistratids 237 Phye story 282–3 and Plutarch 287, 295–301 polemarch, institution of office of 93 and Politics 103–4, 256–77; agreement with 277; conflicts with 273–7 reliability of 110

450

Index

Athenaion Politeia (cont.) research for 266, 268, 270–2, 302, 303, 305–6 revision of 260–73, 326 scholarship on 254–5, 277–8, 325 on Seisachtheia 114, 118, 119–20 on sortition 166, 215, 216, 217, 228 sources 277–327: Androtion 277–8, 283–301, 304–7, 326–7 Cleidemus 282–3, 288, 303 Ephorus 280 Hellanicus 280–1, 287, 303, 322 Herodotus 279, 306 oral tradition 280–1, 305 Philochorus 290–1, 326 Thucydides 279, 306 Xenophon 279, 325 Strategoi, election of 223, 224 superiority to Atthidographers 301–7 tele¯, qualifications for 28–9, 32, 46–7 on trittys system 150, 152 two parts 306, 307–8 Athenian law-code 318 Athens 313, 413, passim agora 347 Cleomenes’ expedition against 424, 433 coinage 392, 396 demos 391 estates, size of 387 Hippias expelled from 423, 431–2 histories of 280–1, 282, 287–8 and Persia, submission to 432–3 and Plataea 422–3 political families 157, 160–1, 208 politics, nature of 208 pottery 408–9

purity of stock 245–6, 252 state archives 308–9 traders 406–7 tyranny 140, 150, 153–4 wars against Aepina 413 and see citizens and citizenship Atthidographers 255 Aristotle and 286–307 see also Androtion; Cleidemus; Hellanicus; Hippias; Philochorus Augustine 372 autarkeia (self-sufficiency) 354 Austin M. M. 417, 418 autochthony 252 axones 31, 38, 315–22, 326–7 Babylon 111–2, 364 Bacchiads 360, 362 Bacchylides 379–80, 381 Badian, E. 230 balance of trade 353 Barker, E. 259 barley 33–41, 46, 66–7, 238 basileus 100 bastards 234–5, 241, 243–4, 248–9, 253 Beloch, K. J. 50, 54, 83, 91–2, 216, 388 on ostracism 188, 190 and trittys system 152, 153 Bengtson, H. 91–2, 193 Blakeway, A. 360, 366 Bloch, H. 185, 266 Boardman, J. 358, 417–18 Bo¨ckh A. 30, 57–8 Boegehold, A. L. 252 Boeotia 422, 424 Bosporus 350 bottomry loans 375–6 boule¯ see Council boundaries 356 Bourriot, F. 178

Index Bowden, H. 418 Bradeen, D. W. 140, 150, 154, 163–4 Braun, T. F. R. G. 417, 418 Brea decree 11 Bresson, A. 418 Britain, Roman 334–5 Brown, W. L. 37 Brucker, G. A. 106 Bugh, G. H. 72 Burn, A. R. 191 Bury, J. B. 426 Busolt, G. 9, 54, 91–2, 172, 193, 216, 297, 317 buying and selling 328, 330–1, 336–7 Byzantium 350–1 Cadoux, T. J. 78 Calderini, A. 193 Callimachus 153 Callisthenes 270 Callixenus 200 capacity, how expressed/ estimated 340, 341–7 Carcopino, J. 193 Carlier, P. 438 Carthage 355, 395 Cartledge, P. A. 438 Carvoran (Hadrian’s Wall) 334 Catana 362 Cavaignac, E. 101, 237 cavalry 11, 14, 15–16, 72, 226, 231 equivalence with Hippeis 7, 8, 19–20, 21, 24, 25–6, 50 Cawkwell, G. L. 368, 438 census classes, see tele¯ Chalcedon 350–1 ´ uboea) 356, 358, 362, Chalcis (E 377, 424 Chalcis (island) 351 Chambers, M. H. 257, 284, 298 Charondas 13

451

Chilon 422, 437 Chios 89, 344–5, 358–9, 391 Choinix 345–6 Chous 344 Chrimes, K. M. T. 35–6 Cimon 234, 237 ostracism of 197, 198, 201, 205, 206, 207 citizens and citizenship (Athenian) 140, 141, 168–9, 233–53 before 451/0 235 and Cleisthenic constitution 131–2, 139, 140, 236 constitutional reforms and 144–7 enfranchisement of new 168–70 number of 243, 246–8 phratries and 143, 144–5, 146, 147, 160–1, 163 qualifications for 131–2, 139, 236, 243–5, 243–5, 247 citizenship laws (Athenian) ch. 6 (233–53) Antipater’s restriction 62–3 Aristophon’s law 241 Demophilus, decree of 242–3 evasion of 245 fourth-century laws 241–2 late fifth century 239–40 Nicomenes’ law 239, 240–1 Pericles 233–4, 236, 237, 238–41, 243, 245–50, 252 purpose of 243–4, 245–51 clans see gene¯ Cleidemus 187, 282–3, 284, 288 Cleisthenes of Athens chs. 4 and 5 (129–232), 286–9, 290–2, 307–9, 313–5, 423–4, 432 in Aristotle 129–30, 132, 181 chronology 177

452

Index

Cleisthenes of Athens (cont.) constitution of 129–79, 292, 313, 423–4, 432–3 Council of Five Hundred 84, 87, 88–9, 164–6, 167 eponymous archon 132 in Herodotus 129–36, 172–3 Isagoras, rivalry with 130, 132, 133 laws 278, 313–5 motives 130, 133–5, 150–1, 163 and ostracism 181–3, 186–9, 193–4, 206–8 tribes 130–1, 132, 143–4, 151–2 trittys system 148–55, 163–4 Cleisthenes of Sicyon 131, 178, 237 Cleitophon 137–8, 313 Cleomenes 43, 88, 170, ch. 11 (421–40) and Aegina 379, 389, 422, 427–8, 435 and Arcadia 429, 435–6 and Argos 426 and Athens 422–4, 432–3, 439 biographical details 422–9 and Cleisthenes, constitution of 423–4, 432–3 daughter 425 death of 429, 437, 439 and Delphi 426, 428–9 and Demaratus, relationship with 424, 428, 435 and Dorieus 352 and Ionian Revolt 425 and Plataea 422–3 cleruchies, see colonies Cloche´, P. 193 Cohen, R. 33–4, 373, 400, 409–11, 413, 414–15 coinage 37–9, 54–5, 403, 414–6 Aegina 389, 392, 394–6, 414–16 Arcadia 435–6, 439

Athens 37–9, 392, 396 Carthage 395 Corinth 392, 395–6 electrum 394 fractional 393–4, 415 origins 393–4, 414–6 ‘Pheidonian’ 390 Phoenician 395 Purpose see origins Sicily 395 Siculo-Punic 395 Silver 394 South Italy 395 and trade 394 standards 396–9 and trade 392, 393–4, 396 Colaeus 385 Collingwood, R. G. 421 colonies, colonisation 11–12, 28, 63, ch. 9 (349–70) causes of 349–70 due to land hunger 349–51, 367–8 trade, as cause of 352–3, 359–67 commerce, see trade ‘commercial rivalry’ 356 concubines 250 Connor, W. R. 72 conscription 18–19 Cook, R. M. 364, 366, 384, 393 Corcyra 350 Corinth 178, 350, 352, 356–7, 360, 362, 392, 395, 404 corn 328, 331–2, 337, 338, 362 and see wheat, barley corn route 364, 406–7 Cornford, F. M. 356 Council of Five Hundred (Cleisthenes) 84, 87, 88–9, 142, 144, 154–5, 161, 164–6, 167, 178 Council of Four Hundred (Solon) 83–9, 141, 142

Index Council of the People (Chios) 89 craftsmen 357 Craterus 235, 309–10 Cratinus 320 Crawley, R. 174 Crete 225 Crius son of Polycritus 427–8 Croesus 430 Crosby, M. 347, 348 Croton 395 cult associations, and political power 141, 157–60 Cumae 363–4, 366 customs duties 355, 356, 407 Cylon 237 Cyrene 161, 173, 349–50, 365 Cyrus 430 daktylos 345–6 Damasias 81–2 Damia and Auxesia 408 Damon 205 Darius 427, 431, 436 Day, J. H. (and Chambers, M. H.) 257, 284, 298 De Sanctis, G. 54, 83, 91–2 debt 110–23, 125 see also Seisachtheia decrees, Athenian 309–10 Delian League 201, 204 Delphi, Delphic Oracle 289, 302, 426, 428–9, 435, 437 Demades 251 demagogues 279 Demaratus 424, 428, 429, 435 demarchs 147–8 demes 131–2, 137, 139–40, 142–6, 162–3, 165, 169, 177, 178–9, 253, 290 deme assemblies 146 registers 242, 244, 250 Demetrius of Phalerum 251 Demiourgoi 82

453

democracy 129 majority voting, introduction of 73–5 Demonax 173–4 Demophilus 242–3 demos 390–1 and power 85–6 and voting 73–5 Demosthenes 12, 18, 66, 239–40, 251, 283 Develin, R. 108 Devereux, G. 438–9 diapse¯phismos, – pse¯phisis 145, 242 Didymus 320 Diodorus 13, 62, 181–2, 194, 213 Diogenes Laertius 89 Dionysius of Halicarnassus 139 divine punishment 429 documents see archives Dorians 411, 428, 437 Dorieus 437 Dover, K. J. 186, 197 Dow, S. 318 dowries 12–13, 30 drachma 38–41, 43, 45–7, 68, 397 Draco 7, 11, 12, 27, 273, 275, 321 Dracontides 314 Dunbabin, T. J. 360, 361, 362, 366 Duncan-Jones, R. P. 348 economics 339 Egypt 333 Ehrenberg, V. 92 eisphora 44, 57–60, 61 election to Council 165–6 of magistrates 89–108, 144, 208, 217–9 Eliot, C. W. J. 136–7, 138, 152, 155, 162, 173 emporia 358–9, 363, 364, 365–6 Naucratis 399–400, 401, 402, 403–4 Emporion (Spain) 364, 366

454

Index

engue¯ (betrothal) 234, 249 Epaminondas 435 Ephialtes 197, 216 Ephorus 181, 279–80 epikleros 12, 30 epimortos 124, 126 Eratosthenes 342 Eretria 350, 356, 358, 427, 428, 436 Erythrae 166 eunomia 379 Eupatrid Order 6 monopoly of archonship 63–4, 80–3, 85 system 82, 140–1, 145, 147, 163–4 Euphorion 319–20 Eusebius 79 Eutocius 342 exile 206 exports/imports 353–9, 362 festivals 263, 378, 439 Figueira, T. J. 348, 412, 413 fines 394 Finley, M. I. 393 Five Thousand 274, 293 Forrest, W. G. 344–5, 438 Four Hundred, revolution of 273–5, 276, 279, 293 Foxhall, L. 72 fruit 345–6 Fuks, A. 313–14 Fustel de Coulanges, N. D. 94–5, 127 Gaius 117 Gallant, T. W. 127 Gamoroi 361 Gardner, P. 407 gene¯ (clans) 6, 147, 159, 162–3, 178 Gibbon, E. 421 Gilbert, G. 193, 278

Glotz, G. 33–4, 193, 373, 400 gno¯rimoi (notables) 390–1 gold 431 Gomme, A. W. 233, 238, 242, 243, 247, 298–9 Gorgo 425 Graham, A. J. 367–8 grain see barley, corn, wheat Gravisca 413 Grayson, C. H. 347–8 Gregory the Great, Pope 333 Griffiths, A. 438 Grote, G. 130, 180, 220 Grundy, G. B. 356 Hammond, N. G. L. 75–6, 108, 182, 190, 233 Hands, A. R. 190 Hansen, M. H. 107–8, 327, 418 Harding, P. 325–6 Harmodius (and Aristogeiton) 129, 153, 423 Harpocration 13, 289, 291, 292, 305 on ostracism 184–8 Harris, E. M. 127 Harrison, A. R. W. 113, 191 Hasebroek, J. 355, 357, 373, 400–1 Haverfield, F. 334 Headlam, J. W. 94, 97 Heichelheim, F. M. 41 Heisterbergk, B. 92 Hektemoroi, ch. 3 (109–28) Heliaia 142 Hellanicus 187, 281, 282, 287–8 Hellenion (Naucratis) 399, 403, 404 Hellespont 406–7 helots 388, 418, 433, 436–7, 439 helot revolt of 490 (?) 436, 439 Henchir Mettich inscription 116 Heraeum, Argive 434 Hermippus 297

Index Herodotus 178, 217, 225, 237, 382, 390, 406–7, 408, 421, 434 on Aeginetans, wealth of 388–9 and Aristotle 270, 279 on Athenian command in Marathon campaign 226–7 on citizenship 131–2, 168 on Cleisthenes 129–36, 172 on Cleomenes ch. 11 (421–40) on colonization 349, 350, 358, 361, 363 on Council of Four Hundred 87–8 on Cyrene 349 and divine punishment 429 method 421–2 on Naucratis 399–406, 417, 418–9 on Nicodromus 376–7 reliability of 421–2 on Sostratus 384–5 sources 425 Hero, Heronian formula 341–2, 343, 347 Hesiod 114, 334, 352, 389 Hesychius 124, 125 hetairai 251 Hignett, C. 53–4, 79, 91–2, 101, 233, 254, 288 on Aristotle 224, 256–7, 264, 284, 290 on Cleisthenic constitutional reforms 138, 143 on Council of Four Hundred 83–4 on ostracism 188, 189, 190, 209–10 on Solon’s nomothesia 77–8, 79 on sortition 92, 215–16, 219 on tyranny 211–12 Hill, G. F. 398 Himera 395 hipparchs 11, 15, 27

455

Hipparchus 182, 183, 193, 198, 199–200, 203 Hippeis 14, 55, 64–5 Aristotle on 25–6, 28–9 cavalry, equivalence with 7, 8, 19–20, 21, 24, 25–6, 50 and military service 13, 22, 23 original qualifications for 46–51, 54 Pollux on 29–30, 56–7 as qualification for office 9–10, 217 taxation 57–8 Hippias (Athenian tyrant) 423, 424, 431–2 Hippias of Elis (polymath) 281 Hippobotai of Chalcis 16 Hippocrates 200 historia (research) 269–70 Holm, A. 193 Homer 96–7 hoplites 8, 13–15, 16–17, 63, 65 in Aegina 387–8 reforms 126–7 Zeugitai, equivalence with 19–20, 21–2, 23–4, 25, 26–7, 50–1 Hopper, R. J. 172 Horden, P. 414 Horoi 115, 120, 121, 122, 125 horse, as status symbol 15–16 Humphreys, S. C. 251 Hyperbolus: ostracism of 192, 194, 196, 198–9, 200, 202–3, 209, 213, 214 imports/exports 353–9, 362, 364 income 40–3, 44 industry, Aeginetan 383–4 infantry, early mounted 65 inheritance 235, 239, 249, 351 Ionian Revolt 425 Isaeus 9, 61

456

Index

Isagoras 130, 132, 133, 146, 161, 208–9, 229–30, 423 ise¯goria 171 Isocrates 18, 60–1, 218, 240, 251, 283, 382 on politics and law-making 265, 271 on sortition 101–3 isonomia 171, 423 Isthmian Games 77–8 Italy, South 361, 363, 364–5, 367, 395 Jacoby, F. 6, 91–2, 242, 243 on Androtion 187–8, 283–5, 289, 290, 291, 294–5 on Aristotle 268, 278, 315–16, 317–18 on Athenaion Politeia 255, 293, 296, 299–300, 301, 302–4, 306, 307–8 on Athenian state archives 308–9 on Athens, history of 280–1 on citizenship 168, 233–4, 238, 246 on Cleidemus 282, 288 on Council of Four Hundred 84 on eupatrid monopoly of archonship 63 on Herodotus 172 on ostracism 182, 185, 187–8, 192 on Philochorus 290, 291, 293 on Solon’s Nomothesia 78 on sortition 216–17 Jaeger, W. 258–9, 264, 269, 324–5 Jarde´, A. 34, 35, 67 Jeffery, L. H. 70–1, 72, 173–4 Jones, A. H. M. 61–2, 63, 66, 114 justice 328–9, 341, 379, 381 Kagan, D. 187, 193, 415 Kahrstedt, U. 20, 22–3, 91–2, 166, 167, 308

on tele¯ 52–3, 68–9 katalogos (register), hoplite and cavalry 16–21, 23–6, 65 Keil, B. 51, 294, 296–301 Keynes, J. M. 393 Kim, H. 415 Kinship 139 King’s Peace 12 Kirsten, E. 384 on Aegina 373, 374, 375, 376, 392, 396, 410–11 kle¯ro¯sis ek prokrito¯n 89–108, 215, 228, 276, 321 Knight, D. W. 229 Kolakretai 9, 55, 83, 289, 290 Kraay, C. M. 37–8, 393–4, 397–8, 415 Kraiker, W. 383 Kroll, J. H. 414–15 kyrbeis 311, 318, 320, 326 and see axones Lade 425 Lampis 385, 410 Lampon 380–1 Lampsacus 431 land, agricultural 349–52 land-hunger 349, 351, 367 ownership of 42, 55, 170–1, 356 redistribution 125 value 32–40, 68 Landtman, G. 112 Lang, M. 341–4, 345–6, 347, 348 Larisa 437 Latte, K. 238 Le´crivain, C. 92 Ledl, A. 91–2, 297 legitimacy see bastards Lelantine plain, war 358 Lemnos 12 Leonidas 425

Index Leontini 361, 362 Leotychidas 428 Lesbos 63 Leuctra 435 Le´veˆque, P. 129 Lewis, D. M. 136–7, 159, 172 and trittys system 149, 152–3, 156–8, 162–3 Libanius 242 Lindus 402–3 loans 111–2, 114 and see debt mercantile 375–6, 410 Locri 361 Lolling, H. 172 lot, see sortition Luria, S. 11–12 luxuries/necessities 353–5 Lydia 430 Lygdamis of Naxos 430 Lysander 377 Lysias 12, 314 Macan, R. W. 408–9 Macedon 427, 431 Maeandrius 422, 430 magistrates 142, 151 class membership and 9 election of 89–96, 166, 208, 273, 276 see also archons and archonship; sortition, Strategoi majority voting, introduction of 73–5 Manichee 372 Mantinea 435 Massalia 364 Marathon campaign 226–7, 230, 376, 428, 436 Marcellinus 237 marriage 234–5, 237, 239, 242, 244, 246–8, 249–50 dowries 12–13, 30 Mathieu, G. 257, 264

457

measures, inaccurate/ fraudulent 333, 338–9, 347–8 and see capacity, metrology Medici, Cosimo de’ 105 medimnos 33–7, 39–43, 45–7, 68, 338–9, 342 medism see Persia Megabazus 351, 431 Megacles 198 Megara (central Greece) 204, 213, 351, 356 Megara (Sicily) 377 Melesias 378 Mendelsohn, I. 111, 112 Menon 200, 205 mercenaries 394 merchants 354–7, 365, 371, 384–5, 390, 391 and see Aegina Meritt, B. D. 261 Messana 367, 436 Messenia 352, 434, 439 metals 351, 353–4, 356, 364, 384, 431 me´tayage 117, 122 metics 51, 60–1, 242, 247, 248, 406 in Aegina 374–5, 376, 392, 410 metra see metrology metre¯te¯s 33–4, 37, 41, 341–2 metrology 294, ch. 8 (328–48) Aristotle on 328–32, 336–7, 338, 341 Athenian corn 337–8 capacity, measures of 341–7 for corn and wine 328–33 cubic units, derived from linear measures? 340–1, 342–3, 346, 347 official 328–9, 330–1, 332–4, 336–7, 340, 346 standards 38, 39–43, 335, 396–8 variations in 331–2, 333, 334, 336–7, 338–9, 348 weights 347–8

458

Index

Metroon 308–9 metronomoi 332–3, 335, 347 me¯troxenoi 234–5, 238, 241 Meyer, E. 5, 40–1, 51, 91–2, 193, 296, 385 Miletus 204, 213, 356, 399, 400, 404, 407, 417, 425–7 military command 92–4, 96–9 reforms 142 military service 8, 13–14, 19–20, 21–3, 48 Miltiades 227, 237 Mnesitheides 215–16 modernist/primitivist 371–2, 399 modius 333–4 Mo¨ller, A. 411, 418–19 money see Coinage morte¯ 124, 126 mortgage 115, 121 Mu¨ller-Stru¨bing, H. 220 Mycenae 434 Myrmidons 389 naucraries/naukraroi 147–8, 161, 165, 290, 305, 315, 321 Naucratis 395, 416–19 administration 399–401, 403–5 and Aegina 399–406, 416–9 archaeology 416, 418, 419 citizens 401 coins 403 and East 417 emporion 399–401, 403 establishment of 399–400, 417–8; date 404–5, 418 ‘factory’ 400 Hellenion 399, 403, 404 magistrates 401 and Miletus 417 and Pharaoh 417 as polis 401–3, 417–8 as port of trade 418 pottery 404

prostatai tou emporiou 399–404, 417 proxenoi 402–3, 418 residents/traders 418 and Rhodes 418 and Aegina 399–406, 416–19 sanctuaries at 399–400, 401, 404–6 as trading colony 363, 364 naukle¯ros 385, 410 naval service 8, 14, 17, 18–19, 21 Naxos (island) 377, 430–1 Naxos (Sicily) 395 necessities/luxuries 353–5 Nehemiah 118, 123 Nicias 203 Nicodromus 376–8, 380 Nicomenes 239, 240–1, 245 Nissen, H. 261–2 nomoi and politeia 271–2, 273 of Solon 278, 310–17 nothoi see bastards numerals 343–4 Oenussae Islands 358–9 office, qualifications for 7, 8–10, 27–8 Ogden, D. 252 oikist 365 ‘Old Oligarch’ see Xenophon Old Testament 118, 123 Olympic Games 78 oral tradition 280–1 Orestes 437 Osborne, R. 252 ostracism 180–215, 229–30 abuse/misuse of 207, 211, 213–5 and see Hyperbolus Alcibiades and 134, 203, 205 Androtion on 182–3, 184–8, 197, 305 Aristeides 198, 200, 204–5, 206, 207

Index ostracism (cont.) Aristotle (Ath. Pol.) on 181–8 Cleisthenes and 181–3, 186–8 Harpocration on 184–8 Hipparchus and 183–8 introduction, date of 181–9 Philochorus on 181, 194, 292 preliminary vote 210 purpose of 189–214, 229–30 democratic envy 192 against Hipparchus 193–4, 200 against ‘Leader of opposition’ 201–3, 206–7 party-political 189–92, 195, 213–4 against Persian sympathizers 202, 214 against potential tyrant 193–6, 199, 204, 206, 212 against powerful men 194–6, 204, 206 and stasis 195, 201–3, 229–30 quorum 210 safeguards 207, 210–11 statistics 229 victims of 182–3, 192–4, 196–207, 209, 213, 214 virtues of 209 ostracophoria 189, 191, 199, 203, 207, 210 Pangaeum 431, 432 Panhellenic festivals 378 Pantaleo 333 Partheniai 352 Pasion 60–1 patrioi nomoi 313 Patterson, C. 252 Pausanias (regent) 382, 436 Pausanias (periegete) 439 peasants 352 peasant proprietors 113–5, 121–2, 124, 126

459

Peiraeus 363, 391, 427 Peisistratids 58, 140, 153–4, 181–2, 211, 432 Peisistratus 89, 140, 153, 237–8, 432 Peloponnese 398–9, 406–7, 421–40 Peloponnesian League 427 Pentakosiomedimnoi 7–9, 24, 53–6, 217 Aristotle on 28, 32 and military service 13, 22, 23 original qualifications for 32–46, 54–5 Pollux on 29, 30, 56–7 treaurers, selection from 8, 9 n. 15, 90, 91 taxation 57–8 Pericles 205, 240, 251 and citizenship laws 233–4, 236, 237, 238–41, 243, 246–50, 251–2, 313, 355 son 240 Persia, Persian wars, medism 202, 211–2, 214, 217, 222, 236, 284, 388, 398, 407–8, 426–8, 430–1, 432–4 petalism 204, 213 Phantocles 11 Pheidon and coinage 390 Philochorus 238, 291, 293 Androtion, connexion with 291–3, 326 Cleisthenes 181, 292 on ostracism 181, 194, 292 Philodemus 266 Phocaea 358–9, 364 Phoenicians 359, 395 phratries 138–9, 140, 141, 142–3 and citizenship, admission to 143, 144–5, 147, 160–1, 163, 169, 235–6, 244, 246 expulsion from 145 membership of 160, 240

460

Index

phratries (cont.) political functions 160–1 as subdivision of tribes 165 Phye 282 phylarchs 11, 131, 151, 225, 226, 231 phylobasileis 147 Pindar 378–82, 412, 437 piracy 391 Pithecussae 363, 367, 368 pithos (jar) 341–6 Plataea 387–8, 422–3 Plato on colonization 367 on trade 354–5 Laws 10, 27, 44, 95–6, 268–9, 367 Republic 27, 217, 354 Pleistarchus 425 Pliny 62, 71, 116, 185, 351 Plutarch 29, 157, 192, 214, 219, 240 and Androtion 283, 297–301 on Cleisthenes 129 on Corcyra 350 on Council of Four Hundred 83, 84–5 on debt 110, 114, 120–1 on demes, names of 163 Hektemoroi in 120, 122 on Lampis 385 on Lelantine war 358 on medimnos 39–40 methods 298–9, 327 on Seisachtheia 110, 114, 120 see also Solon (Plutarch) polemarch 55, 56, 92–3, 221, 222–3, 226–7, 230–1 appointment by sortition 99, 100, 228 Herodotus on 227 status of 231 Poletai 9 politeia and nomoi 271–2, 273

of Solon 278, 310–17 political office, qualifications for 8–11 political parties, non-existence of 208 political rights, and land ownership 170–1 Politics (Aristotle) 62–3, 93, 161, 168, 169, 217, 218–19, 238, 243, 244 Athenaion Politeia, chronological relationship with 258, 266, 268, 272–3, 275–6 Athenaion Politeia, comparison with 103–4, 256–7, 273, 276–7 on change of constitution 68–9 date of 258–60, 267–8, 323–4 Four Hundred, revolution of the 273–4, 275 on ostracism 194–6, 206 sections of 258–60 Solonian laws 311–12 on time¯mata 68–9 Pollux 70, 157, 397 on share-cropping tenants 124, 125 on taxation 56–60 on tele¯, qualifications for 29–30, 56–7 Polybius 350–1 Polycrates 430 Population pressure 362, 367 statistics 387, 409, 414 port of trade 418 Poseideion 363 Poseidippus 13 pottery 341–2, 366–7, 383–4, 390, 404, 408–9 and see capacity prices 67 priesthoods 158–9, 161 primitivist/modernist 371–2 Prinz, H. 399–400 profit 42, 353–9

Index Pronapes 9, 10, 61 proportional representation 165 prostatai tou emporiou 399–404, 417 prostitutes 250 proxenia 141, 402–3, 418 prytany system 154, 167 Psammetichus 238 pseudo-Xenophon see Xenophon punishment, divine 429 Purcell N. 414 Pytheas 418 quadriremes 261–3, 324 Raubitschek A. E. 76–7, 181, 187, 190 registers, deme 242, 244, 250 military see katalogos religious associations see cult rent 111–12, 117 research see Aristotle retail see wholesale revolt of Helots c. 490 (?) 436, 439 revolution of the 400, 273–5, 276, 279, 293 Rhegium 352, 367 Rhodes (island) 402–3, 418 Rhodes, P. J. 72, 325–7 Rihll, T. E. 127 Robinson, E. S. G. 37, 398 Roebuck, C. 401 Rosivach, V. 72 Ross, W. D. 259, 260, 269 Roussel, D. 178 Ruschenbusch, E. 251–2, 288, 313–14 Salamis 379 Salaries 394 sale, cash 339 wholesale/retail 328, 330 Samos 356, 399, 400, 404, 422, 430 Sanctis, G. de see De Sanctis

461

Sandys, J. E. 174, 255, 258 Sardis 433 Schevill, F. 106 Scho¨mann, G. F. 94, 220 scorpion 372 Scythians 429 Schwahn, W. 234 sea-power, thalassocracy 354, 439 Seisachtheia 109, 114, 123–4, 289, 293 Aristotle on 110, 114, 118, 119–20 in Plutarch: Solon 110, 114, 115–16, 120 Se¯ko¯ma 344–6 Selinus 361, 395 Sellasia 434, 439 Seleucus 320 Selling see buying Seltman, C. T. 400 Sepeia, battle of 426–7, 433–4 sextarius 333 share-cropping 109, 113, 115–6, 120–6 and rent/debt 111–13 in Roman World 116–7 terms 116–17 ships 380, 389, 412 and see warships Sicels 359, 361–2 Sicily 356, 359–61, 364, 367, 395 Siewert, P. 177–8 Silver 353–4 and see metals as standard of value 38 Simplicius 258 Siphnos 382 Skolia 423 slaves and slavery 17, 60, 238, 241, 243 in Aegina 388, 409 children of 248–9, 250, 251–3 and debt 112, 114, 118, 119, 123–4

462

Index

snakes, winged 429 Solon chs. 1, 2 and 3 (1–128), 129, 286–9, 291, 310–17 amnesty law 79 Androtion on 288, 289 archonship 75–80 chronology 75–80 constitution 6–7, 9, 166, 273, 276, 310–17 Council of Four Hundred 83–9, 141, 142 eupatrid monopoly, destruction of 80–3 Horoi, removal of 120, 122 and Kle¯ro¯sis ek prokrito¯n 92–103, 276 laws 39, 79–80, 235, 241, 278, 313, 320–1 laws forged, interpolated etc. 321 laws no longer used 320–1 laws, survival of text of 310, 318, 327; and see axones magistrates, election of 89–104, 166, 273, 276 nomoi and politeia 310–17 nomothesia, date of 75–80 politeia see nomoi property classes as qualification for office, introduction of 80–1 reforms 125, 126–7, 286–9 and share-croppers 117, 123, 124, 125 sortition 89–104, 166, 218–22 and Voting 73–5 as one of Seven Wise Men 313 see also hektemoroi, Seisachtheia Solon (Plutarch) 45–6, 63, 79, 110, 113–16, 120–2, 287, 293, 295, 297–9, 301 sortition 89–108, 316–7

archons, in appointment of 215–23, 316–7 fixing of 99–100 in Florence 105–7 Isocrates on 101–3 reforms 215–20, 227–8 trittyes, allocation of 152, 156 Sosicrates 78–9 Sostratos 384–5, 413 Souk el-Khmis inscription 116 Spain 384 and see Emporion Sparta 15, 352, 421–40 and Samos 430 Sepeia, battle of 426–7, 433–4 Spence, I. G. 72 Spivey, N. 108 Stanley, P. V. 127 Stanton, G. R. 229 stasis 195, 201–3, 229–30 Stephanus of Byzantium 288 Stesichorus 437 Stesimbrotus 237 Stewart, J. A. 328 Strabo 351, 358, 361 on Aegina 386, 389–90, 417 strategos strategoi 11, 131, 151–2, 223, 225, 227, 230–2 election of 22–4 original function of 224–5 re-election 219, 223, 226–8 replace archons as supreme magistrates 215–28, 230–2 with specialized functions 260–2 Stroud, R. S. 326 Sumner, G. V. 185, 186, 197 Susa 425 Sutherland, C. H. V. 396 Sybaris 395 Symmories 282 Syracuse 204, 213, 352, 360–2, 365, 366, 395 Syria 364

Index Tacitus 351 Taras 352, 391 Tartessos 385 taxation 8, 44, 56–60, 334, 348, 394 direct/eisphora 57–9, 61 income 44 symmory system 59 taxiarchs 151, 225, 226 Tegea 435 Teisamenus 313–14 tele¯ (census classes) ch. 1 (1–72), 170, 317 membership of 8–11, 61–2, 68–9 military categories, equivalence with 19–24, 48–9 qualifications 28–32, 46–7, 51–6 see also Hippeis; Pentakosiomedimnos; Thetes; Zeugitai Tenea 361 Tenedos 391 thalassocracy see sea-power Thasos 341–3 Thebes 407, 422–4, 435 Themistocles 211, 227, 228, 237, 283, 313 and ostracism, introduction of 190, 191–2 ostracism of 198, 201, 205, 207 Theognis 81 Theophrastus 181, 266, 278, 305–6, 318 Theopompus 279 Theramenes 283 Thermopylae 425 Thesmothetai 100, 221 Thessaly 398–9, 407, 429, 437, 439 Thetes 7, 8, 13, 14, 28–9, 55, 70–1, 126, 167 Council of five Hundred, admission to 167 equivalence with ‘those below hoplite class’ 15, 18–22, 23

463

Pollux on 30, 56–7 Thirty, the 314–5 Thompson, W. E. 156, 173 Thomson, G. 36–7 Thrasybulus 314 Thucydides son of Melesias 198, 201, 205, 207, 237, 379 Thucydides the historian 134, 174, 196, 237, 279, 313, 359, 361, 397, 422, 431, 436 on Aegina 377, 407 on trade 355, 359 Four Hundred, revolution of 273–4, 275, 276 on Athenian military organization 13–14, 18, 19, 20–2, 213 Thyrea 377 timocracy 6 time¯ma 10, 25–7, 39, 48–9, 60–2, 69 and see tele¯ Timotheus 239 Tiryns 434 Tissaphernes 398 Tod, M. N. 89 trade, traders ch. 10 (371–420) passim in classical period 368 as cause of colonisation 352, 359–67 and coinage 392, 393–4, 396 commercial exchange 353, 356, 363 exports 353–4, 355, 356, 358, 359, 366 imports 353–5, 356, 358–9, 361–3 wholesale, large-scale 373, 375 trading colonies 359–67 treasurers of Athena 27, 42, 55–6 census class membership and 8–10 selection 90, 91

464

Index

treaties 355 tribes: establishment of 130–1 reforms 130–1, 132, 138–40, 142–4, 147, 149, 151–7, 160, 162, 165, 173–8 subdivisions of 165 trittyes 132, 156–8, 162–3 arrangement of 137, 138 assignment to tribes 148–57, 161–3, 177–8 heterogeneity of 154–5 in Cleisthenes’ constitution 148–57, 161–4 and sortition 152, 156 as subdivision of tribes 165 tyranny 211–12 Athenian 140, 150, 153–4, 423–4, 431

‘commercial rivalry’ 356 over boundaries and land-ownership 356 warships 380, 414 Wayte, W. 66 weights and measures, see metrology Weil, R. 255, 257, 259, 260, 267–8, 269–70 Weinberg, S. S. 383 Welter, G. 373, 374–5, 376, 409 wheat 33–6, 66, 67, 333, 356 Whitehead, D. 72 wholesale/retail 328, 330 Wilamowitz-Moellendorff, U. von 193, 277, 289, 296–7 Wilcken, U. 41, 45–6, 67–8 wills 351 wine 66–7, 328, 331–2, 337, 429 Winterscheidt, H. 373, 376, 388

Ure, P. N. 409 Vallet, G. 366 van Groningen, B. A. 46, 67–8 Vidal-Naquet, P. 129 Villard, F. 366 Vinogradoff, P. 113 Vita Menagiana 317 voting procedures 107–8 majority 73–5

Xanthippus 194, 198, 206 xenia (hospitality) 141, 380–2, 412 Xenoi, xenai (foreigners) 241–2, 247 Xenophon 14, 25, 279, 314, 325, 355 pseudo-Xenophon (‘Old Oligarch’) 354

Wade-Gery, H. T. 6, 84, 101, 129, 221, 228, 237, 278, 378 on selection of polemarch 92–3 on sortition 96–7 wages 43 Waggoner, N. M. 414–15 Walker, E. M. 150–1, 193, 233 Wallace, R. W. 108 Walters, K. R. 252 wars, origins of 356

Zancle 367, 395, 436 Zeugitai 7, 8, 29, 55, 126 and archonship 28, 217 equivalence with hoplites 19–20, 21–2, 23–4, 25, 26–7, 50–1 original qualifications for 30, 46–51, 54 Pollux on 29, 30, 56–7 taxation 57–8