American Journal of Ancient History (Vol 4.1) 9781463206680, 1463206682

Citation preview

American Journal of Ancient History

American Journal of Ancient History

4.1

The American Journal of Ancient History is a peer-reviewed academic journal covering ancient history and classical studies. It was established in 1976 and edited by Ernst Badian until 2001. It is continued by the American Journal of Ancient History: New Series, edited by T. Corey Brennan.

American Journal of Ancient History

Volume 4.1 Edited by

Ernst Badian

gp 2017

Gorgias Press LLC, 954 River Road, Piscataway, NJ, 08854, USA www.gorgiaspress.com Copyright © 2017 by Gorgias Press LLC Originally published in 1979 All rights reserved under International and Pan-American Copyright Conventions. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise without the prior written permission of Gorgias Press LLC. ‫ܐ‬

1

2017

ISBN 978-1-4632-0668-0

Printed in the United States of America

TABLE

OF

CONTENTS

E. Fredricksmeyer:Three Notes on Alexander's Deification ...............

June W. Allison: Thucydidesand HOAYHPAl'MOI;¾NH

..............

1

10

K.H. Kinzl: Betrachtungen zur )•lterenGriechischen Tyrannis............

23

Roger S. Bagnail: The Date of the Foundation of Alexandria ...........

46

John Buckler: The Re-establishmentof the Boiotarchia (378 BC) .........

50

Frank

65

M.

Clover.

Count

Gainas and Count

Sebastian

..................

Catherine Reid Rubincam: Qualification of Numerals in Thucydides......

77

TABLE

OF

CONTENTS

E. Fredricksmeyer:Three Notes on Alexander's Deification ...............

June W. Allison: Thucydidesand HOAYHPAl'MOI;¾NH

..............

1

10

K.H. Kinzl: Betrachtungen zur )•lterenGriechischen Tyrannis............

23

Roger S. Bagnail: The Date of the Foundation of Alexandria ...........

46

John Buckler: The Re-establishmentof the Boiotarchia (378 BC) .........

50

Frank

65

M.

Clover.

Count

Gainas and Count

Sebastian

..................

Catherine Reid Rubincam: Qualification of Numerals in Thucydides......

77

THREE

NOTES

ON ALEXANDER'S

DEIFICATION

Pausaniasreportsin his descriptionof Megalopolis(8.32.1):

Christian Habicht in his Gottmenschentumund griechischeStiidte, the classic modern treatment of the subject, notes: "In Megalopolis sah Pausanias(8,32,1) ein ursprfinglichffir AlexandererbautesHaus. Es ist ungewiss, ob es ein Kultgeb•iude gewesen war (so Boehringer-Krauss, Alterttimer von PergamonIX (1937) 89, zweifelndNilsson,Gr. Rel. II 140, Anm. 4). "2 E. Boehringerand F. Krauss(loc. cit.) refer to examplesof sacredbuildingswhoseconstructionallowedan easyconversionto private use,and assumewithout argumentthat the building noted by Pausanias wasoriginallya sanctuaryof Alexander.3 M.P. Nilsson,on the otherhand, says(1oc.cit.): "Ob man aus Paus. VIII 32, 1 rrX•q•o• ... •rro•q• (in Megalopolis)auf einenKult schliessen darf, ist zweifelhaft.TM We can agree. But Nilsson failed to note the immediately following information about the statue of Ammon (above). And this makesa difference.5 We know quitea bit about the associationbetweenAlexander and Ammon. In January 331 13cthe oracle of Ammon at Siwa, or the high priest, proclaimed Alexander as the god's son. Alexander acceptedthe filiation with alacrity and hoped to have it acknowledgedas widely as possiblein the Greekworld.6Now, if we hear of a buildingat Megalopolis whichwas"built for Alexander",and by that buildingwefind an "imageof Ammon", it is reasonable to surmise that it was a shrine dedicated to Alexander as son of Ammon.7 A priori, it could also have been a shrine

dedicatedby Alexanderto Ammon, but there is no evidencewhateverof any cults of Ammon founded by Alexander anywhere, while cults for Alexander are attested in various places.8 ¸ 1980by E. Badian. All rights reserved.

2

E. FREDRICKSMEYER

As for the date of the cult, thereismuchto support,and nothingagainst, the hypothesisthat it dated from Alexander'slifetime. Although Megalopolis did not participate in the anti-Macedonian uprising after Alexander's death in 323 BC, the Macedonian regent Antipater was opposedto his cults?and after Antipater'sdeathin 319 BCuntil probably 303 BCMegalopoliswasunderthe protectionof Antipater'ssonCassander, who hated Alexander.•0 But we do not know of any occasionat any time after Alexander'sdeath whenanyoneat Megalopolismighthavewishedto proposea cult for Alexander.And, as far as we know, a posthumouscult for Alexander would have been unique in mainland Greece. On the other hand, during Alexander'slifetime we can identifytwo occasions on whichit wasverymuchin theinterestof the Megalopolitans to institutea cult for Alexander.The terminuspost quemis his proclamation as sonof Ammonat SiwaaboutJanuary331 BC.It istrue that Ammon was commonly identified with Zeus, and there is evidencethat Alexander thoughtof himselfasthe sonof Zeusevenbeforehis pilgrimageto Siwa.• But the image seen by Pausanias,with the ram's horns, is distinctly Ammon, and identified by him as Ammon, not Zeus. The relationship betweenAlexander and Ammon originatedwith Alexander'svisit to Siwa. However, we know of no cults of Alexander which were founded in

reaction to his proclamationas Ammon's son about January 331 BC. Alexander'scultsin the Greekcitiesof Asia Minor probablyhad already been instituted in 334/3 Bc, as an expressionof gratitude for their liberation from the Persiandomination.12If they were foundedat some later time during Alexander'scareer,thereis no evidencethat the occasion was the proclamationat Siwa. Now Megalopolis,againstthe perennial threat of Sparta, had beena constantally of the Macedonians,first of Philip (since352 BC)and then of Alexander.•3 After Antipater'svictory over Agis III sometime betweenthe late fall of 331 and the springof 330 BC,TM the councilof the HellenicLeagueimposedon the Achaeansand the Eleans,who had sidedwith the Lacedaemonians, a penaltyof onehundred and twenty talents,to be paid to Megalopolis(Aeschin.Ctes.165;Curt. 6.1.17-21). It is possiblethat on this occasionthe Megalopolitans,in gratitudeto Alexanderasthepresidentof theLeague,established a cultfor him as the son of Ammon.

Onemightobserve,contra,that Alexanderwasmovingeverdeeperinto Asiaand wasnot likely to returnsoon,if ever,whileAntipaterremainedin formidable proximity.•5 It was Antipater to whom the Megalopolitans owedtheir mostimmediategratitude.It washe who had savedthecity by lifting Agis' siege,engaginghim in battle, and defeatingand killing him. And it was Antipater who summonedthe Council which awarded the Megalopolitansthe hundredand twentytalents.Antipater,wehavenoted, was not in favor of payingdivine honorsto mortals.•6On the other hand,

ALEXANDER'S

DEIFICATION

3

Antipaterwouldnothavevoicedhisopposition exceptin private,and whatever munificence hemightbestow ona GreekcityduringAlexander's absence, hedidsonotinpropria persona butonlyasAlexander's agent. Everyone knewwhotherealboss was.ThusanAlexander cultat Megalopolisin the aftermathof the defeatanddeathof Agisis a definite possibility.

Buta morelikelyoccasion camein 324BC.This,however, ispartof

another problem.

We have evidenceof discussions at Athensand at Sparta in the fall and winter of 324 BCon the proposalto accordAlexanderdivine honors.l? If the proposalwas madeat Athensand Sparta, it is reasonableto speculate that it wasmadealsoin othercities.This muchis generallyacknowledged. The controversysetsin with the questionwhetherin the eventcultswerein fact instituted in those cities, and whether the initiative came from them or

from Alexanderhimself.I will not discussherethe wholevexedproblem, but focus on one crucial item of evidence. ms

Arrian says(7.23.2) that, in the springof 323 BCat Babylon,

The crux of the passagehasbeenthe &•. Somescholarshavetaken it in

the senseof "as",that is, "in the capacityof", and considerthepassage as evidencefor Alexandercultsin Greeceat thistime, whileothershavetaken it in the senseof "asif (but not really)"andthereforerejectthepassage as evidence for Alexandercultsin Greecein 323BC.Thereisnoneed,or space, here to rehearsethe variousargumentspro and contra which have been advanced.They are inconclusive,and the questionremainsopen.19Fortunately,however,the disputecanbeconclusively settledby investigation

of thewayArrianactuallyusesthephrase&•... 8•0ev.20 Apart from 7.23.2, it occursalsoat 4.18.4, 6.13.2,7.8.2, and Ind. 15.9.2] In eachcasethe phraseconveysan ambivalence.From the perspective of the present it indicateswhat was then consideredas real or true, but subsequentlywas shown to be not real or true. There is of course no

equivalentexpression in Englishto conveythisambivalence, but it canbe renderedadequatelyby adding to the verbs"consider","act", or the like,

the qualifyingremark"mistakenly(or wrongly)as it turnedout". In the

4

E. FREDRICKSMEYER

following translationsI will in each case,for greater clarity, italicize the word or expressionenclosedand marked by &g ... 8•qOe¾. 4.18.4. Oxyartes'wife and daughtershad taken refugeon the Sogdiana rock,

6';'r•xO•F•vou , 6•:•x•x•,•6•:6• Acp•u•'r•xe• &•' "Oxyartes having put them there for safetyat the time of his own rebellion againstAlexander, thinking (mistakenly,as it turned out) that the place was impregnable." 6.13.2. Alexander'ssoldiersthought that he had died of his wound. As the ship with Alexander reclining on board approached the camp, Alexander ordered the awning to be removedfrom the stern so that all could see him.

"But they still disbelievedit, thinking (mistakenly,as it turned out) that it was Alexander's corpse that was being conveyed,until at length ... Alexander lifted his hand toward the crowd. But they raised a shout ....

"

7.8.2. Intending to dischargehis superannuatedand disabledveterans, Alexander made a speechpraising them.

"Alexander said thesethingsno doubt (but mistakenly,as it turned out) topleasethe Macedonians,but they,thinkingtheyweredespised by Alexander, not unreasonablyfor their part were annoyed with what he said."

Ind. 15.9. In this passagethe phraseis usedin a negativeconstruction and therefore requires a slight adaptation in the translation. Nearchus treats parrots as marvels.

ALEXANDER'S

DEIFICATION

5

"But I, becauseI myselfhaveseenmany of them and alsoknow others who have an understanding of the bird, will not discussthem (mistakenlyas it would turn out) as anythingremarkable." Now our passage,7.23.2: "And there arrived at this time embassiesfrom Greece,and their ambassadors,themselvescrowned, approachedAlexanderandcrownedhim with goldencrowns,assacredenvoys(mistakenly,

as it turned out) who had cometo honora god." And Arrian proceeds immediatelyto showwhy theyweremistaken.The supposed god not long thereafter

died.

Thus our passagemustbe taken to meanthat theenvoyson the occasion honoredAlexander in the capacityof "sacredenvoyscomein honor of a god", that is, they paid him divine honors. If this is correct,the passageprovidesevidencethat as the resultof the discussions in late 324 BCAlexandercults werefoundedin (at leastsome) citiesof Greeceand that envoysfrom thesecitiesappearedbefore Alexanderin the springof 323 BCto honor him accordingly.We may note,in addition,that if thecitiesin questionwereonlyAthensand Sparta,thetwo most famous cities in Greece, the one having been vacillating and ambivalent in her attitude to Alexander, the other persistentlyhostile,

Arrianprobablywouldhavenamedthem.Thevagueexpression rr?e•eZ• •x r• •EXX•8o• suggestsan indefinite number of cities, possiblythe member states of the Hellenic League, or some of them. The nearly simultaneousinstitution of Alexander cults in severalor more Greek cities,

in turn, presupposes a unanimity of attitude, planning, and action that would be most uncharacteristicallyGreek and in fact inexplicablein a purely practical sense,unlessthe cities were acting under a strongconstraint,and this can only havecomefrom Alexanderhimself,eitherby an outright requestor in someother, more diplomaticbut still unmistakable way. Someinitiativefrom Alexander,in whateverform, maybeconsidered as certain. 22

We have seenthat Megalopolis,of all Greekcitiesthe most loyal, may haveinstituteda cult for Alexanderin 331 or 330 BC.If not, wemaybe sure that she was now, in late 324 or early 323 BC, one of the cities which accorded

Alexander

divine

honors. III

In the last yearsof his life Alexandercherishedand propagatedhisdivine filiation with undiminishedand evenincreasingfervor.23It is reasonableto speculatethat he hopedto be honoredofficiallyasthe sonof ZeusAmmon

6

E. FREDRICKSMEYER

by the citiesof Greece,and in 324 BChe may havesenttheminstructionsto this effect. At any rate, his preoccupationwith his divine sonshipwas a well-knownfact. An appealto this relationship,sanctionedasit wasby the god'sown authority at Siwa,couldfacilitatefor hissupportersin the Greek citiesa proposalfor deification.2n We have seen reason to think that at Megalopolis Alexander was honored as son of Ammon. There is reason to think, as well, that Alexander was honored similarly at Athens. In his speechagainstDemosthenesin 323 BC Hyperides discussesDemosthenes'role in the Harpalus scandal of 324 BC:

"When you thought that the Areopaguswould report thosewho had the gold you becamehostileandcreateda disturbancein thecity soas to obstruct the inquiry. But when the Areopaguspostponedits statementon the groundsthat it had not yet discoveredthe truth,

•

[,,• • •o6X]o•o, [•]• •p•[•o•,,o•,] .... [Desunt col 31 versusfere decem.]

col. 32: ......

oa'rouq......

•[3o6•.e-r[o ......

] a-rqa•[

Despite the mutilated state of the latter part of the passageit seems reasonably clear that Demosthenes' statement concerns the proposal (probably by Demades)of Alexander'sdeification in 324 BC,and it also seemsclear from the contextthat Hyperidesis quoting,not somechance remark by Demosthenesduring the course of the discussionsof the deification proposal,but an announcementwhich expresseda changein Demosthenes'attitude, or at least which under the circumstancescame as a

surprise.Now Dinarchus, in his speechagainst Demosthenesin 323 BC, criticizedDemosthenesfor hisinconsistencies and citesexempligratia that, in the matter of Alexander'sdeification, he first opposedbut later supported the proposal.26It is thereforereasonableto infer that Hyperidesin his speechhas quoted the words of Demosthenes'announcementof support for Alexander'sdeification. There is no reason to doubt the authenticity of the statement. Demosthenesmade it in publicand Hyperidesnot long afterwardsquoted it in public.Thereweretoo manywitnesses and it wastoo idiosyncraticto be made up. The sarcastic"and of Posidontoo if he wishes"is obviously Demosthenes'attempt to savefacewith the peoplein viewof histurn-about in the matter? But the words"Let him be the sonof Zeus"suggest,since the god had already long ago acknowledgedAlexander as his son, that the proposal was to acknowledgeAlexander officially as the son of Zeus

ALEXANDER'S

DEIFICATION

7

Ammon, and sincethe issueat hand was Alexander's deification, there is a

very real possibilitythat the proposalwas the deificationof Alexanderas son of Zeus Ammon. 28

So much for Athensand Megalopolis.It isjust possiblethat other cities of Greece honored Alexander likewise. Lucian, with reference to

and thereforenot individualsbut communities,says(DMort 13.2):

veGCx0d06ovveC&C8p0•xovvoc [Ammon] Die.29 University of Colorado

E. Fredricksmeyer NOTES

1. The text is that of M.H. Rocha-Pereira,vol. II (Teubner 1977).The app. crit. gives:8] r6 F om. L del. Schubart-Walz.I have corrected(I),.Xt•to,. to (I)tMrcrmu,

apparently a typographical errordueto theauthor'suse(throughout) of theiota adscriptum instead of the subscripture.So also at 1.25.7 (p. 57: 1.14) she has A'•tz•'rp•t 8• •. 'Awt¾6vo•.insteadof 'Av•.¾6•ou. 2. (Munich 1970)2 29 n. 3. 3. Cf. F. Taeger, Charisma I (Stuttgart 1957) 217, and F. Schachermeyr, Alexander der Grosse(Vienna 1973) 529 n. 336. 4. Also at Geschichteder griechischenReligion II (Munich 1961)2 147 n. 12. 5. I do not mean to suggestthat no one has noted the connection,but that Nilsson'sfailure to do somightexplainhisscepticism aboutthebuilding.H. Hitzig and H. BIQmer,Des PausaniasBeschreibung yon GriechenlandIII (Leipzig 1907) 231, notead loc.:"Die HermedesZeusAmmonbeidemursprtinglichAlexanderd. 6r. geweihtenBau erkl/irt sichdaraus,dassAlexandersichbekanntlichgem als Sohn des Ammon verehren liess, wie ihn denn auch MQnzen der Ptolemaier mit

Widderh6rnernabbilden."The earliestknown depictionsof Alexanderwith the horn of Ammonare on coinsissuedby PtolemyI about318 Bcandby Lysimachus in 296 BC.(See A.R. Bellinger,Essayson the coinageof Alexander the Great. NumismaticStudies11 (New York 1963)86 f.) Thesedepictionsmay well be based on contemporaryprototypes.Cf. R. Hinks, Greekand Romanportrait sculpture (London 1935)9. SeealsoClem. Alex. Protr. 4.54.2,and Ephippus,FGrHist 126F 5=Athen. 12.537E.However,we shouldnotethat the statueseenby Pausaniaswas that of Ammon,not oneof Alexander.J.6. Frazerin hiscommentaryon Pausanias (London2 1913),ad loc. (IV 349), has nothingto say on the matter. 6. On Alexander'svisit to Siwa, see Callisthenes,FGrHist 124 F14a= Strabo 17.C p. 814; Arr. 3.3-4; Diod. 17.49.2-51.4;Plut. Alex. 26.6-27;Curt. 4.7.6-32;Just. 11.11.2-12. On his attitude to Ammon, see Callisthenes,FGrHist 124 F36=Plut. Alex. 33.1; Arr. 4.9.9; 7.8.3; 7.29.3; Plut. Alex. 28.1; 50.6; Mor. 339E; Curt. 4.7.2830; 6.9.18; 8.5.5; Satyrus,FHG III 164 F18=Athen. 6.250F. Cf. P.A. Brunt, Arrian. History of Alexander and Indica, LCL, I (1976) App. V, and A.B. Bosworth, "Alexander and Ammon", in Greece and the eastern Mediterranean in ancient

8

E. FREDRICKSMEYER

historyandprehistory.Studiespresentedto Fritz Schachermeyr on theoccasionof his eightieth birthday, edited by K.H. Kinzl (Berlin and New York 1977) 51 if. 7. Cf. -r&Xe[0•¾gpe{ou at Priene, Inschr. Prien. 108.75,with commentary. 8. Habicht (above n. 2 and text) 17-36, 245-52.

9. Suda,s.v.'A¾'r•rr0•rpo•: p.6vo• •k •&• •6Z•

0•6• x•X•

AX•po•

o6Z

10. Diod. 18.49,54-55, 68, 74; F. Staehelin,RE 10(1917) 2293 if.; H. Bede, Das AlexanderreichaufprosopographischerGrundlageII (Munich 1926)201 f. It was very possibly during the period of Cassander'sascendancythat the oikia was deconsecrated.

11. Bosworth (above n. 6) 68 if. 12. Habicht (above n. 2 and text) 17 if. and 245 f. Habicht's dating remains attractive becauseis providesa strongmotivation for the cults. 13. Von Hiller, RE 15 (1932) 132if.; E. Meyer, Der Kleine Pau½ 3 (1969) 1141. 14. The dateof the battleof Megalopolisis disputed,but thisdoesnot affectour argument. See,most recently,R.A. Lock, "The date of Agis III's war in Greece", Antichthon 6 (1972) 10-27 (for a late date), and E.N. Borza, "The end of Agis' revolt", CP 66 (1971) 230-35 (for the traditional early date of late September). 15. Cf. Aeschin. Oes. 165, speakingshortly after the battle at Megalopolis: "Alexander had withdrawn to the uttermostregionsof the North, almost beyond the borders

16. Above

of the oikoumene."

n. 9.

17. Din. Dem. 94; Hyp. Dem. 31-32; Timaeus, FGrHist 566 F155=Polyb. 12.12b.3; Plut. Mor. 219E-F; 804B; 842D; Ael. VH2.19; 5.12; Athen. 6.251B; Val. Max. 7.2. ext. 13;D.L. 6.63. On thechronology,seeE. Badian,"Harpalus",JHS81 (1961) 4143. 18. Hyp. Epit. 21 constitutesat bestequivocalevidence.It could be taken to mean that a cult of Alexander was instituted at Athens and elsewhere, and abolished after his death. But it may also mean, instead, that at the time of the deliveryof the speechin the hll of 322 •c cultsof Alexander(still) existedin Asia Minor. It shouldbe noted,however,that the secondinterpretation(whichperhaps shouldbe preferred)doesnot excludethe possibilitythat cultswere institutedat Athens and elsewherein Greecein 324/3 •c and abolishedafter his death. Seethe discussions of the passageby Habicht(aboven. 2 and text) 28-30and 24647, and E. Bickerman,"Sur un passaged'Hyp•ride (Epitaphioscol. VIII)", Athenaeum41 (1963) 70-85. 19. See most recently the argumentsof Bickerman 75 ("as iff) and Habicht (above n. 2 and text) 247 f. ("in the capacityof'D. 20. I am gratefulto ProfessorEugeneN. Borzafor helpingme completethe set of references that I had collectedfor thisusage,althoughhisinterpretationof them may well differ from mine. 21. Arrian uses$q0ev without &{ at 7.14.5 and at Ind. 27.10, but thesetwo instancesdo not affect our question. 22. •. Ael. VH 2.19; Plut. Mor. 219E; Hyp. Dem. 31-32; Din. Dem. 94. 23. Arr. 4.9.4;7.8.3;7.29.3;Ephippus,FGrHist 126F5=Athen.12.538A-B; Plut. Alex. 28; Diod. 18.3.5;Cure. 10.5.4;Just. 12.15.7;C. Seltman, Greek coins(London 1955)2 213.

ALEXANDER'S

DEIFICATION

9

24. Cf. Arr. 4.9.9; Curt. 4.7.28; Just. 11.11.11. 25. Hyp. Dem. col. 31. The text and translation are by J.O. Burtt, Minor Attic orators, LCL, II (1954).

26. Dem.94:x• 'ro'r•p.•vypd•covx• •rr•yop½6tav p.'q•v•vo•½•v •ov

0½•v•

•7. The quip about Posidon perhapscan bc explainedby referenceto Aulus Ocllius(15.• 1):ferocissimoset iramaneset alienosab omnihumanitaw,tamquame mari genitos, Neptuni flios dixerunt. Cf. J.P.V.D. Balsdon, "The 'divinity' of Alexander", Historia 1 (1950) 385 n. 119. •8. •is of coursedid not preventtheadditionalproposalandpossibleadoption of specialcult titlessuchas,at A•hcns,possibly,ofAnik2tos •eos (Hyp. Dem. 3•) and/or of •eos

Triskaidekatos (Acl. VH 5.12).

•9. I •sh to expressmy sincereappreciation•o the Journal'sanonymous referees,and especiallyto the •itor, throughwhoseobservations and suggestions this paper has bccn greatly improved. Any remainingfaults arc solelymy own responsibility.

THUCYDIDES

AND

HOAYHPAFM0•YNH*

In his book, Thucydides,J.H. Finley summarizesthe character of the

Athenianswith theword•o3.v•p•¾•oa6v3.1Five yearslater,V. Ehrenberg traced the role playedby •o3.v•p0•¾•oa6v3 in Athenianforeignpolicy throughthe fifth centuryinto the fourth.2 Most recently,A.W.H. Adkins has analyzedmuch of the sameevidencefrom what he characterizesas a 'socio-political'perspective in orderto uncoverthe ethicalconnotationsof the word amongthe Athenians.3 Sinceit is likely that theselasttwo studies will continueto be viewedasmajor sourcesin the elucidationof the concept

polypragmosyng,, it is important to understandthe limitations of our enlightenment.In dealingwith political contextsboth authorsnecessarily considerThucydides'History, where the conceptseemsto them to find clearexpression.In thisdiscussion I shallquestionsomeof the conclusions drawn by Ehrenbergand Adkins regardingthe significanceof•o•.•:O0• Y•oa6v• in the History. My own conclusions will emergefrom both an evaluationof a particularmethodof argumentationusedby scholarsand a reevaluationof the History• value as evidencefor the importanceof the concept,taking into considerationthe virtual absenceof the word 'polypragmosyn•'. 4 The term occursexactlyoncein the entirework, at VI 87.3. Whetheronetakesit to expressa singleconceptor a family of concepts,the frequency of its occurrencemust in part determine the direction of argumentsstructuredtowards proving its significance. The method of argumentationemployedby Ehrenbergand followedby Adkins is one basedon assumptions:they have presented'rational reconstructions'of theconnotationsof the word 'polypragmosyn6' by examining those qualities in the History which seemto be identical with those the word characteristically expresses. 5 Rational reconstruction is a legitimate method of arguing a thesisfor which there are no verifiable data in the source.This kind of argumentation,in the absenceof data, mustfirst of all be consistentwith what the source claims, since it is not objectively demonstrable,but only interpretive.6 Secondly,it follows that the argument doesnot containany omissionsof data even--or especially--ifthey contradictthe argument.In the caseof •o3.•f•0•y•o•6v• both Ehrenberg 10

THUCYDIDES

AND

IIOA¾IIPAI"M0•¾NH

11

and Adkinsseemto satisfythe first criterion:their argumentsare consistent with their interpretationsof Thucydides.In respectto thesecondcriterion, however, it would seem that the omission of the fact that the word

'polypragmosyn•'appearsonly once (VI 87.3) vitiates their attemptsto reconstruct the significance of the concept polypragmosyng from Thucydides.Both scholarshave,therefore,left out an importantpieceof verifiableevidencewhich,whenincluded,is foundto beincompatiblewith their arguments.7 Ehrenbergsaysat the outsetthat there can be no middle term between •oXurr0•y•zo•;6¾ • and &rr0•y•zo•;6¾•. 8 Adkins refersto this observation when he says,'to decrypolupragmosuneis not necessarilyto commend apragmosune'and vice versa,but what we haveis 'an Aristoteliantriad, a virtue betweentwo vices,polupragmosune-arete-apragmosune'. 9The neatnessof the polarity which Adkins and Ehrenbergfind so compellingis upset when we realize that, except for the one instanceof the term 'polypragmosyn•',Thucydidesusesonly negativeforms: 'apragmSn'and 'apragmosyn•'.But the assumptionunderlyingthe establishmentof the polarity goesundetected,viz. that the appearanceof thesenegativeprefixedformspermitstalk of the word'polypragmosyn•'. In thelight of this, the most obvious questiongoesunasked by Ehrenbergand Adkins: why does Thucydides,who so frequently summarizeseventsand policy by abstractnouns,use 'apragmosyn•'with a negativeinsteadof 'polypragmosyn•'?It is extremelydifficult to arguefor the importanceof a term which is scarcelyin evidence. Rather, it seemsto make greater senseto examinethe possibilitythat the fact that this word only appearsonceis instructivein determiningits connotationsas an expressionof a concept. Adkins (seebelow) and K. Kleve considerthe passagein which the term occurs, yet fail to note that it occurs only there.•0 P. Huart and A.G. Woodhead both note that the word occursonly once in Thucydides;the former finds no significancein the fact and the latter arguesthat the word has positiveconnotationsin VI 87 and in assessing Pericles'policy.TM Neither Ehrenbergnor Finley dealswith the passageor with the frequency of the word.

We can now turn to the ways in which noting the uniquenessof the word's appearanceaffectsmuch of what has been acceptedby scholars about the concept.While Ehrenbergsaysthat in the History all condemn &rr0•y•zooG¾•, he also claims of rro•.urr0•y•zo•;6¾ • that the Athenians 'proclaim it as a principleof foreignpolicy'.12Two basicerrors,of which Ehrenbergand othersare apparently not aware, underlie this statement: one, a tendencyto conflatereferenceto the conceptwith referenceto the word; and two, an assumptionthat 'ouk apragmSn'implies not only polypragm6n but also 'polypragmfn'.13It is evident, when one is made aware of the conflation,that neither Athens'opponentsnor the major

12

JUNE

W. ALLISON

proponentsof empire in Athens 'proclaim it'. In connectionwith the analysisof the Athenians by one of theseantagonists,the Corinthians, Ehrenbergsaysthat the speech'doesnot mentionrcXzo,z• but is a setof variantson the themeof r•oXur•O0t¾•toa6,• qand howit isneversatisfied'.TM It is not significantthat the word 'pleonexia'doesnot appear when it is not worthy of note that 'polypragmosyn•'is also lacking. If it is not the word 'pleonexia' to which Ehrenbergrefers, then he has failed to provide an argumentfor the absenceof the themepleonexiawith the samefervor with which he hasarguedfor the presenceof the themepolypragmosyng.While it is possibleto talk about a family of conceptswhich may fall under the generic rubric 'polypragmosyn•', one must also deal with the actual appearanceof the rubric in the text. We are perhapstoo easily given to argumentsbasedon key wordsor on what we take to be slogans.This seems to be the casewhen we considerEhrenberg'sremarks on the Corinthian speech:'If we try to think of an outstandingdescriptionof the r•O0•¾•to¾, the first passagewhich comes to mind will probably be Thucydides'descriptionof the Athenian character.'•sInasmuchas neither 'polypragmosyn•'nor indeed'polypragm6n'occurshere,this is not the first passagethat comes to this reader's mind. There is no indication that Thucydideswould haveconsideredthe qualitiesof the Atheniansoutlined by the Corinthiansas demonstrations of •:oXu•O•zy•to•,• • simplybecause they disavow Statementsattributed to the Atheniansthemselves,includingthosewith whom Thucydidesassociatesmajor policy decisions,both foreign and internal, provide no further help. Alcibiades, whose character might be mosteasilyconnectedwith what we think •oXuOdy•to•,•means,usesthe negativeforms three times in the speechurging the Sicilian expedition; clearly he condemnsthe 0•r•O0ty•o•6,• q of Nicias, and he statesrather emphaticallythat the city whichis not 0•r•O0•y•.o•,• wouldbe destroyedby a changeto 0•r•O0ty•o•6,• q (VI 18.6-7). He doesnot poseas an alternativea policy of rmXu•O0ty•.o•6¾• q. Nowhereare any formsof the term appliedto Cleon, so that the two most conspicuousadvocatesof expansion and aggression do notfor somereasonwarranttheattributermXur•O0•y•o•¾. But the qualities they do exhibit in the History have elicited the label from scholars,as if it were also Thucydides'tag: Ehrenbergsaysthat audacity, personalambition, and the 'ficklenessof the masses'are 'expressedby r•oXu•O0ty•.ooG¾•q', to which he adds: 'It is not the only term used by Thucydidesto coverthat quality of characterwhichwasat the bottom of Athenian imperialism.'•6It isclearthat Ehrenberghasassumeda definition of the conceptand his argumentsare presentedas if they resultedfrom an investigationof the use of the word in Thucydides.

Thucydideshas includedin his portrait of Alcibiadessentiments expressedearlier in the History by Pericles.Periclestoo tried to engender

THUCYDIDES

AND

HOAYHPAPMOY.

YNH

13

confidencein the Athenians through exhorting them to avoid

?.o•6•, andindeedhecondemns asuseless themanwhois&rrpdyy. to,;and hetoofailsto offerrroXurO•¾y. omS,•eitherasa privatemodeoflivingoras a public policy. Sevenof the elevenoccurrences of &rrp0•yy. to, and &rrf•0•y?.oa6,• comein contexts connected withAlcibiades andPericles. In noneof the passages (II 40.2;63.2,3; 64.4;VI 18.6,7 twice)are we given licenceto inferrroXurO0tygzoa6,• simplybecause itsantithesis occurs.The Corcyreansuse&rrO•¾?.oa6,• to describetheir formerforeignpolicy(I 32.5);Hermocrates suggests to theSiciliansthattheirownquarrelswould cometo nothing(&rrp0•yy.6,toq) if theycametogetherin peace(IV 61.7). And it is onceusedby the Corinthians(I 70.8) of the Atheniansjust as Pericleshad usedit. The last occurrencecomesimmediatelyafter the sole appearance of thenounrroXurp0•y?.oa6,• at V187.3-4.Mostof theusesare, therefore,found in the speeches of Alcibiadesand Pericles,and the concept expressed by the word 'apragmosyn•'is not widespreadin the History. In the caseof Periclesthe passageat II 64.4 is particularlydeservingof notice.Here Periclescontrasts the&rrOdyy. to,withthemanwhohasthewill to act (8• zt). Finley'scommentis apposite: This extraordinary passagesetsforth the full implicationsof the doctrineof rcoX,rc?•¾y.o•;6,•, the doctrinethat the capacityand the will to do forever causeand justify change.... To Thucydides,this will to do, which is thus identified with the Athenian democracy, undoubtedlyseemedfundamental to history.•7

While the 'will to do' and the abilityto act uponsoundpolicy(¾•t3•z•)with vigor may indeed be characteristicof the Athenians at this time, it is necessaryto point out what now might seemobvious,namelywhat appears

to be a careful periphrasis:6 •z• &rr•¾•zto•... 6 8k 8• zt x•[ a6z6• ' 18 •3o,Xoy.•oq. When Thucydidesdoesthis kind of thing in antithesis,it is not withoutsomepoint. Althoughthereisno nominalform between&rrp0•¾•zoo%• and rroX,rrpot¾•zoo6,• derivedfrom the sameverbal stem (see abovep. 11),it appearsthat for Thucydides thewordrroX,rrootyy. oo6•• did not carry the positive connotations he desired for antithesis with &rrp&¾y. to•. In otherwords,8p• zt is not, as Finley supposes, synonymous with rroX,rrp•¾•zo•6•,but is a middleterm between&rro•¾•zo•6• • and the other extreme,rroXX&7:p0{etetet•. In this particularpassagenot only is the word 'polypragmosyn•'absent,but theconceptisalsonot in evidence.I fail to seehow the denial of the antitheticalconceptapragmosyngleadsto the emphaticassertionofpolypragmosyng,especiallysinceThucydidesusesa circumlocutionwhen giventhe perfectopportunityto usethe word itself. In fact we should not talk, as Kleve does,as if Periclesor the Corinthians, Athenians or non-Athenians,consideredusingthe word and 'avoided' or

14

JUNE

W. ALLISON

'refrained from' doing so.•9 A brief examination of the Thucydidean affirmationof 8p•, •:t against&r•p0•¾•oa6,• producessomeevidencethat

the historianhimselfmay havedeliberatelyavoidedr•oXt•r•O0•¾•oa6, •. As expected,8•, andr•p0•aazt,occurfrequentlywithoutemphasis,but the termsalsoappearconspicuously in the text asdefinitiveexpressions of policyor politicalphilosophy. 20The passageat II 64.4 isnot unique:it isto Periclesthat Thucydidesfirst attributesthe belief that action is a suitable policyfor a stateand its people.In the FuneralOrationPericlesrepeatsthe notion threetimes(II 35.2, 37.2,and 40.4). In the first passagehe recognizes that every man measureshis own capabilitiesto accomplishsomething

(8p•a0d•:•) againstthe actionsof acknowledged heroes;in the secondhe underlinesthe notion that private citizensare free to act in private as they choose (x0•0' '•8o• •:• 8p•); and in the third his insistenceon the superiorityof the doer is clear: Not by passiveacceptance(r•0•aZo,•:z•)do we possess friends,but by acting(8?•,-rz•). Thusthe one who performs(8•0•0•) a favor is the more secure,since he keeps the recipient in his debt by his kind intentions.

This sentimentPericlesthenappliesto the Atheniansasa wholeon a public basis.The readershouldcomparethe passagesat I 140.1,141.7,142.7, and II 64.1 for similarideas.Oncethe emphasison 'doing'hasbeenestablished for Pericles,it is repeatedby the other men in the History whom modern scholarshave describedas imperialists.For example, Cleon alsojuxta-

posesrc•Xz• with 8p•.¾ (III 38), andthewordsof Alcibiadesat VI 16.2not only recallthoseof Pericles,but are madeto echothe Melians'at V 102.But what of r•oXX&r•?&aazt,,whichis the extremeantithesisof We are too willingto acceptthe notionthat this phrasecan be put into servicealong with r•oXt•r•?0•¾•oa6,• when referringto the Athenians. Adkins (301-7) explores the phrase in various authors and so I shall consideronly Thucydideanevidence.I merelywishto bring to the reader's attention the fact that the verb phraseoccursas rarely as the term it is meantto parallel.Amongthe 185occurrences of r•p0• in the History, the phrasexoXX&xp0•**•[• occursonly once? No one seemsto have noticedthat Euphemus,the only one to sayxoXuxp•¾•oa6•q,is alsothe only one to usethe phrase:oXX&:p•**e[•; and he repeats•oXX0•in an emphaticclaim for Athenian methods(VI 87.2): We saythat... we are compelledto engagein manyinterests r•?&aaet,)becausewe alsoprotectand watchout for many(r•oXX&).

Euphemusis not one of theprostatai tou d•mou nor one of the major

THUCYDIDES

AND

HOAYHPAFMOY•¾NH

15

figureswhosecharacteris at all clearlydefined.Adkins,I believe,has correctlynoted the peculiarityin the speech: Prima facie,it is true, he useslanguagedifferently,sincehe is willing to term Athens'foreign policypolupragmosune.::

Buthe and Klevealsoarguethat Euphemus is 'ironically adoptingthe expectedlanguageof his opponents'. •3 For thisto be true, the senseof the word mustbe strongerthan'meddling',whichAdkinssuggests, especiallyif its absenceelsewhereis significantand if Euphemusthinksit is a term of disparagement. •4 Raubitschekhas demonstratedthe similaritiesand differencesbetweenEuphemus'speechand that of the Atheniansin Book I and shownthe affinity of hiswordsto thoseof the Atheniansat Melos? In particular, Raubitschekseemsto have underlinedthe flippant tone of the ambassador's remarks,when,after pointingout that Euphemusrefusedto argue from the 'glory of Marathon' and the liberation of the Peloponnesiansfrom the Persianthreat, he says:'Euphemusgliblydeclares... "all are free from envy(anepiphthonon)who providefor their own safety".TM In thissamemannerEuphemus,therefore,alsousesboth•roXX& and xoXurrp•¾•o•6v•,sinceinterferenceis perhapswhat is expectedby thosewho (like the Camarinaeans)cannotappreciatethe powerAthenshas or understandAthenian perceptionsof that power. I believeAdkins' and Raubitschek'sunderstandingof the condescending tone of the speechaids in explaining the appearanceof the noun, but I do not think that the connotations of •roXurrp•¾•o•6v• indicatethatit wasusedby Euphemus in a weak sense,nor is there any evidencethat Athens' opponentsever used the term. To my mind the mostsatisfactoryexplanationis that Euphemus useda term which was generallyconsideredpejorativeif found in serious contexts,regardlessof the political bias of the speaker.The rhetorical sarcasm by whichhe prefaces andunderscores •roXurrp•¾y. oo• cannotbe overemphasized.In a sententiousantithesishe had introduced xf>&aaz[,•,brazenly assertingthe necessityof aggressiveprotectionof Athenian

self-interests.

As if to force home to the Camarinaeans

this

frightening,but incontrovertibleposition,he nominalizesthe wholeethos of Athenianbehaviorin the monstrouscompoundrmXurrp0•¾y.o•6•. We might momentarily pursuea related line of argument. Adkins arguesfor the negativeconnotationsbehind and particularly its condemnationby the agathoi. He claims that the literature of the period was written by the agathoi. This generalization

impliesthat Thucydides, as oneof theagathoi,condemned •6•. As we have seen,Thucydidesnever indicatesthat he condones •roXurrp•¾y. oo•v•, but neitherdo any of the leaders,whetherthey be consideredagathoi or not. Certainly the historian would not have included

16

JUNE W. ALLISON

Cleonin thisgroupand yettheconceptpolypragmosyng is no moredirectly associated with him than it is with Nicias, Pericles, and Alcibiades.

Further, Thucydidesprovidesno reasonfor usto associateEuphemus'use of the word with the policiesof any of the major figures.In the settingof his speech,regardlessof homepolitics,Euphemusis compelledto presentthe most convincing,rhetorical argument to the Camarinaeans,for whom no real choiceexistsand whosedilemmaisreinforcedby thepresenceof forces of both sides.To usetermssuchas 'agathoi'and 'pacifist'(thus Kleve),or labelssuchas 'imperialists','peaceparty' and 'war party' is treacherous enough,albeit convenient,but then to align the peopleso labeledon one side or the other of an alleged policy or political concept, which is hypothesizedon evenlesssubstantialgroundsthan the labelsgivento its proponentsor detractors,is dangerousindeed? One further example of the dangerof falling into this trap may be helpful. On the Melian Dialogue, which is clearly an undisguiseddefenseof imperialism, Adkins concludes: Anyone (or any group or polis) who supposedhimself to possess arete would be likely to usepolla prattein or polupragmosuneto censurethe activitiesof anyoneelse... who wasactingoutsidehis own sphereof influencein sucha wayasto crosstheboundariesof the censurer'ssphereof influence.... 28

He alsonotesthat the term •:o3.ur•f>0ty•to•v• is 'significantlyabsent'from the dialogue? If it was as prominentas we havebeenled to believe,the Melians or Thucydides misseda brilliant opportunity to censurethe Atheniansby usingit of them. While it may havebeen'likely' that either

Meliansor Thucydides wouldusetheterm,onlyEuphemus infactdoesuse it.

It seemsto me that the role assignedto the conceptpolypragmosyng in the History has been enlargedbeyondthe evidenceprovided by the text. Although we may wish to continue using the term to describecertain qualitiesand attitudeswe deem important for a generalunderstandingof the fifth-centuryAthenian,we mustbewary of foistingour ownpropensity for terms which summarizecertainattitudesback upon Thucydides.It is abundantlyclearfrom evena cursoryreadingof the essays on

•toa6v•that most of our notionsof what the word 'polypragmosyn•' meansderivefrom fourth-centurysources,wherethereislittle doubtabout the authors'intention to refer to the concept,becausethe word is found with great frequency and in contexts where the connection with the Atheniansis explicitlymade.30We shouldnote, however,that neitherthe noun nor the verb (•:oXurrp0ty•ov•½tv) is found in Demosthenes,Isaeus, Andocides,Dinarchus,or Lycurgus,and Lysiasusesthe verb only twice.

THUCYDIDES

AND

IIOA¾IIPAFMOY•¾NH

17

By contrastwe find a greaterfrequencyin Plato, Xenophonand, particularly, in Isocrates.I can only speculateon the reasonsfor this surprising grouping. First, it would appear that the termswere not commonplaces which people readily acceptedor anticipated, since they are virtually unattestedamong the practisingorators. Secondly,it may be saidthat the terms were no longerapplicableto the Athenian political positionand henceappear only in historicalor theoreticalworks. But, as I haveshown

andwill furtherargue,it isquestionable whetherr:oXvr:p0t'r[zo•6,•q wasever a commonlyappliedattributeof Athensin thefifth century.And third, the greateremphasison the•oXvr:p&r[zo•¾ in fourth-centurycomedysupports the notion that the term possessed sarcasticor humorousovertones. In a discussion which treatsthe relation of the word to the conceptit expresses, it is appropriateto considerthe frequencyof the word in other fifth-centuryauthors.It has beenarguedherethat it is not likely that an author would repeatedlyassumehis audience'sawarenessof the importanceof a conceptwithoutprovidingthe singlelinguistictokenexpressive of the concept,if his vocabularycontainedsucha word. The frequency, therefore,is an ancillary indicationof how significantthe conceptwas.

No instances of thewordr:oXvr•p0t'r[zo•6,•q or itsrelatedformsareto be found in Pindar, Aeschylus,Sophocles,or Euripides.3• I cannotpretendto offer a universallyacceptableexplanationfor this, but I might enumerate somepossibilities:one, xoXu compounds,with the exceptionof Homeric or epic-sounding words,wereconsideredunpoeticor lackingseriousness and hencenot suitablefor lyric or tragedy;two, in the casesof Pindar and Aeschylus,the word may not yet have been coined; three, even in those works which we think refer directly to Athenian political enterprisesthe wordmay not havebeenusedbecauseit wasinappositein seriouspolitical contexts.

The verbr:oXur:p0t'r•zo,•[,• is found in Herodotusat III 15.2.Although LSJ saysof this occurrence:'esp. meddlein stateaffairs, intrigues',it is evidentthat this definitionis basedon the contextand is not necessarily a regular meaning of the word. Psammenitus,who has been sparedby Cambyses,'would have recoveredEgypt', Herodotus tells us, 'if he had beensmartenoughto restrainhimself','to keepquiet'or 'to lie low'. Since we have not been told, nor are we told, until after the digressionon the Persiancustomof restoringsonsof royalty, what kinds of interfering activitieshe becameinvolvedin, thewordhasto carrya generalmeaningof the sort I have suggested.We have no other instanceof the word in the History to showwhat Herodotususuallytook it to mean. Further, hedoes not assumethat the readerknows,as is demonstratedby the explanation whichhe providesat 15.4:Psammenitus engagedin conspiracy.I think it probable that Herodotus' very explanation has given rise to the lexical definition

of this occurrence.

18

JUNE W. ALLISON

In the Acharniansof 424 Aristophanesusesthe noun at line 833, whereit seemsto meansomethinglike 'sorryfor buttingin' or'I guessI put my foot in my mouth', even though Dicaeopolis,who is

unquestionably themostinventive,aggressive, andambitiousof all Aristophanicheroes,is speakingto the Megarian.Pherecrates, perhapsa slightly oldercontemporaryof Aristophanes, usesthe verbwith similarcolloquial color) 2

The adjectiveoccursin the Birds at 471. Pisthetairosis convincingthe birds of their regal heritage. To their naivet• and ignorance of their mythological position he says, 'for you are ignorant and not

rcp&yy. to¾'.Sincein and of itselfthelinedoesnot conveypoliticalassociations, one must, if one thinks there are such associations,claim 1) that

Pisthetairosimpliesin the statementthat he isa rcoX•rcp0•¾y. to¾,in thesense of an ambitious, self-motivated,and enlightenedperson;2) that the traits of thischaracterin the playadequatelydefinethe termin all contexts;and, most crucial to such an argument, 3) that AristophanesintendsPisthetairos and his traits to stand in some sense for the Athenians. This last has

beena major issuein Aristophanicscholarshipand remainsa moot point. The extremesmight be representedby Whitman, who viewsthe play as essentiallya fantasyof words,and Arrowsmith,who seesit as a political fantasy,whosecentralthemesare •:p• andrco•.•rcpot¾[a. offtSv•. 33We areon more solid ground with the Plutos, where(line 913), as Adkins observes,

•oX•p(zyy.o¾&t¾is connectedwith e6epyez&t¾ in specificbut ironic referenceto the city.34

Finally,I shouldmentionwhatmaybe a fifth-centuryusageapproximatinga definitionin a serious context:Democritus' 'it isshameful, while being•o•,vrcpoty•o•o•zot inregardtoforeignmatters (z&60veZot), toignore one'sown affairs(-r&o[x•]tot). TMThe problemwith thefragment,however, liesnot in the understanding, but in the fact that it belongsto that groupof

fragments whoseauthorship is disputed. Theyareperhaps to beassigned not to the fifth-century Democritus,but to the mid-fourth-century Democrates.

From theseoccurrences two generalconclusions can be drawn:one,the

fewappearances of theword(s)in fifth-century authorsandthecontexts in whichtheyappearcastconsiderable doubtontheseriousness withwhich we are to take the characterizationsof fifth-centuryAtheniansas

rcpdyy, oYeg. Secondly, thereis no evidence amongThucydides' contemporarieswhichis stronger thanwhatwe find in the Historyandwhich mighthavebeenusedto providesomesubstantiation that thehistorian impliespolypragmosyng in thosepassages wherehe usesthe phrase'ouk apragm6n'.

In concluding,I wouldlike to point out what may seemobvious,but since I have not seenit stated elsewhere,there is some need to do so here:

THUCYDIDES

AND

HOAYHPAFMOZYNH

19

the abstractnoun, which we are prone to think of as a key word, or in extremecasesa politicalcatchword,asfar asI cantell, onlyoccurstwicein fifth-centuryliterature,viz. in Thucydidesat VI 87.3and in Aristophanes, Acharnians,line 833. Again, I do not intendto impugnthe efforts of those scholarswhoattemptto recreatethe politicaltemperandmilitaryvisionof the Athenians, but I think that, had this last observation been made long

ago and had scholarsbeena bit more carefulin their own useof wordsand concepts,we would not now be so far down the road to enshrining •oX•p0•¾•o(•6v• l asthe nationalcharacteristic of fifth-centuryAthenians. June W. Allison

The Ohio State University NOTES

*I would like to thank ProfessorsS.V. Tracy and A.G. Woodheadfor their time and advicein the preparationof this paper and extend my thanksto the editor of this journal and its anonymousreadersfor their careful and incisivecomments. I. J.H. Finley, Thucydides(Ann Arbor 1963) 127-8,143•, 220. When referring specificallyto the distinctionbetweenthe word and the conceptI shalltransliterate and usesinglequotesfor the word and italicsfor the concept.In directquotations from secondarysourcesthe authors' own usageis followed. 2. V. Ehrenberg,'Polypragmosyne:A Study in Greek politics',JHS 67 (1947) 44-67, reprintedin Polls and Imperiurn(Zurich 1965)466-501. 3. A.W.H. Adkins, 'Polupragrnosuneand "minding one's own business":A study in Greek socialand political values', CP 71 (1976) 301-27. 4. In fact, of the twelve r•o•t•- compoundsonly three occur more than once: r•o•t•,•0p•0r•o•,r•o•t•T•, and the Homericr•o•6Tpor•o•.In part the scarcityof suchforms in the History suggests that Thucydidesavoidedthemexceptfor special effect.

5. This form of argumentationis describedby W. Stegmiiller, 'Towards a rrational reconstructionof Kant'smetaphysicsof experience',Part I, Ratio 9 (1967) 1-37; Part II, Ratio l0 (1968) 1-18. 6. Stegmiiller, Part II, I. 7. It might be argued that the singleappearanceof a word in a work which containsrelatedtermsor carefullyleadsup to the oneappearancemakesthat one instanceunusuallysignificant.This is not pertinentto the presentdiscussion,since no scholarhas soughtto make this claim and the contextof the oneinstancedoes not particularlymark it as a key political term. 8. Ehrenberg (above n. 2) 46. 9. Adkins (above n. 3) 315. 10. K. Kleve, "Ar•p0•¾M. oo4,•l and r•oXt•r•p0•¾M. oo4,•l, two slogansin Athenian politics',SO 39 (1964) 86-7. 11. P. Huart, Le vocabulaire de l•nalyse psychologiquedans l'oeuvre de Thucydide(Paris 1968)38545.A.G. Woodhead,Thucydideson thenatureofpower (Cambridge, Mass. 1970) 43, 49-50: 'Pericleswas by no meansan opponentof

20

JUNE

W. ALLISON

polypragmosyne,quite the reversein fact, for it was the inactive, uninvolvedman, the dmp0•¾[z•ov, whom he rebuked'(50). 12. Ehrenberg(above n. 2) 47. 13. The confusionbetweenwordandconcepthasbeenthesubjectof widedebate amonglinguistsand philosophers. For generaldiscussions seeG. Ryle, Theconcept

of mind (London 1949), R. Carnap, The logicalsyntaxof language(New York 1937), and W.V. Quine, Word and object (Cambridge,Mass. 1960). In From a logical point of view (New York 1953) 2-4, Quine deals with the confusion over

what is,whatis said,and concepts.E.g. in speakingwe are notliableto confusethe Parthenonwith the Parthenon-idea,but in dealingwith a Pegasusthe distinction betweenbeing and the conceptcollapsesin verbal description.W. Sellars,'Language as thought and as communication',Philosophyand phenomenological research (1969)506-27,seemsto meto haveclarifiedtheextentof theconfusion by reducingit to a distinctionbetweenthecontexts:'personexpresses' and 'utterance expresses'(526). The former may take as an object 'concepts',and the latter, 'words'; this is not the sum of the distinction betweenthesetwo, but it servesfor the

presentdiscussion. The only Thucydideanscholarwho has madean attemptto uncoveran awarenessof conceptsby the historianby consideringhis syntaxis F. Solmsen,'Thucydides'treatmentof words and concepts',Hermes99 (1971) 385408.

14. Ehrenberg (above n. 2) 50. 15. Ehrenberg, 46.

16. Ehrenberg,49. More recently this view has been repeated,as regards individuals,by H.R. Immerwahr,'Pathologyof power and the speechesin Thucydides',Thespeeches in Thucydides, ed. PhilipA. Stadter(ChapelHill 1973) 30, and by W. Arrowsmith, 'Aristophanes'Birds:The fantasy politics of eros; Arion n.s. 1 (1973) 131 f. We must remember that such evaluationsare deduced from Thucydides,who lays the blame for the failure in Sicily on the dgmosand Alcibiades'enemies.Cleonisalsonot withouthismerits.Thucydidesmakeslittle of hismanysuccesses asa generalbeforethecatastrophein the north.SeeWoodhead, 'Thucydides'portrait of Cleon',Mnemosyne13 (1960) 289-317;for an opposing view see W.K. Pritchett,'The Woodheadeaninterpretationof Kleon'sAmphipolitan campaign', Mnemosyne 26 (1973) 367-86. 17. Finley(aboven. 1) 154.W. Nestle,' 'A•p0•¾•oo6v•',Philologus81=n.s.35 (1926) 131-2,and especiallyDer Friedensgedanke in derantiken Welt, Philologus,

Suppl. 31.1 (1938) 22-3, first madethe observationthat ,:6 $½0m,:qO,.ov is the antithesisof &:O0•¾•t•v, whoseoriginshe findsin Gorgias.Seealso H.T. WadeGery,'Thucydidesthe sonof Melesias:A studyof Perikleanpolicy',JHS 52(1932) 224-5, for a full discussionof the passage.He prudently does not mention •:oXu•:p0•ylzoo6v' •. 18. In this casethe [zkv... 8k antithesisis extendedby the additionof a third memberintroducedby 8k:e[ 8• •r[½[xqx•xr•:0t[, qo0ovqae[ (II 64.4). 19. For example, Ehrenberg:

It is alwaysxoXu•:O0t¾[zoo-6v• that isat theverycentre,andwefind it most emphatically expressed in thosepassages in whichAthenian imperialism is mostforciblyeitherattackedor defended--in the second speech of the Corinthians, in thelastgreatspeech of Pericles, andin Alcibiades' speech beforethe Sicilianexpedition(aboven. 2, 51).

THUCYDIDES

AND

HOAYHPAI-'M0Y•¾NH

21

Kleve (above n. 10) ingeniouslynotesthat Euphemus'usereflectsthe policy of the 'war party', but 'that responsiblepoliticiansof the war party, however,preferred not to use the word appearsfrom Pericles'speech,where it is avoided' (87).

20. I give a list of the passages in which 8p•¾or rrpd•zt,• occursfrom the speeches of the Atheniansat Sparta, Pericles,Cleon, Diodotus, Nicias,Alcibiades, and from the Melian Debate. Those marked with an asterisk are particularly noteworthybecausetheyseemto describea definitepolicy.•p&;•;zt¾:I 140,141.7'; II 64*, 65.7;III 38.4,45.7*; V 89', 105;VI 10.2,13.2,15.4', 16.4,5,91.8p•¾:173.2*, 78.3, 142.7'; II 35.2*, 37.1,2', 40.1,4', 60.4, 64.1,4'; III 38.1',4, 40.4, 43, 46.5; V 102', 105.2; Vl 16.2', 21', 83.

21. Of the two other passages which contain the word rroXX0•, one (VIII 108.2)

usesthe wordZp•yxzz•z andin theother(VIII 87.5)rro•,•,& withXf•yxzz•zimpliedis the object of &•zXt3•zg. 22. Adkins (above n. 3) 311.

23. Adkins,311.Kleve:'Euphemus musthaveusedthewordrroX•rr?•z¾•oo'6• in bravado.Usinghisadversaries'description of hisownparty'spolicygavehisspeech a certain air of piquancy...' (above n. 10, 87). 24. Adkins, 314. He considers'meddling'a weakword 'in replyto Hermocrates' talk of aggressionand slavery'.It is evidentthat translation is a seesawaffair. Against Adkins' 'meddle'see Woodhead, Nature of power (above n. 11) 43. 25. A.E. Raubitschek,'The speechof the Atheniansat Sparta', Thespeeches in Thucydides(above n. 16) 36-8. 26. Raubitschek, 37.

27. Kleve adopts preciselythis line of argument: he assumesthe validity of referring to a war party or peaceparty; he then arguesfor the political significance of such parties in the fourth century. From there it seemseasyto argue for the 'slogan'in the fifth century,and in particularin contextsdealingwith the Peloponnesian War.

28. Adkins,314. Ehrenbergand AdkinstreatrroXX& •Od•z• andrroXurf>0ty•o•6v3 as synonyms.This seemsincorrectfor severalreasons.Words derived from the samestemsare rarelystrictsynonyms.In thiscasenot only istheverbal stemthe samein theverbalphraseandthe compound,but in boththedescriptive modifier/prefix (rroXX0[ andrroXu-)comesfrom the sameword. We mustassumea shift in semanticimpactto justify the creationof the compound.Moreover,the appearanceof the phraseandthe compoundsideby sidein the sametextssuggests that we look for variations in the respectiveconnotations.The cumbersome compoundxoXuxf>0tyy. o¾&[•,probablycreatedfrom the adjectiverroXurr•0•yy.o•¾, seemsto be a linguisticexaggerationof xoXX&xp0[•e[•; Euphemus'useof both

xoXX&rrp&$•ztvandrroXuro0tyy. o•6•• seemsto demonstrate justthispointandto supportthe notion that the corresponding denotationis also an exaggeration. There is, further, a reassuringconsistencyin the humorousflavor of rroXu xp0•y•o•6¾•and in thefrequencywith whichit occurredin comedy,tojudgefrom fifth- andfourth-century comicpoets(e.g.Timocles,Diphilus,andHeniochus each entitleda comedyPolypragm6n).For a discussion of compoundformationsee Schwyzer, Griechische Grammatik I 434, 447 f., 584.

29. Sincethetermoccursoncein theentirework,itsabsence hereishardlyuseful asevidence for anything,because it failsto distinguish in anysignificant way one contextfrom another.Adkinsmay havederivedthe supposed significance from a

22

JUNE

W. ALLISON

tooreadyacceptance ofEhrenberg's remarkonthedialogue: 'Nolonger dowemeet thewords•:oXu•poq, F•oo-6v• and&r•p•¾•o•6v•'--which ispreposterous, sincewe havenot yet met 'polypragmosyn•'.

30. The citationsfrom fifth- and fourth-century authorsare takenfrom the

several authorlexica,Kock,CAF(andMeineke), Nauck,TGF,Diels-Kranz, and Bergk,PLG. The readershouldconsult thediscussions of Ehrenberg, Adkins, Kleve,andK. Dienelt,'Apragmosyne', WS66(1953)94-104,for observations on

fourth-century usage. HereI shallsimplynotethefrequency for majorfourthcentury authorswhodealwithpoliticalandmilitarymatters: Plato:noun,2;verb,

12;Xenophon: noun,1;verb,4;Isocrates: noun, 5;verb2;adjective, 3;Lysias: verb,

2; Oxyrh. Historian:verb, 1; Demosthenes: none.

31. Norisanyinstance tobefoundinNauckorintheindices tothelyricpoets.

32. •y• 8' &v&v'rz•r•ot•t, • •oXu•p•y•6vzt(Meineke).

33. C. Whitman, Aristophanes andthecomichero(Cambridge, Mass.1964) 173.SeeArrowsmith, 'Aristophanes' Birds:Thefantasy politics oferos'(above n. 16),andespecially hisnotein histranslation, TheBirds(AnnArbor1961)110. 34. Adkins (aboven. 2) 317.

35. B.80(Diels): •[•Xp6v 'r&60ve[•:zo•,u:zp0ty•ov&ov'r0t &y•oe•v 'r&o[x•t•.It is

difficultto tell whether 60v½•is to be takenasreferring to foreignaffairsas opposed to internalpolitical concerns or 'foreign' in thesense of otherpeople's business. SeeThucydides I 141.7for xp&eectv with

BETRACHTUNGEN

ZUR •.LTEREN

GRIECHISCHEN

TYRANNIS

Der hier vorgelegteVersucherhebtnicht den Anspruch,allgemeingtiltige

Antwortenzur vielschichtigen Problematikder ',•lteren Tyrannis'zu bieten.• Vielmehr sollenanhanddreier Testf•illeinsbesondere zwei Fragen angesprochenwerden.Zum einenwerdendie antiken Quellennach ihrem Aussagewerthin durchleuchtet. Zum anderen wollen wir untersuchen,ob

derg•ingige Begriff',•ltereTyrannis'generell undundifferenziert anwend~ bar ist. Um den Rahmen diesesAufsatzesnicht zu sprengen,wurde eine Analyse der Entwicklung des antiken Tyrannisbegriffesbewusstausge~

klammert. A priori legenwir ferner zwei Oberlegungen als gegeben zugrunde. Die eine ist rein historischerNatur: Es m6ge nicht als eine Gegebenheitvorausgesetztwerden, dass sich unter sehr verschiedenen politischenBedingungengleichsamgesetzm•issig Herrschaftsformeniden~

tischer Pr•igungherausbildenmussten. 2 Die zweite Oberlegunggilt Herodot. Es ist m/Sglich,eine 'herodotischeChronologie',die ein in sich geschlossenes Ganzesdarstellt, zu erarbeiten, welchevon Herodot selbstin vielen einzelnen Str•ingen ausgesponnenwurde.3 Diese herodotische

Chronologiel•isstsich dann schliesslichin unsereabsoluteChronologie umsetzen(w•ihrendesftir Herodot selbstverfehlt wire, mit den Begriffen der absolutenund relativen Chronologie gesondertzu arbeiten).4 Hieraus ergibt sich die Feststellung,dassHerodot kaum etwas zur griechischen Geschichte zu berichten weiss, das wesentlich tiber ca. 600 v.u.Z. zurtick-

reichtund in unsererTerminologienochalsGeschichte geltendarf. Unsere drei Testf•ille sind' Pheidon von Argos, Kleisthenesvon Sikyon und Peisistratos von Athen und seine S/Shne.

I. Pheidon von Argos5

Unter den Freiern um die Hand der Agariste,der TochterdesKleisthenes von Sikyon,6 nennt Herodot auch die beidenPeloponnesierOnomastos von Elis (6,127,3) und Leokedes, 7 Sohn des Pheidon.œr apostrophiert diesenPheidon als tyrannosder Argiver und ftigt hinzu, dasser den

23

24

K.H.

KINZL

Peloponnesiernta metra bescherthabeund sichganz unerh/3rtunter allen Hellenen betragen habe, hatte er doch die Agonotheten der Eleier ihrer Funktion enthoben und selbst den olympischen Wettkampf geleitet (6,127,3). In der modernen Literatur werden diese Angaben als in sich widersprfichlichabgewertet.Da nach5,67,! Kleisthenesin einenKrieg mit Argosverwickeltwar, k/3nneer keinenargivischenFreier akzeptierthaben. Ferner sei, angesichtsdes Gewaltaktes des Pheidon in Olympia, die gleichzeitige Anwesenheit des Eleiers Onomastos und des Sohnes des Pheidon, Leokedes, h/3chstungereimt. Diese Kritik setzt voraus, Argos und Sikyon seienin einemmassivenmilitiirischenKonflikt gestanden,der in die Zeit der Brautwerbungfalle, sodassKleisthenesund Pheidondamals Kriegsfeindegewesenwiiren. Ferner stfitzt sich dieselbeKritik auf die Annahme,eshabeeinemachtvollekriegerischeInterventionder Argiver in Elis stattgefunden,weshalb die beiden Poleis auf lange Sicht zutiefst verfeindetwaren. Derlei mag zwar in Herodot hineininterpretiertwerden, ohne dassdie Grfinde jedoch zwingend wiiren. Ffir Herodot geh/3rtjener Pheidon derselbenGeneration wie Thales, Solon, Kroisosund Kleisthenesvon Sikyonan. Innerhalbder Chronologie Herodots dfirfte die sikyonisch-argivischeFeindschaft lange vor die Brautwerbungfallen (s.u.S. 26). Wenn unsereim folgendenvorgebrachte Deutung der historischenund politischenRolle des Kleisthenesund des Pheidon angenommenwird, bestehtkein Anlass, dassdie zwei Adligen, ihre Familien

und ihre Heimatstiidte

ffir alle Zukunft

verfeindet

sein

mussten.Die Bedeutungdes Aktes der Hybris des Pheidon in Olympia erkliirt sicham zwanglosesten, wennwir ihn als einenpers/3nlichen Exzess verstehen.Pheidon hatte sichim Rahmen einereinzelnenOlympiadezum Schiedsrichteraufgeworfen. Von argivischen Hoplitenaufgebotenund Kriegffihrung schreibt Herodot nichts. Herodots Bemerkungensollten auchnicht sogedeutetwerden,dasswegendesolympischenZwischenfalles dem Pheidon der Titel Tyrann gebfihre. Anderseits erscheint die Mittellung, Pheidon sei als metrologischerReformer tiitig gewesen,nicht ungereimt.Wurden doch seinenZeitgenossen(wie etwa Solon) ebenfalls derartige Neuerungen gutgeschrieben.Im kollektiven Gediichtnis der Argiver, ja 'der Griechen', lebte dieser Pheidon als ein ungew/Shnlicher Adelsherr weiter, dem man die 'pheidonischenGewichte' zuschrieb,der sich in Olympia ganz unerh/Srtaufffihrte, und dessenSohn Leokedes wiirdig war, sichum die Hand der TochtereinesArtgenossen in Sikyonzu bewerben;die Summedessen,weshalbman sichseinererinnerte,trug ihm die Bezeichnungz6p0•vvo• ein--sei es durch Herodot oder schonfrQher. Verstorbenmagdiesetgewissungew6hnliche,fiberdie Norm destypischen archaischenAdelsherrnhinausragende Pheidonvon Argoserstgegendie Mitte

des 6. Jahrhunderts

v.u.Z.

sein.

Dieset scheinbar so problemlose Sachverhalt wird zum nahezu un-

ZUR •LTEREN

GRIECHISCHEN

TYRANNIS

25

16sbarenRfitsel,sobaldwir die Welt Herodotsverlassen.Da begegnenwir nun dem Herakliden- und Temenidenk6nig von Argos reit Namen Pheidon. Hier ist nun vorweg grundsfitzlichfolgendeszu beachten:Eine Scheidungvon nichthistorischer(mythischer)und historischerKunde hat es ftir den antiken Geschichtsschreiber nicht gegeben--auchwenn diese

nattirlichobjektiviraruerbestanden hat;wennesaberim Dbergang vonder mythischenzur nichtmythischenZeit genealogische Leerrfiumegab, galt es diesezu tiberbrticken,und zwar wurde die Chronologiedesgeschichtlichen Zeitalters nach der Vergangenheithin zerdehnt, wfihrenddie der Heroenzeit nach unten hin verlfingert wurde, bis sich die zwei dann irgendwo trafen. Hinsichtlich Argos ist eine weltere grundsfitzlicheBeobachtung vorwegzunehmen:"Die lokale tiberlieferungvon Argos beginnt... in den spfiteren epen, unter denen die Phoronis das wichtigste ist";8 daran schliessensich rein genealogischeWerke von Akusilaos von Argos und Hellanikos von Lesbos, und erst Hellanikos verfasste einen historiographischenTraktat tiber diese Polis9--alles weltere geh6rt dem vierten oder noch spfiterenJahrhunderten an. •0 Die nichtherodotischenNachrichten datieren K6nig Pheidon unterschiedlich, vom frtihen neunten bis etwa zur Mitte des achten Jahrhun-

derts.•l Sachlich erfahren wir ausserdem Tatbestand des K6nigtums des Pheidon und seinesHeraklidenstammbaumesso gut wie nichts,wasweiter von Herodot abwiche, als von ihm Berichtetes auszuschmticken; dass Pheidon in Aigina die erstenSilbermtinzengeschlagenhabe,•2ist freilich m6glicherweiseeine Ausnahme. Zwar ist es objektiv unsinnig, die hohe Datierung mit Silbermtinzprfigungzu vereinen, doch weissdies erst der Numismatiker der neuestenZeit; 13von Mass- und Gewichtsreformen ist es nur ein kleiner Schritt zur Mtinzprfigung im Denken einesantiken Autors.

Ob der Verbindung reit Aigina in irgendeiner Form echte historische Kunde zugrunde liegen k6nnte, lasse ich offen; archaische Reformer kamenwelt herurn.Am interessantesten istjedochdie frtihestedirekt tiberlieferte Nachricht ausserhalb Herodots, nfimlich Aristoteles, Politik

1310626f., wonachPheidonvon Argosund andere'r670•¾¾o• [30ta•.z•0t• 6rr0t?Xo6o-•.ErstenssuggeriertAristoteles,dassPheidonder erstenSchichteder Tyrannen angeh6rte,wobei er auchandereTyrannen, die auf andere Art zur Tyrannis aufstiegen,zu nennen weiss: nfimlich Phalaris, Panaitiosvon Leontinoi, Kypselos,Peisistratosund Dionysios. Panaitios tritt uns spfiternochmalsentgegen(1316a30ff.) in Gesellschaft von Tyrannen, von denenkeiner filter ist als Myron, Vater desKleisthenes von Sikyon. Es ist schwerdenkbar,dasssichAristoteleseinenTyrannen, der um erheblichmehralsein ganzesJahrhundertden anderenvorausging, vorgestellthabe. Wie dem auch sei,die Umwandlungin einenK6nig hatte sichbereitsvollzogen,und Aristoteleshat diesealsgegebenhingenommen, ohne jedoch die herodotischeBezeichnung'Tyrann' aufzugeben.Es ist

26

K.H. KINZL

unzweifelhaft, dass sich die nichtherodotischen Nachrichten nicht auf

einen gemeinsamenNenner bringen lassen--doch f•ir unsere Zwecke brfichte eine Klarlegung der Quellenlagekeinen neuen Anstoss. Zusammenfassend

zum

Problem

Pheidon

sei darum

auf

unsere

Ausf•ihrungenoben S. 23 if. verwiesen,die durch die sonstigeantike

Oberlieferung nichtentkrfiftetwurden. TM Die GestaltdesK6'nigsPheidon heraklidisch-temenidischen Gebl•itesentspringtder Welt desMythosund ist alssolchef•ir den Historiker ohneBelang.t5Der von Herodot als Tyrann

apostrophierte argivischeAristokratnamensPheidonhingegenist allein als historisch zu betrachten und als solcher Gegenstandder Kritik des Historikers.

II. Kleisthenesvon Sikyon16

Antike Nachrichtenrankensichin gr6ssererF•ille um Kleisthenesund sein Geschlecht;Herodot widmet ihm zwei Exkurse: 5,67-68 und 6,126-130; letzterer bietet die Novelle vom Wettbewerb um die Hand seinerTochter,

wfihrend der erstere, dessenZentralmotiv die Phylenreform des Kleisthenesbildet, sich als das bei weitem komplexere Gebilde erweist. Zunfichst gilt es jedoch, Herodots Nachrichtendaraufhin zu durchleuchten, ob und wie sich Kleisthenesund sein Wirken in die zeitliche Dimension bei Herodot fiigen. Die Lesung, die sich in 6,125,1-126,1, sowiedurch sorgffiltigesKorrelationierenmit absolutenchronologischen Werten innerhalb Herodots, aufdrfingt, erbringt nicht nur einenzeitlichen

Fixpunktim LebendesKleisthenes, sondern aucheinigeweitereOberraschungen.Alkmeons Besuchder Residenz des Kroisos in Sardes nach dessenHerrschaftsantrittffillt in die dem Jahr 561•7 folgenden Monate oder Jahre, wfihrend Megakles'Auftritt in Sikyon alldem nicht vorangehenkann. Letztere Episodemussin einem olympischenJahr begonnen haben,woraussich HerodotszeitlicheEinordnungdesolympischenViergespannsieges insJahr 556 sogut wie zweifelsfreiergibt;hiermiterschliesst sich das Jahr der Vermfihlung der Tochter des Kleisthenes,Agariste, als 555.t8

F•ir das lodesdatum des Kleisthenesgibt Herodot wenig Anhaltspunkte. Nach 5,68,2 entledigtensich die Sikyonier sechzigJahre nach seinemTod t9der von ihm eingeffihrtenPhylennamen,um die traditionelldorischenanzunehmen;dies kann also nur nach 496/95 geschehensein. Falls irgendein Zusammenhangmit der Rolle der Sikyonier in der KampagnedesKleomenesgegenArgosca. 494 besteht?0wfireKleisthenes bald nach der Verheiratung seiner Tochter verstorben--keinesfallsaber spfiterals um die Mitte der vierziger Jahre des 6. Jahrhunderts.Was die Phylenreform des Kleisthenes betrifft, so k6nnen wir nicht einmal andeutungsweise eine Plazierungin HerodotschronologischemGeb•iude

ZUR •LTEREN

GRIECHISCHEN

ausmachen.TMWenn meine Annahme zutrifft?

TYRANNIS

27

dass die von Herodot als

•:p•¾,-xo• Xo1:o•apostrophierten dionysischen Kompositionen eineWeiterentwicklungder DithyrambendesArion, diedieserzur Zeit desPeriandros in Korinth niederschrieb, darstellen,gewinnenwir einen--rechthypothetischen--Hinweis, eine Datierung etwa im dritten Jahrzehnt des 6. Jahrhunderts

als sinnvoll

zu betrachten. •3

Soviel zu HerodotschronologischenVorstellungen,wie siesichunserer Ansicht nach ergeben. Sehen wir nun, welch ein Charakterbild des Kleisthenessich aus Herodot gewinnen Irisst. Hierbei diirfen wir der Er•hlung 6,126if. untergeordnete Bedeutungeinrgumenund unssogleich der wichtigen Passage 5,67 f. zuwenden. Der Schliisselpunktist'die Phylenreform.Ihre Ursachenund Stossrichtung bleibenunerklgrt,sowohl subjektiv(Herodot) alsauchobjektiv. Wennf/Jr Herodot dasMoment der 'NachahmungTMder HandlungendesSikyoniersKleisthenesdurchseinen gleichnamigenathenischenEnkel den erstenAnsatzpunktbildet, soist die politischeGesamtlagegleichermassen dominant: Kleisthenesbefandsich im Kriegszustandmit Argos.• Die im folgendenvon Herodot beschriebenenReformen werfen einigebezeichnendeSchlaglichterauf die Gestalt des Kleisthenes,der uns als zielbewusstund umsichtigplanendbeleuchtet wird, und dessenganzesTun und Handeln dem Heil seinerPolis gewidmet ist. Sikyons Heil ist selbstverstgndlich nur dann gewghrleistet,wenn ihm die G6tter gewogensind und es innerlichgeeintist. ErsteremZiel sinddie 5,67, letzterem die 5,68 berichteten Massnahmen gewidmet, die unter diesera Blickwinkel

betrachtet

werden miissen.

Die Deutung der Religions-und Kulturpolitik des Kleisthenesmussin dieseraRahmen trotz ihrer objektiventiefgreifendenWirkungenbeiseite gelassenwerden.26Dieser Aspekt der Reformen umfasst:das Verbot einer Kunstform--der Homerrezitationender Rhapsoden--und die F6rderung eineranderen--derdionysischen •:?0•¾,-xo• Xo?ot.--,fernerdie Einfiihrung einesneuen Heroenkultesmit Sitz im 'Prytaneion', die Aufwertung des Gottes Dionysos, sowieschliesslichdie v611igeEntwertungdesAdrastos. Anl/isslich der politisch und gesellschaftlichgewiss einschneidenden Phylenreform2? weissHerodot sichbetriiblicherweise nur auf die groteske Kuriosit/it der neuen Phylennamen zu beschr/inken, die er gem/iss allgemeiner griechischerVolksetymologie mit wenig schmeichelhaften Tiernamen verbindet, bzw. in einem Fall mit der Herrschaft des Klei-

sthenes. Hiedurch habe Kleisthenesdie Sikyonier zur Zielscheibedes Spottesgemacht. Die Sikyonier aber behieltendieseNamen noch lange bei. Herodot sprichtmit keinemWort davon,dassKleisthenes beabsichtigt h•itte, die Sikyonier l/icherlichzu machen;noch wenigeraber spricht er davon, classer die Mitbiirger, die nicht in seiner eigenen Phyle eingeschriebenwurden, verh6hnenwollte; und endlich l/isstsich aus seinen Worten nicht herauslesen,classdie Sikyonier unter Zwang die kleisthe-

28

K.H. KINZL

nischenNamen so lange beibehaltenh•itten. Es widersprichtjeglicher Logik, zu meinen, Kleisthenesh•itte Sikyon Argos gegeniiberstairken k6nnen, indem er die Mehrheit der Biirgerbeschimpfte.Aber ist nicht der neueName seinereigenenPhyle--Archelaoi--ein unwiderlegbarerBeweis hierftir?UnseresErachtensuntermauertder Name Archelaoi das gerade Gegenteilund nimmt zugleichdie Schliisselstellung zum Verst•indnisder Phylennamendes Kleisthenesein. Abstrakte Begriffe geben keine passendeneponymen Figuren als Phylenv•iterab--gewiss nicht im frtihen sechsten Jahrhundert--, wohingegen Heroengestalten die typischen Phyleneponymendarstellen.Ein Archelaosist unswohlbekannt:der Sohn des Herakliden Temenos, des Begrtinders der argivischen Temenidendynastie, welcher von seinenBriidern schn6deseinesErbes beraubt aus Argos vertrieben wurde und in Makedonien dann als Vorvater des

Argeadenhauses beanspruchtwerdensollte.Eine derartigeAbleitungdes PhylennamensArchelaoi fiigt sich auf das beste in die argosfeindliche Reformt•itigkeit des Kleisthenes. Somit ist der Annahme, ein hochfahrender

und

anmassender

Kleisthenes

habe die dominierende

Rolle

seinereigenenPhyle gegentiberden anderendurch eine abstrakte Namengebung propagiert, der Boden entzogen. Die Kehrseitejenes absonderlichen Bildes entbehrt gleichfalls jeglicher Grundlage: auch die von Herodot zitierten, wenig reputierlichen Vierbeiner geben keine brauchbaren Eponymen ab. Linguistisch geraten wir in nicht geringe

Schwierigkeiten, wennesgilt,Herodots'T•0•,'O•-r0•, Xo•z•-r0• von•, 6•o•, Xorooqlautgesetzlichabzuleiten.Die doppeltePr•imisse,dassdie Phylenreform und somit die neuen Namen der gesamtsikyonischen St•irkunggegenArgos dienenmussten,sowie,dassdie Phylennamenvon eponymen Heroen abzuleiten sind, schaltetdie M6glichkeit, von einem "schwerenAffront, nicht nur gegenden heimischenAdel, sonderngegen die dorischen Herren allenthalben",28 zu sprechen,aus. Die fraglichen Bildungen werden also von den Namen von Heroen abzuleiten sein, m6glichst von solchen,denen die Sage FeindschaftgegenArgos attestiert.29 Somit 1/isstsich in Zusammenfassungder herodotischenFakten ein Bild entwerfen,das dem desPheidon von Argos in vielen Ziigen/ihnelt (vgl. oben S. 23 if.). Es wird uns einer jener archaischenAdligen vorgestellt,derenAngedenkennicht verlosch,da siesichmit ausserordentlichen Massnahmendem Kollektivged•ichtniseinpr•igten.Im Fall des Sikyoniers Kleisthenessind es Reformen sowohl auf dem Gebiet des Kultus als auch insbesonderepolitischeNeuerungen,wobei man insbesondereder absonderlichenPhylennamengedachte;er war ein Adelsherr dessenStammbaumsichbisauf seinenUrgrossvaterzuriickverfolgenliess, der einen olympischen Sieg errang, und der Noble aus der ganzen griechischen Welt zur Werbungum seinerTochterHand auf seinenGiitern

ZUR •LTEREN

versammeln

GRIECHISCHEN TYRANNIS

konnte. Sein Todesdatum

29

ist um die Mitte des 6. Jahrhunderts

(oder kurz darnach) anzusetzen.

Wiederum--ihnlich dem Falle des Pheidonyon Argos--schwindetdie scheinbareProblemlosigkeit,sobaidwir HerodotsWelt verlassenund uns den anderen einschligigen Nachrichten aus dem Altertum zuwenden. Doch sehen wir uns in diesem Fall nicht einem Identititsproblem konfrontiert, und die Frage der Chronologie bewegt sich in engeren Grenzen.Es ist hier die dynastische Orthagoriden-Tyrannis, die unsin der ausserherodotischen Tradition gegenfibertritt(verbunden mit genealogischenFragen,wobei Kleistheneszu einemspiten Gliedjener Dynastie wird). Ffir die zeitliche Fixierung jener orthagoridischenTyrannendynastie liegen zwei Angaben vor: der olympischeWagensiegdes Tyrannen Myron (648),30 sowie die Gesamtdauer (100 Jahre)31 dieser Tyrannis.Zur Definierungder Anfags-und Endpunktehat man versucht, ein Papyrusfragmentaus dem 2. Jahrhundertv.u.Z. zu verwerten.32Der Text stehtteilweiseeinerTradition nahe,die von Tyrannenvertreibungen durchdie Spartanerhandelt,die unsauchnochbei Plutarch33und in einem Scholion zu dem attischenRedner Aischines34begegnet.Angesichtsder erheblichen Probleme, die einer schlfissigenInterpretation des Papyrustextesim Wege stehen,35 sollte bis zur K1/irung dieser Fragen das betreffendeFragmentbei der Diskussionfiber die Orthagoridendynastie und ihre Chronologieausgeklammertwerden.Neben dieserTradition, in welcher Sikyon und seine Tyrannis nur wegen des Elementeseiner spartanischenVertreibungsaktionvorkommen, tritt uns eine im grossen und ganzenziemlicheinheitlicheTradition gegenfiber,der vor allem das Moment tief wurzelnden Hassesgegen Tyrannen gemein ist.36 Diese

•dberlieferungsgruppe enthfiltferner eine ausgeprfigte Orakelkomponente.37 Im fibrigen lilgt sich auch Aristoteles,38 der sich fiber die

sikyonischeTyrannisweltersnicht feindselig•ussert,derselbenTradition angliedern,da er keinerleihiervon abweichendeFakten berichtet. Von der sogenanntenSikyonischen Anagraphe abgesehen,wurde im 4. Jahrhundertdie Lokalhistorienur durch Menaichmosyon Sikyon vertreten; hinzu kommendie einschl•igigen Partien desEphorosund die Verfassung der SikyonierdesAristoteles,wobeiletzterewohl gewissyon Menaichmos, der Lokalquelle, abhingen. Menaichmos verfassteferner ein Werk mit dem Titel Pythikos--ein gleichnamigerTraktat des Aristotelesjedoch machte es bald obsolet.39 Aus der GeschichteSikyons zur Zeit des Menaichmoswissenwir nur so viel, dassesvon der Geisselder Tyrannis nicht verschont blieb.40 Was l•ge nilher als die orakelreiche Hasstradition fiber die frfihen Tyrannen dem sikyonischenLokalschriftsteller Menaichmoszuzuschreiben? Andreas--der Urgrossvaterdes Kleisthenes

bei Herodot--wird hier zum Vater desBegrfinders der Tyrannendynastie,

30

K.H.

KINZL

Orthagoras;er seiniedrigerHerkunft gewesen?Orthagoras'Aufstiegzur Herrschaft folgt dem abgedroschenenSchema? das uns immer wieder entgegentritt--nach dem Vorbild der Darstellungder frtihen Laufbahn des Peisistratosbei Herodot (1,59, komplett mit Weissagungen,die der Vater erffihrt). Wenn man die stereotype Darstellungsweise,die feindliche Tendenz sowie die Bedeutungdes Wortes, dem der Name Orthagoras entspricht,und die verdfichtigeAbwesenheitdesOrthagorasbei Herodot unvoreingenommentiberdenkt,mtisstedem Historiker wohl gar mancher Zweifel aufsteigen. 43 Meines Erachtens ist Orthagoras als eine blosse Romanfigur zu betrachten. Am feindseligstenist die Darstellung der TyrannisdesKleisthenes,sowiederjenigenseinesVorg/ingersund Bruders Myron? Dies folgt wiederum der Schabloneder sich unter den Nachfolgern in tibelste Despotie wandelnden Herrschaft; irgendwelchehistorische Glaubwtirdigkeit besitzt all dies keinesfalls.Im Genealogischen

weichtdiesetlberlieferung insoferne von Herodotab, alssiedenvonihm bewahrten

Stammbaum

des Kleisthenes einerseits ausbaut und anderseits

in eine Art Tyrannenlisteumfunktioniert. Wie diesesktinstlicheStemma ausgesehenhaben mag, illustrieren moderne Rekonstruktionen45 des 'Stammbaumesder Orthagoriden'--wissenschaftlichenWert besitzensie keinen. Wir habenhiermit gewissermassen einentoten Punkt erreicht.Auf der einen Seite besitzenwir das Zeugnis des Herodot, in sich sinn- und gehaltvollund ohne aus sichherausnicht 16sbareWiderspriichlichkeiten. Demgegentiber steht, mit Herodot nicht vereinbar, eine gemischte Allianz--illustrierende Aristotelespassagen,Fragmente skandalstichtiger Kolportagevon der Art desNikolaos, Papyrusfetzchen unklarer Herkunft, u. dgl. mehr. AngesichtsdieserSachlagestehenunsdrei verschiedeneWege zur Wahl: entweder alle Quellen irgendwie miteinander in Einklang zu

bringenzu suchenoderdie nichtherodotische fJberlieferung--bzw. Teile daraus--als grunds•tzlich glaubwtirdig dem Herodot vorzuziehen oder aber die 'gemischteAllianz' zu verwerfen und Herodot als die einzig 'orthodoxe',authentischeQuelle zu behandeln. 46Meine eigeneEntscheidung f•llt zugunstenHerodots aus. In der Person des Kleisthenes erkennen wir den Archetypus des hocharchaischenAdelsherrenwieder. Sein politischesGewicht ist durch ererbten

Grund

und Boden samt den darauf wirtschaftenden

Kleinbauern

bestimmt. Hierdurch wiederumist er in der Lage, als eineftihrendeGestalt im politischen,kulturellenund milit•rischen Bereichaufzutreten.So ist er auch eines Platzes im Andenken

seiner Landsleute

versichert.

Doch um

eine hervorragendeRolle bei der Neuordnungder Phylenzu spielen(ja, diesevielleicht sogar initiiert zu haben), um sichals Patron dionysischer Dicht- und Kompositionskunstzu verewigen,um in kriegerischerLageals milit/irischer Anftihrer aufzutreten--zu alldem bedurfte es keiner dynastischenTyrannis samt gruseligemZugehOr. Wir dtirfen Herodot folgen

ZUR ,•LTEREN

GRIECHISCHEN

TYRANNIS

31

und Kleisthenesals jene Einzelerscheinung(die in vieler Beziehunghervorstachund folglich im Denken Herodotsden Titel Tyrann verdiente) sehen,alswelcheer ihn darstellt.Doch miissenwir Herodot gegenfiberdie rechten Proportionen wiederherstellen.Das Entscheidendezur Abgabe eines historischen III.

Verdiktes

ist allein der Sachverhalt.

Peisistratos

von Athen und seine Sfhne 47

In den Aufbaujahren des kleisthenischenStaates,als Athen sich an allen Fronten bedroht sah, ging eine Gesandtschaftan den Hof des persischen Satrapen zu Sardes? Sie erreichte, was ihr Mandat war, doch als die Gesandtenzurfickkehrten,wurde ihr diplomatischerErfolg negiert,und man wollte nichts yon dem Ausgehandeltenwissen. Nun ist es zwar unbestreitbar,dassdie Rfickffihrungder Peisistratideneine unabdingbare persischeBedingungffir ihre Parteinahmezugunstender Athener darstellte.Weshalbaber sollteein athenisch-persisches Arrangementgerade an diesem Punkt scheitern?Die Athener konnten schwerlicherwarten, dass

die Perser nicht irgendwiein Athen vertreten sein wollten--ein yon den

Perserngegiingelter Hippiassolltegewisseingeringeres libel seinalsein yon den Spartanern freie Hand gegebenerIsagoras, oder als ein yon niemandemgezfigelterHippias, wie man ihn vor 511/10 kannte. Es ist offenbar, dassdie Perser nicht begriffen (weil dies den Gesandtenselbst anscheinendnoch gat nicht so richtig klar war), dassder kleisthenische Staat nicht nur aus launigem Herumreformierenbestand, sondern in kompromisslosester Weise die Staatsverwaltungauf das rechtsstaatliche Prinzip und auf die Idee der Rechtssicherheit grfindete? Dies und nichts weniger ist es, was man einer persischenHilfeleistung zuliebe durch Riickffihrungdes Hippias h/itte opfern mfissen.Also hatte man in Athen keineandereWahl alsdie eigenenGesandtenzu desavouieren. 50Hiermit ist bewiesen, dass Hippias krimineller Vergehen schuldig war, begangen w/ihrend seinesTerrorr•gimes, das zu seinerVertreibung geffihrt hatte;er war in der Tat ein Tyrann gewesen,Athen hatte in der Tat unter einer Tyrannis gestanden.Als Historiker fragenwir, wie eshierzugekommenist. Deshalb spannenwir den Bogenzurfick zu den sechzigerJahrendes 6. Jahrhunderts und der bunten Karriere des Vaters des Tyrannen, also des Peisistratos,Sohn des Hippokrates. Im ersten Drittel

des 6. Jahrhunderts

ist Athens

weitere

Geschichte

wesentlichbestimmt worden, wie man wohl ohne zu fibertreiben feststellen

darf. Es ist auch eine Epoche,in welchertief wurzelndearistokratische Zwietracht

offen aufbricht.

In und fiber alldem steht die Gestalt des Solon.

Auch andereNamen sindunsfiberliefert, seiesdank der Archontenliste,sei

esdurchihre hervorragendeRolle im Wechselspiel desMachtkampfesder aristokratischen

Staseis. Die Pers6nlichkeit

des Peisistratos

nimmt

eine

32

K.H.

KINZL

prominenteStellungin dem politischenRingenein, beginnendmit seinem Hervortreten als milit•irischer Anftihrer (m.E. unzweifelhaft in der offi-

ziellen Funktion eines•pXto¾•roX•,•pXo•)5• gegenMegara in den frtihen sechzigerJahren.5:Bis zu seinemTode vier Jahrzehntesp•iter,im Archontenjahr des Philoneos?3 sollte er immer wieder entscheidenddie attische Politik mitbestimmen.Dem traditionellen Bild entsprechendgeschahdies in der Form von drei Perioden, w•ihrend derer er als Tyrann tiber Athen gebot, und welche durch zwei Abschnitte in seinem Leben, die er in politischbedingtemExil anderweitsverbrachte,jeweils klar voneinander getrenntsind. Dem kritischenAuge Julius Belochskonntenicht entgehen, dasshier irgendetwassichnicht rechtreimenwill. Doch indemer dieersten beiden Episodenpeisistratischer'Tyrannis' zu einer einzigenverschmolz, hat er das Kind mit dem Badeausgeschtittet•4--undzwar, wie unsscheinen will, da er sich doch nicht g•inzlichyon der Tradition zu 16senvermochte. Der Kernpunkt der Frage der 'drei Tyranneis'des Peisistratosstellt sich unsdar alsdie Aufgabe,die StellungdesPeisistratosin und zu athenischem Staat und athenischer

Gesellschaft

zu ermitteln.

Ein solcher Versuch kann

jedoch nur dann sinnvoll erscheinen,wenn es uns irgendwiegelingt, das historisch-faktischeaus seiner anachronistischtiberlagerten Umgebung herauszusch•ilen. Lassensichdie dergestaltgewonnenenFakten dann yon neuem, doch unter anderen Vorzeichen als etwa bei Herodot, zu einem geschichtlichsinntragenden Ganzen wieder zusammenftigen,so sollte

unser Experiment zu einem Ergebnis geftihrt haben, welches das zu fordernde

Mass an historischer Wahrscheinlichkeit--und

somit Glaub-

wtirdigkeit--besitzt. Im Verlauf desHin- und Herwogensder aristokratischenStaseis • wurde

dem Peisistratoseine aus xo•uv•q>6•o,bestehendeLeibwachevon der Volksversammlungbewilligt. Wir erkennen aufgrund dieses Sonderarrangements,dass Stasis per definitionem nicht Gewaltanwendung einschliesst,denn ansonstenh•itte Peisistratosbewaffnete Anh•inger einzusetzenvermocht,um seinepolitischenGegnerzu terrorisieren.Hiermit

nicht genug,besetzenPeisistratos und seinekeulenschwingenden 'Leibw•ichter'die Akropolis.Nun mag zwar diesoderjenesDetail der archaischenAkropolisstrittigsein--unbestreitbar ist nut dieeinfachegeologische Tatsache,dassesauf dem Plateauselbstkein Wassergibt (vgl. Hdt. 5,65,1). Kurzfristigmag der altehrwtirdigeBurgfelsenmit seinenKultst•itteneine

hervorragendeFluchtburg abgeben--als Regierungszentrumist die Akropolis kaum geeignet. 56 Die gegnerischenMachtgruppierungen bleibenihrerseitssoweitunangetastet,dasssiesichzu einer konzertierten Aktion versammeln k6nnen. Der Mann auf der Akropolis wird samt seinen keulenbewehrtenGenossenaus Athen hinausgeworfen,und man

kehrt zur Tagesordnung der Stasiszurtick.Als abet Megaklesdie Gefahr,

ZUR •LTEREN

GRIECHISCHEN

TYRANNIS

33

den kiirzeren zu ziehen,droht, besinnter sichdarauf, seinepolitischeBasis zu verbreitern, nicht un•ihnlich der Taktik, die sein Sohn ca. 50 Jahre

darnach gegenIsagorasanwendenwiirde? Besonderswichtig war offensichtlich,wer die dem PeisistratossichzuneigendeMachtgruppierung58auf seineSeitebringenkonnte--und dieswiirde durcheinedirekteAllianz mit Peisistratos

selbst am sichersten

zu erzielen

sein. Jenem war es offen-

sichtlichdringendgenugdaran gelegen,nachAthen heimzukehren,sodass er mitspielte.Doch als Megakleserfuhr, dassPeisistratosan der ihm zur Frau gegebenenTochter desersterenEmpf•ingnisverhiitung iibte, warf er ihn stracks wieder hinaus. Falls hier irgendeinerder Beteiligtenden BeinamenTyrann sichverdiente,so Megakles.Peisistratosaberbliebvoile zehn Jahre im Exil. In dieser Dekade m•igen sich die innerpolitischen Spannungenetwasabgektihlthaben, Megaklesmag seineVorrangstellung zu konsolidierenvermocht haben, und es m•igen auch neue politische Konstellationen entstanden sein, so vor allem um die Halbbrtider Kimon,

Sohn des Stesagoras,und Miltiades, Sohn des Kypselos,sowie Hippokleides, Sohn des Tisandros, der irgendwie zu deren Verwandtschaft z•hlte?

Die zweite Heimkehr des Peisistratoswar yon langer Hand und griindlichgeplant. GewissverbrachtePeisistratosdie zehnJahredesExils nicht nur in Gedanken an die Rtickkehr selbst, sondern auch an Massnahmen, die der Rtickkehr folgen soilten und eine abermalige

VertreibungnachM•iglichkeitausschliessen wtirden.•berspitzt formuliert stellt sich die historischeFrage nach dem Wesender Unternehmung desPeisistratosetwafolgendermassen dar: War esPeisistratos'Bestreben, sich und seinerFamilie die Riickkehr nach Attika zu erzwingenund ein sicheresVerbleiben daselbstzu bewerkstelligen,oder war er einzig vom finsteren Tyrannenstrebenerfiillt, der Polls seinen Willen aufzuzwingen und seinearistokratischenGenossenzu demtitigen?Wer die letzten etwa eineinhalb Jahrzehntehis zum Tod des Peisistratosunvoreingenommen betrachtet,dem dr•ingt sichletztererEindruck durchausnicht als gegeben auf, wie wir sogleichsehenwerden. Was die Rtickkehr selbstbetrifft, so gibt es hier freilich keinen Zweifel: sie war mit Gewalt in der Form yon fremdenHilfstruppenund S6ldnern6ødurchgesetztworden, und zwar so

griindlich,dasseineganzeAnzahlfiihrenderAdelsherren sogleich einmal

dasweitesuchte--aber dochnichtalle,wiedasBeispiel desMiltiadesdes •lteren zeigt,dersichm•ichtig genugfiihlte(undesoffenbar auchwar), einfach auf seinen Giitern sitzen zu bleiben. 61 Zwar verliess auch er bald

darauf Attika, doch nicht als politischerFltichtling,sondernals respektgebietenderOikistes der thrakischenChersonesos,der somit gewissermassendie Grenzen Athens an die Meerengenvorverlegte. 62Doch die Entwicklung w•ihrend der folgenden Jahre vermittelt nicht das allzu

34

K.H.

KINZL

vertraute Bild eines despotischregierten Athen. Zum einen sind die geflohenenAdligen nicht langeim Exil verblieben,wof•ir wir in zweiganz besondersprominenten F•illen greifbare Hinweise besitzen.

Kimonder,•ltere,VaterdesMiltiadesdesJ•ingeren, hatteseinen Sohn bei sich in Athen; dies kann sich schwerlich auf eine andere Zeit als die des Heranwachsensdes Miltiades beziehen,also aufdie Jahre um 540.63Ferner

gelang ihm das ganz ungew6hnlicheSpektakel, mit demselbenViergespann dreimal hintereinander olympische Siege davonzutragen; den zweiten Sieg aber •iberliesser grossz•igigdem Peisistratosund kehrte darauf nach Athen helm. Es liegt nichtsnilher,alsdiesenzweitenSiegauf die der gewaltsamenR•ickkehrdesPeisistratosfolgendeOlympiade(544) zu datieren.64Das zweiteprominenteBeispielist die Familie desMegakles selbst.Auch sie waren gefliichtet. Doch im Jahr 535 oder 534 oder 533 h6ren wir von der ersten Trag6dienauff•ihrungdes Thespis.65 Hierzu bedurfteThespisder Untersti•tzungeinesM•zenas, dem an dionysischen Belangen besondersgelegen war. Ich m6chte diesesM•zenatentum der Familie des Megakles zuschreiben,66womit ein deutlicher Hinweis aufeine relativ baldige R•ickkehr dieser Familie gewonnen w•re. Zum anderen aber fehlt auch nur der geringsteHinweis auf eine Weiterf•ihrungjener Adelszwietracht,wie sie uns vorher (und dann wieder gegen Ende des Jahrhunderts,wenn auch unter bereitsverfindertenVorzeichen)begegnet war.

Wer diese Sachlagemit dem Hinweis, wir h•itten es nun eben mit einer Despotenherrschaftzu tun, bagatellisierenwill, iibersieht,dassein solches Willkiirr•gime zu Widerstand h•tte f•ihren m•issen--w•hrend unsere Quellen keine Spuren vom Wirken politischerEmigranten enthalten,die doch im Falle des Despotentumsdes Hippias67 wohldokumentiert zu dessen schliesslicherVertreibung vielf•ltig und entschiedenbeitrugen. Ebenso undenkbar aber ist es, dass die attischen Adeligen sich zur R•ickkehrbequemth•tten, um unter einemDespotensichausschliesslich der

Musse

zu widmen.

Mit

aller

Entschiedenheit

aber

ist die

letzte

verbleibende 'Erkl•rung' zu verurteilen, nfimlich, dass Peisistratosein Wohlt•iter und Tyrann in einer Person gewesensei, mit dem sich der attischeAdel abgefundenh•.tte:die Geschichtelehrt, dasses ein solches Wesennicht gebenkann. Wer in das Athen des'Tyrannen' zur•ickkehrte, sanktionierte dessen Tyrannis und muss als mitschuldig betrachtet werden--dieser Schlussfolgerungkann der Historiker nur dann ausweichen, wenn er die Jahre, die der ersten Periode des Schocks der gewaltsamenRiickkehr des Peisistratosfolgten, als eine Epochefruchtbarer Zusammenarbeit

und friedlicher

Koexistenz

der Adelsh•.user

neu

definiert. Ein solcherpolitischerNeubeginn(der sich unter Umst•inden bereits nach der zweiten Vertreibung des Peisistratosanzubahnenbegonnen haben mag, vgl. oben S. 33) •urde als solcherbereitsden Keim zu

ZUR •LTEREN

GRIECHISCHEN

TYRANNIS

35

einermerklichenVerbesserungder allgemeinenwirtschaftlichenund somit

sozialenLagein sichgetragen haben. 68Die durchMiltiadesden•lteren erfolgreicheExpansionattischenTerritoriumswar in doppelterHinsicht wirksam: einerseitsbegleiteteihn eine Anzahl Athener, womit eine VerringerunghungrigerM•indererreichtwurde(aucheinevergleichsweise verschwindend kleineGruppeAuswandererkonnteunterprek/irenVersorgungsverh/iltnissenins Gewicht fallen), anderseitskonnten athenische Handelsschiffenun unter geringerenSchwierigkeitenzwischenAthen und den Schwarzmeerh/ifenverkehren, was dem Handel im allgemeinenund der Getreideeinfuhrim besonderenzugutekommen musste(zumal Sigeion bereits durch die Verwaltung durch den PeisistratossohnHegesistratos Athen wohl nahestand).Ferner darf wohl angenommenwerden,dassdie nun beginnendeVerbreitungder erstenSilberm•inzen69(der sogenannten Wappenm•inzen)eine stabilisierendeWirkung aus•ibte:wenndie st/indig am Rand desRuins stehendenKleinstbauernnun in einemgutenErntejahr ein paar solcher Silberst•ickeerwerben konnten, wtirden sie in einem

schlechten Jahrdiesegegen daszumblossen Oberleben N/•tigeeintauschen k6nnen.

Die wohlvertrautendrei Perioden tyrannischer Herrschaftsaus•ibung durch Peisistratosstelltensichuns in neuemLicht dar. Im Archonjahr des

Komeas(561/ 60) wurdeeinemprominentengewesenen (und Areopagiten)eine Leibwachebewilligt--die aristokratischenGegenkr/ifte jedoch obsiegten,der Rechtsstaatwar der Unterlegene. Einer der Exponentenjener aristokratischenGruppen, denender Rechtsstaatnichts galt, inszenierte nach einer mittleren Zeitspanne die Heimkehr des Peisistratos,um seineeigenn•itzigenZiele zu f/Jrdern.Die Allianz scheiterte nachnicht allzulangerZeit, und Peisistratosbefandsichwiederumim Exil. Schliesslich,nach einemJahrzehnt, erzwangsichPeisistratosin einer wohlvorbereitetenmilit/irischen Aktion--in mancher Hinsicht vergleichbar der Kampagnejener Exiladligen, die sich 511/10 die R•ickkehr nach Athen erk/impften, wobei sie 1/3blicherweise auch noch den Tyrannen vertrieben--den Wiedereintritt nach Athen, nicht lange vor der EingliederungLydiensins persischeReich und dem Fall desletztenlydischen Mermnadenk6nigs Kroisos. Zwar konnte sich Peisistratosnicht r•ihmen, den Tyrannen vertrieben zu haben, doch zeigt die friedfertige und fruchtbare Entfaltung Athens, dass das politischeLebens Attikas von weitestgehendem aristokratischemKonsensgetragenwar. Ein Macht•ibergang dynastischenStiles beim To.d des Peisistratos (gegenEndedesAmtsjahresdesArchontenPhiloneos,im Fr•ihjahr 527TM) auf einen oder alle seine S/3hne1/3stsich vor dem Hintergrund unserer obigen Ausf•ihrung in ein historischesTruggebilde auf. Zun/ichst einmal hat sich wohl kaum etwas ge/indert.Da jedoch politischerFriede immer eher als eine Art politische Anomalie zu gelten hat, darf es nicht

36

K.H.

KINZL

verwundern, wenn der ffir Athen so heilsameZustand desGleichgewichtes der Krfifte schliesslichzu verfallen beginnenmusste.Dieser Auf16sungs-

prozesserstrecktesich fiber viele Jahre.TMEr begannm.E. in den spfiten zwanzigerJahrendes6. Jahrhunderts;die ErmordungdesHipparchoswar gleichsamdas ersteFeuerzeichenam Horizont frir dasschliessliche Fanal. Die nach dem Mord

an seinem Bruder entfachte

Schreckensherrschaft

des

Hippias hinterliessim kollektiven Bewusstseinder Athener ein Wundmal, das nie ganz verheilen sollte. Wir sind somit zum Ausgangspunktdes Athen gewidmetenAbschnittesunsererBetrachtungengelangt(s. oben S.

31 if.), der Kreis ist gewissermassen geschlossen. 'Die Tyrannisin Athen' ist ein Phfinomen des ausgehenden6. Jahrhunderts und trfigt Charakteristika, die bereits zukrinftigen Formen der Tyrannis eignen, wie sie einem Aristoteles

etwa wohlvertraut

waren.

Das kollektive

Trauma

des

Volkes der Athener, die politischePropagandades 5. Jahrhunderts,die philosophischeund politologischeAuseinandersetzung mit verschiedenen Typologiender Regierungsformen, ja die politischeErfahrungsweltdes4. Jahrhundertsriberhaupt--dieseMomente vereinensich in einem eigentrimlichenZusammenspiel,welchesschliesslich darin resultiert,dasssozusagendreidimensionaleGeschichtein zweidimensionaleTheorie verflacht wird, zum Zweck der Vertiefung des theoretischen Aspektes. Dem Verfasser stellt sich, was uns konventionellerweiseals eine 'verspfitete' Epoche einer langewfihrendenTyrannis mit dynastischemMachtribergang von Pallenebiszur Vertreibungder Peisistratidenmit zweivorangehenden Versuchender EtablierungeinerTyrannisangebotenwird (und wodurch-die zwei Exile mitgerechnet--ein halbesJahrhundertin Anspruchgenommen wird), vielmehr dar als eine bunte Kurve im Auf und Ab der Zwietracht der Adligen, die dann in die Tyrannis der Jahre bis 511/10

mfindet--jedoch nichtalseineepigonenhafte Formder•lteren Tyrannis, sondernvielmehr schon ein frfiher Vorl•iufer sp•iter sich auspr•igender Tyranniserscheinungen.

Es war unser Bestrebenaufzuzeigen,dassdieselbenGriechen,die sich etwa ein wahres Chaos an lokalbedingt verschiedenenG6ttern und Heroen zu schaffenwussten,die eszuwegebrachten,mit einemGeniestreichnicht

nur ein, sonderngleichein Dutzendodermehrphonetische Alphabetedem Abendland herbeizuzaubern, dass diese Griechen sich schwerlich in zwei

oder drei klar definierte Regierungsformen(etwa: Monarchie, Oligarchie, Tyrannis) formiert haben werden. Selbstverstfindlichkonnten wir in so engemRahmen nicht einmal im politischenBereichdie Vielfaltigkeit des archaischenHellas umspannen, noch hatte dies unser Bestrebensein drirfen.Doch hoffenwir, ebendiesebunteVielfalt deutlicheraufgezeigtzu

haben.UnsereTestfallesollenunsgezeigthaben,dassim geographischen Zentrum der griechischenWelt ein weites Spektrum politischer Lebens-

ZUR )•LTEREN

GRIECHISCHEN

TYRANNIS

37

formen Ausdruckfinden konnte. Die grossenErftillungender griechischen Geistesgeschichte fanden nicht in den peripherenGebietenstatt, sondern im Herzen Griechenlands--wo sich Athen zum Katalysator und Kristallisationspunktherausbildensollte. Die Einfltissefremder Hochkulturen waren die unentbehrlicheGrundlage ftir das Aufbrechen der Griechenin unergrtindeteRegionen des menschlichenGeistes,die 'Entdeckungdes Geistes'.Doch war es in jenem engen Raum, wo die Griechenganz unter sich waren (beeinflusst zwar, jedoch nicht beherrscht yon fremden Milchten), in einem Umkreis yon nicht mehr als etwa 100 km um den Isthmus yon Korinth, dass sich jener Kulturreichtum, der die Frucht verschiedenfiltigsterVielheit verbunden mit engstenKontakten ist, entfalten konnte, um schliesslichdie Statik des Adelszeitalterszu brechen. Es scheint denn auch verfehlt, Gesellschaftsformen,die sich in den Aussenregionen herausbildeten, 72 wie z.B. die territorial orientierte sizilische Militirdiktatur, oder die durch Lyder- bzw. Perserherrschaftden Griechen Kleinasiensdiktierte Gouverneurstyrannistiberall und unverindert auf dem Festland wiederfinden zu wollen--scheint es vor allem verfehlt, die

Ftille und Vielfalt desarchaischenFestlandshellenentums im politischen Bereich in die Zwangsjacke eines simplifizierenden,pseudo-juristischen Terminuszwingen zu wollen,nimlich des'Begriffes'der'Tyrannis',wo wir es doch in Wahrheit mit den verschiedenartigenManifestationendiverser Adelsr•gime zu tun haben. Trent University,Peterborough

K.H.

Kinzl

Anhang: Pap. Rvl. Gr. 18 Col. Iist weitestgehendzerst6rt. Dies erschwertdas Verstilndnis von col. II. Sicher zu lesen sind die Namen Chilon (als Ephoros bezeichnet), Anaxandridas, Sikyon, Aischines und Hippias; der Sachverhalt der Aufhebungvon Tyrannenherrschaftenist gleichfallssicherauszumachen. Plut. mor. 859 C-D handelt ebensovon Tyrannenvertreibungenund nennt ftir Sikyon einen Aischines. Das Scholion zu Aischin. 2,77 erwilhnt ebenfalls Austreibungenvon Tyrannen (es ist nicht ganz einzusehen,

weshalb)und schreibtbeztiglichSikyonsrobe $rr• KXzt•0•,•ou•. Dies ist verwunderlich,da die Scholiastenpsyche einenbestimmtenNamen (sei er auch noch so unbekannt)einer Gruppenbezeichnung vorzuziehenpflegt, und da kein ScholiastdieseSubstitutionvon sichausvorgenommenhaben wird (schonaus Unwissenallein). Wenn wir nun den Text desAischines,an den sich das Scholion kntipft, aufschlagen,lesen wir von spartanischen Harmosten. Sodann nennt Aischineseinen Spartanerals Nauarchen,der aber in Wirklichkeit

Harmost

war. Sein Name: Cheilon--so

alle Aischi-

neshandschriftensowie Harpokration s.v.; dagegenlesenwir Milon im

38

K.H.

KINZL

Papyrustextder Hell. Oxy. 6(1),3 und 8(3),1, welcheLesungEd. Meyer, TheopompsHellenika und Poralla, Prosopographieder Lakedaimonier, Nr. 535, beibehalten,w/ihrend Bruce, Hist. Comm. Hell. Oxy., die Frage offenl/isst. Ich m/Jchteernsthaft die M6glichkeit zur Debatte stellen, ob

nicht der Rylands Papyruseine arg verdorbeneWiedergabeeinesTextes darstellt, der zumindest teilweise aus einem Kommentar zu Aischines2,77 f. stammt. Hierbei w/ire der Name Chilon der desvon Aischinesgenannten Harmostendes4. Jahrhunderts;der Name Aischinesw/ire alsderjenigedes Redners zu betrachten, der sekund/ir in den Text eingedrungenw/ire.

Damit w/irde sich auch die Frage von selbererledigen, ob und wie der 'sikyonischeTyrann Aischines'mit den 'Orthagoriden'zusammenh/ingen k6nnte (vgl. H. R•dolph, Chiron I (1971) 75-83). Sovielstehtjedochfest: der ber•ihmte Ephoros Chilon und Anaxandridas waren keine Zeitgenossen;nur der letztere w/irde sich chronologischmit der Vertreibung desHippiasverbindenlassen,dochkeineder reichenQuellenbest/itigtdies; und schliesslich 1/isstsichdie stilistischeund intellektuelleKompetenzdes Schreibers

des Bruchstiickes

als bestenfalls

unterdurchschnittlich

beur-

teilen. Solangedie hier vorgetragenenFragen in all ihren Verzweigungen nicht v/Jlliggekl/irtsind,m/Jchteichvorschlagen, den RylandsPapyrusNr. 18 aus der Diskussionum die sikyonischeTyrannis und ihre Chronologie herauszuhalten.

NOTES

1. /3berarbeitete FassungeinesOriginalbeitrages in: K.H. Kinzl (Hrsg.),Die iiltere Tyrannisbiszu den Perserkriegen (1979)(WdF DX) 298-325.Esseihierden zweianonymenBegutachtern unddemHerausgeber dieserZeitschriftfar wertvolle Verbesserungsvorschl/ige gedankt. 2. D.h., wenn vollkommenidentischeHerrschaftsformenneuartigerPr•igung

unteransonstv611igverschiedenen/iusseren Bedingungen (unterFremdherrschaft in Westkleinasien; in Freiheit, aber unter voneinander ganz verschiedenen Umst•inden,auf dem Festlandund in Grossgriechenland) entstandenseinsollten, so bed•irfte eben dieserSachverhalteiner Erkl•irung,da er nicht als 'nat•rlich' bezeichnet

werden kann.

3. Richtungweisend H. Strasburger,'HerodotsZeitrechnung',Historia 5 (1956)129-161=ders. in: W. Marg(Hrsg.),Herodot2(1965)(WdF XXVI) 688-736 ('berichtigteund erweiterteFassung'). 4. In diesemeinenPunkt ist Strasburger(s. Anm. 3) nicht gen•igendpr•izise,

wie auchP.J. Rhodesin seinemvorz•iglichen Beitragzur Peisistratoschronologie (Phoenix 30 (1976) 219-233).

5. Lit. bei Berve518;ferneretwaA. Andrewes,CQ43(1949)70-78;ders.,CQ 1 (1951)39-45;N.G.L. Hafnmond,CQ 10(1960)33-36;G. Z6rner,Kypselos und Pheidon yonArgos(Diss.Marburg/Lahn1971)66 ff.; 129if.; R.A. Tomlinson, ArgosandtheArgolid(1972)81ff.; 188ff.;T. Kelly,A historyofArgosto500B.C.

ZUR •LTEREN

GRIECHISCHEN

TYRANNIS

39

(1976)94 if.; 112ff.; C.G. Starr, Theeconomicandsocialgrowthof earlyGreece (1977) 115; 186. 6. Man giesstdas Kind mit dem Badeaus, wenn man der ganzenEpisodeallen historischenWert abspricht, well sie novellistischeund Mfirchenz•ge trage (auf welcheFrage hier nicht eingegangenwerdenkann). Auch wenn die hier genannten noblenHerrn sichnie im Lebengetroffenhabensoilten,brauchenwir sienichtauch selberals unhistorischerklfiren;eshat noch niemandden Einfall gehabt, Solon als M•irchenfigur zur betrachten,well er unter den 'Sieben Weisen'aufscheint. 7. Der nicht mit dem Pausan.2,19,2 genanntenLakealas,Vater desMeltas, des letzten Herakliden-Temeniden aufdem Thron der Argiver (K/. Pauly s.v. Lakealas; Meltas), identifiziert werden sollte. 8. F. Jacoby,FGrHist 3 b Text S. 11(Einleitungzur Lokalhistorie•iberArgos). 9. Akusilaos: FGrHist 2 F 23-28 (argivische Stammb/iume); Hellanikos: FGrHist 4 F 1-5; 87-116 (Phoronis); 4 F 36 (Argolika). 10. Die Fragmenteder Lokalschriftsteller•iber Argos: FGrHist 3 B Nr. IX (304314, S. 6-25), mit Einleitungund Kommentar FGrHist 3 b Text S. 11-63;Noten S. 7-42.

11. 9. Jh.: Synkell. p. 499,5 ff. Dindorf=Theopompos,FGrHist 115F 393; Diod. 7,17 (Pheidon sei 10. Nachfahre des Herakles, 7. des Temenos); Marta. Par., FGrHist 239 A 30 (895/93). 8. Jh.: Strabon 8,3,33 C. 358=Ephoros,FGrHist 70 F ! 15 (10. NachfahredesTemenos);Pausan.6,22,2 (EinmischungO1.8 -748 v.u.Z.). J2. Strabon 8,6,16 C. 376=Ephoros,FGrHist70 F 176;Marm. Par. (s. Anm. 11); weitere Stellen FGrHist 3 C S. 86 f. (Kommentar zu 70 F 176). 13. Die bislangniedrigstenDaten erbrachtedie AuswertungdesAsyut-Hortes (s. M. Price& N. Waggoner,Archaic Greekcoinage:TheAsyut Hoard (1975)), und zwar ca. 550 ftir Athen, Korinth und Aigina (a.O. 122). Anm. 241 •iussernsichdie Autoren skeptisch•iber die Aufrechterhaltungder Datierung der erstenMtinzen tiberhaupt(die Elektronpr•igung),mit 640 als terminusante quem (L. Weidauer,

Problemederfrfihen Elektronpri•gung(1975)). 14. Die Inschrift IG 4,614 (vgl. SEG 11,336),die auf ca. 575-550(L.H. Jeffery, The local scriptsof Archaic Greece(1961) Argos,Nr. 7, pp. 156 ff.; 168; pl. 26) datiert wird, und deren Bedeutungunsicherist (es handelt sichum eine Liste mit denNamenvon neunPersonen,dieals8].0t.•op-t'o• .•F•¾0t•o,rro),brauchtunshier nicht weiter zu beschfiftigen:waren sie politischeBeamte, beweisensie dennoch nicht die Unm6glichkeiteinerK6nigs- oder Tyrannenherrschaft(vgl. etwa Spartas Ephoren und das ungest6rteFortbestehender Archonschaftin Athen zur Zeit des Peisistratos und seiner S6hne); waren sie religi6se Beamte, sind sie f•r Beweiszweckegleichermassenungeeignet. 15. Ich lege Wert auf die Feststellung,dassich nicht einen Pheidon in zwei verschiedeneGestalten aufspalte (nach der von Hammond, Bicknell und ihren Imitatoren ftir dieattischeProsopographiegetibtenMethode);vielmehrgilt es,sich klar zu vergegenwfirtigen, dassesimmer nur einenhistorischenPheidongegeben hat, den antikeAutorenmit demdanebenim Mythosangesiedelten K6nig Pheidon verschmolzen

haben.

16. Lit. bei Berve 532 ff.

17. Vgl. Strasburger(Anm. 3) 140=705. 18. O1.55(=560)erg/ibewohl ein zu argeszeitlichesGedr•inge;6,128,2scheintzu

40

K.H.

KINZL

suggerieren,dass die Kypselidenin Korinth noch fiber betr•chtlichen Einfluss geboten,dochistdiesnachHerodotsChronologieauchnochO1.56(-556) m6glich. Wenn diesesDatum als historischauthentischakzeptiert wird, ergebensichh6chst aufschlussreiche Tatbestfindeals 1ogischeFolge: 1) Die yon Peisistratosals Mutter weiterer Kinder verschmfihteTochter des Megakles, dutch deren Heirat die Rfickkehr des Peisistratos(nach Herodot 557/55; s.o. 33 ff.) besiegeItwurde, entstammteeiner friiheren Ehe des Megakles und war folglich keine Tochter der Agariste und Vollschwesterdes Kleisthenes;2) Kleisthenesyon Athen war der erstgeboreneSohn der 555 geschlossenen Ehe des Megaklesund der Agaristeund seinGeburtsdatumist auf Anfang 554 anzusetzen--sodasser zur Archontenwahl fiir 525/24 geradedas dreissigsteLebensjahrerreicht hatte. 19. Herodots Wortwahl schliesstdie M6glichkeit, dass wit die sechzigJahre nicht yore Todesjahr des Kleisthenes,sondern yon der Einffihrung der neuen Phylennamenan zu rechnenhiltten, zwar nicht aus, doch scheintmir die konventionelle Auffassungnatfirlicher. 20. Hdt. 6,92,2; vgl. K. Wickeft, Der peloponnesischeBund yon seiner Entstehunghis zurn Ende desarchidarnischenKrieges(Diss. Erlangen-Nfirnberg1961) 19.

21. A priori lfisstessichnichtausschliessen, dassder olympischeWagensiegund die Verm•ihlung der Tochter des Kleisthenesder antiargivischenReform nach Herodots Vorstellungen vorangegangensein k6nnten; anderseitsabet wfire es schwierig,eine solcheLesungdes Herodot reit schliissigenBeweisenzu stfitzen. 22. Ffir die Begrfindung roeinethiergefiusserten AnsichtsiehemeinenAufsatzin Klio 62 (1980) 177-190.Vgl. noch oben S. 34 mit Anrv. 66 unten. 23. Ffir Herodot stand Periander um 560 im Ni,.dergang (vgl. Strasburger (Anm. 3) 160--733);vgl. Anm. 18. Unter den modemen Historikern ist Beloch, GG 12,2,274ff., der prominentesteVertreter der niedrigstm6glichen Chronologieffir die Kypseliden:610-540. Arion und Perlander:Hdt. 1,23 f. 24. Hdt. 5,67,l; 69,1. Es ist hier nicht m6glich,aufzuzeigen,dassHerodot Punkt um Punkt die Parallelitfit herzustellenvermag(hierzu s.die Anm. 22 zitierte Arbeit,

'Exkursus').•berdies kanndie 5,69,1beobachtete Ionerverachtung desAtheners Kleisthenes

nut dutch

das yon Herodot--ffilschlich--konstatierte

Moment

der

Dorierverachtungdutch den Sikyonier Kleisthenes,das sich natiirlich nut gegen die Argiver richter, in den Bereich des Vergleichbaren geriickt werden--

Verachtunghier, Verachtungdort. •brigens m6ge dennochan diesetStelle vermerkt wetden, dassdie doch ffir jeden Kommentator befremdlicheBemerkung Herodotsx0•[o6zoq(sc. Kleisthenes)S:zeO[Sd),•"Ito¾0•q (5,69,1) Sinn gewinnt,wenn sie mit seiner,ebensoanscheinendverschleiertenNachricht, die cru¾¾eYe•q des

Isagoras06oum ... A[[ K0•o•, und dies sei alles, was er fiber die Familie des Isagoras wisse, in Verbindung gebracht wird. Die tiefere Bedeutung dieser Mitteilung erhellt aus 1,146,2(vgl. bes.Athenian Agora 3, Nr. 551). 25. Die sikyonischenReformenkonnten selbstverst•indlich nicht durchgeftihrt werden,w•ihrenddie Sikyonierund Kleisthenesim Feld standen(ebensowenig wie die kleisthenischenNeuerungenin Athen im Verlaufeiner milit•irischenKampagne ausgeffihrt werden konnten); es kann also nur ein allgemeiner Zustand der politischenSpannung(der in Krieg umschlagenmag) gemeintsein.Man beachte unter Umstiinden, dass Herodot diese Phrase dazu dient, den antiargivischen Grundton der Handlungen des Kleistheneszu motivieren (man mag sogarfragen,

ZUR .•LTEREN

GRIECHISCHEN

TYRANNIS

41

ob Kleisthenes--vergleichbar derPropaganda gegenAiginazurDurchsetzung der Flottenplfinedurch Themistokles--etwa selbsteine solcheKriegsstimmung kiinstlicherzeugte). 26. S. oben. Anm. 22.

27. EssolIrewohlaucheinmilit/•risches Momenthierbeimitspielen--wie wohl auchim Fallderathenischen Phylenreform (eingew/Shnlich vernachl/•ssigter bzw. iibergangener Aspekt).Ichselbsthabediesg/•nzlich unbeachtet gelassen in meinem Aufsatz'Athens:Betweentyrannyand democracy'in: Greeceand the Eastern Mediterranean .... Studies... Fritz Schachermeyr (1977) 199if. 28. Berve 29.

29. Vgl.z.B.denb/Jotischen Heroennamen Hyas;jedochvgl.gleichfalls denmit Dionysosin Zusammenhang gesetztenHyes(s. RE 9, 88 ff.) (freundlicherHinweis von H. Schwabl). Choireai ist als Ortsname belegt, auch wenn sich, wie ftir die Oneatai, kein Personennameunmittelbar anbietet. Im tibrigen sollte die

Emendierung von Hdt. 5,68,1durchSauppe(•r•[ wegendesDoppelsinnes yonZo•Ooq nichtbedenkenlos in unsereStandardtextedes Autors tibernommenwerden.Der 'Hohn' im Namen Choireatai liegt m.E. nicht in der Assoziationreit dem Tier (wasauf eine Duplizierungyon Hyatai hinausk/•me); s. bes. Suda X 601 Xo•po½... r•p& Kopt¾0•ot•8• r6 Tuv•tze•ov•[So•ov.Es sei schliesslichnoch darauf hingewiesen,dasssichaus Herodot wederdie Existenzyon vier Phylen vor der Reform desKleisthenesklar ablesenliisst,noch mit Sicherheit auf vier kleisthenischePhylen geschlossen werdenkann (es m/Sgenmehr gewesen sein). Eine letzte Bemerkungnoch: der Athener Kleisthenesliesssich die neuen Phylennamenyon Delphoi sanktionieren(Aristot. Ath. pol. 21,6); hatte man in Sikyon diesen Schritt verabs/•umt? 30. Pausan. 6,19,1 f.

31. Aristot. pol. 1316a30;Diod. 8,24.

32. Pap.Ryl.Gr.18.Lit.Berve 531;zuletztD.M. Leahy,Historia17(1968)1-23. 33. Plut.

mot.

859 C-D.

34. Schol. Aischin 2,77. 35. S. Anhang. 36. Diod. 8,24; Plut. mot. 553 A-B; Pap. Oxy. 1241,col. IIl. Vgl. KI. Pauly4,363

s.v. Orthagoras 2. 37. Diod. 8,24; Plut. mot. 553 A-B; Pap. Oxy. 1365 (=FGrHist. 105 F 2=551 F lb). 38. Aristot. pol. 1315b12; 1316a30. 39. Menaichmosyon Sikyon, FGrHist 131 F 1 (Sikyoniaka); F 2 (Pythikos). F. Jacoby,FGrHist 3 b S. 476, zur sikyonischenLokalgeschichtsschreibung. Vgl. 131 T 3 (Aristot. Pyth.). 40.

Berve 305.

41. Es ist objektiv richtig, dassein Opferschliichternicht niedriger Herkunft gewesensein k6nne, doch deutet die Hasstradition diesdennochnicht anders(Pap. Oxy. 1365=FGrHist 105 F 2=551 F lb; vgl. Diod. 8,24). 42. Pap. Oxy. 1365=FGrHist 105 F 2=551 F lb; vgl. z.B. Nikolaos yon Damaskos, 90 F 57 (Kypselos). 43. Vgl. Aristoph. Ekkles. 916; auch V. Pisani,Paideia 13(1958) 143.S. bereits A. Momigliano, A&R 10 (1929) 153. 44. Nikolaos, 90 F 61.

42

K.H. KINZL

45. Vgl. etwa die sch/3neSymmetrievon BervesStemma (Berve 758). 46. Die kriegerischeNote, die bei Herodot anklingt und von Aristoteles(pol. 1315b17)herausgestelltwird, setztsich in der Einbeziehungdes Kleisthenesin die Vorg/ingedessog. Ersten Heiligen Krieges(hierzu bes.G. Forrest,BCH80 (1956) 33 if. mit den Stellen)und der Zuweisungder Zerst/3rungdeswestlichenNachbarn Pellana (Anaxandridasvon Delphoi, FGrHist 404 F 1; Pap. Oxy. 1241,col. III) fort. Inwieferne in beiden F/illen echte historischeKunde zugrunde liegt, vermag ich nicht zu entscheiden.Pellanaist in der Tat einmal zerst/3rtworden, und diesmag den Sikyoniernanzulastenseinundaufdie Zeit desKleisthenes datiertwerden;zu beweisenist es nicht. Der sog. Erste Heilige Krieg mit der Vernichtung von Kirrha/Krisa durch Kleisthenesist schonstark mit Legendenhaftemdurchwoben (gleichdem sog.LelantinischenKrieg);meineeigeneZuversicht,dasswir hiernoch viel authentischhistorischesMaterial vor uns haben, ist gering. 47. Lit. Berve 539 ff.• vgl. F. Schachermeyr,RE 19,156if. K.H. Kinzl (Hrsg.), Die •iltere Tyrannis(1978) 94 ff.; ders.,RE 19,150ff. Das Problemder Chronologie desPeisistratosist ein vielverhandeltesThema; zur herodotischenChronologievgl. Strasburger(Anm. 3), sowieganzbesondersRhodes(Anm. 4), dessenErgebnissen ich reich im folgenden anschliesse. 48. F. Schachermeyr,'Athen als Stadt des Grossk/Snigs', GB I (1973) 211 ff.: Forschungenund Betrachtungen(1974) 75 ff.; ders.,Die Siegerder Perserkriege (1974) 19. 49. Vgl. meine Ausffihrungena.O. (Anm. 27), passim. 50. Vgl. Schachermeyr(Anm. 48) 219:83.

51. Hdt. 1,59,4:• •,• •p6• MzT0•p•0•Tz•oF• •s'rp0•'r•T• •. Esistausgeschlossen, dassPeisistratosnie Mitglied des Areopags gewesenw/ire, w/ihrend es umgekehrt doch unwahrscheinlichschiene,wenneineeponymeArchonschaft,in der Archontenliste offiziell verzeichnet,absolut spurlosvergessenworden w/ire. Auch dem Alter nach wiirde sich dies gut einfiigen, vgl. Anm. 52. 52. Zum Zeitpunkt vgl. etwa Berve 544. 53. Aristot. Ath. pol. 17,1, und zwar vermutlich nach der Bestimmung der Archontenfiir dasfolgendeJahr,alsoim Fr/.ihjahr527;s. M.E. White, 'Hippiasand the Athenian Archon List', Polis and Imperium, Studies... E. T. Salmon (1974) 81 if.

54. K.J. Beloch, RbM45 (1890) 469 f.; GG 11,327f. mit 328 Anm. 2; GG 12,1,368 if.; 12,2,288 if. Heute darf man diese Hypothesen Belochs als iiberwunden betrachten.

55. Die Ausdrucksweise Herodotsin seinerDarstellungder Ereignisse,die im Jahr der Archonschaftdes Komeas(Aristot. Ath. pol. 14,1,561/60 v.u.Z.) einen H/Shepunkterreichen(1,59,3 ff.), scheintmir starke Ankl/ingean seineWortwahl 5,66,1 zu verraten. Dies ist nur allzu nat/.irlich,sobaidman sichvergegenw/irtigt, dass er fiber Ereignisse,die zwar durch ein halbes Jahrhundert voneinander getrennt sind, unter ein und demselbenGesichtspunkt(letzlich doch eben auch 'anachronistisch')berichtet. 56. Zwar wird sich keine Regierung in Athen der Kontrolle der Akropolis begeben,doch ist esein ungleichschwierigeres Unterfangen,mit der Besetzungder Akropolis beginnendKontrolle •iber die Stadt und ganz Attika zu gewinnen.Was auch immer der ungliicklicheKylon eigentlicherreichenwollte, seinund seiner Gef/ihrtenScheiternexemplifiziertdiesenSachverhaltund erweistihn alsTatsache;

ZUR •LTEREN

GRIECHISCHEN

TYRANNIS

43

man miisstePeisistratosSchwachsinnigkeit unterstellen,um sicheinzureden,er h/itte diesnicht gesehen.SeineAbk/Smmlinge fliichtetensich511/10 dorthin, kurz darnach Isagoras und Kleomenes--um binnen kurzem gedemiitigtdas Feld zu r/iumen. Wenn in Aristophanes'Lysistrata (freundlicher Hinweis von A.E. Raubitschek)die in SexstreiktretendenAthenerinnendie Akropolis besetzen,so dochweil siedort ebenam unzugiinglichsten seinwfirden,abgesehen davon,dasses sichum irreale Komik handelt. Einem anonymenBegutachtermeinesAufsatzesffir diese Zeitschrift verdanke ich den Hinweis auf einen weiteren Vorfall (ca. 298 v.u.Z.; siehe FGrHist 257a F 2:Pap. Oxy. 2082), der ebenfallsmeine Auffassung sch/Sn illustriert.

57. Die Taktik bestehtdarin, dassman sichder Unterstiitzungeinerhinl/inglich grossen Anzahl von adligen H/iuptlingen versichert, die ein jeder eine Anzahl yon Stimmen ihrer kleinen Gefolgsleutein die Volksversammlungmitbringen; vgl. meine Ausfiihrungena.O. (Anm. 27) 201 mit Anm. 11 sowie Gymnasium 85 (1978) 314. 58. Die 'Hyper-' (Hdt.) oder 'Diakrioi', eines der meistverhandelten Themen, vgl. Berve542; E. Kluwe, Klio 54 (1972) 101if. Fiir unsereEr/Srterungen l/isstsich eine Stellungnahmehierzu vermeiden,ohnedassunsereBeweisfiihrung beeintr/ichtigtwiirde. 59. KI. Pauly 4,735 f. s.v. Philaidai 2. 60. Das dauernde Verbleiben einer S/51dnertruppem/Schtemir fragwfirdig scheinen. W/ire eine solche durch Sold entlohnt worden, h/itte dies eine stete finanzielle Belastungder Privatfinanzen des Peisistratosbedeutet. Wfiren ihnen ihre DienstedutchZuweisungvon Land vergoltenworden,hfittePeisistratosdamit

eine st•ndige Quelle der Zwietracht gestiftet. Der Umstand, dass Hippias offensichtlichsichwenigstenszum Teil auf S/51dner stfitzenmusste,unddasseinige in Athen verblieben sein k/Snntenund dann im Verlauf des sogenanntenDiapsephismos 'ausgebfirgert'oder'eingebiirgert'wordenw/iren,istdurchausm/Sglich, darf uns aber nicht zu vorschnellenRiickschlfissen auf die hier in Frage stehende Zeit verfiihren. Es scheint bezeichnend,dass die Bliitezeit der mit skythischen BogenschiitzenbernaltenVasen ca. 530 bis 490 anzusetzenist; vgl. M.F. Vos, Scythianarchersin archaic Attic vase-painting(1963). 61. Hdt. 6,35,1; die Darstellungschliesstsichnahtlosan 1,64,3an. Man beachte anderseitsdie parallele Ausdrucksweise6,35,1•5•66,1, mittels welcher Herodot

sehrgeschickt die,•hnlichkeiten undVerschiedenheiten derjeweiligen Situationen sichtbarmacht (fiber diesesStilmittel s. RhM 118 (1975) 193 ff.). 62. Esgehtaus HerodotsBerichtfiberdie Kolonisierungder Chersonesos durch

Miltiadesden ,•lteren v/511ig unbezweifelbar hervor,dassdieseVorg•ngesich unmittelbarnach Palleneabspielten(zur Kompositionstechnik s. Anm. 61); sie lassensich unm/Sglichin Herodots Darstellungder Umst/indeder sogenannten erstenoder zweitenTyrannenherrschaften einfiigen.Fiir Herodot istesein Faktum gewesen,dass Kroisos seineErkundigungenfiber Athen bald nach dem Zusammenstossbei Palleneeinzogund darnachnochlangegenugjener m/ichtigeK6nig

war,alswelcher erauchMiltiadesden,•lterenausdenH/indenderLampsakener zu befreien vermochte(Hdt. 6,37), ob dies nun zu unsererabsolutenChronologie stimmenmagodernicht.Gewiss,wennmanvonderArchonschaftdesHarpaktides als Fixpunkt (d.h. 511/10) 36 Jahre (Hdt. 5,65,3) zuriickrechnet,so entstehtein arges Gedr/inge um 546 herurn. Doch stand dem Herodot dieset, in absoluter

44

K.H.

KINZL

Chronologiefixierte, Zeitpunkt nicht zu Gebote; ob 36 Jahre als mathematisch exakte Bestimmungder Zeitspanne,oderetwanur alsrundeZahl (drei Dutzend)zu verstehensei, ist durchausnicht yon selberersichtlich(wobei zu bemerkenwiire, dassbei Herodots Art und Weise, zeitliche Distanz zu bezeichnen,man selbstbei Zurfickrechnenab dem Archonschaftsjahr des Harpaktidesunter Akzeptierung der Zahl 36 auf--in unseremZeitrechnungssystem ausgedrfickt--548/47kommen kann); schliesslichist zu bedenken, dass in der ausserherodotischenT.radition Archontendatenffir die Vertreibung der Peisistratiden,ffir den Tod des Peisistratos, nicht aber ffir Pallene aufscheinen.Angesichtsdieser Sachlageist es angebracht,HerodotsBerichtunangetastetzu lassen,aber anzuerkennen,dassdie yon ihm geboteneJahresanzahl36 als u.U. um bis zu zwei Jahre ungenausein k6nnte--eine wahrhaftminimaleUngenauigkeit.Berufungauf absoluteDaten,die grossteils ausserhalbHerodotsgewonnenwurden,widerlegendiebiervorgetragene Ansicht nicht (vgl. etwa Berve 565). 63. Hdt. 6,103,4. Miltiades der Jfingerekann schwerlichnach 555/54 geboren sein,da er 524/23 eponymerArchon war (Dionys. Halikarn. 7,3,1; $EG 10,352). 64. Die drei Siege des Kimon werden gew/Shnlichzwischen 536/32/28 und 532/28/24 hin und her geschoben;vgl., ffir letztere Datierung, j fingstM.E. White (Anm. 53) 87 f. Indesist esv611igunzutreffend,zu vermuten,dassKimon kurz nach seinem dritten Sieg aus dem Weg gerfiumt worden sei. Herodots Ausdrucksweise legt keinerlei derartige Ausdeutungnahe (6,103,3; man vgl. z.B. die Formulierung 5,71,2 betreffend die Zeit der Kylonaffiire); zeitliche Niihe suggerierter bloss hinsichtlichder Entsendungdes Miltiades des Jfingerenauf die Chersonesos in Beziehungauf die Ermordung des Kimon (dessenimmer wieder aufgetischter 'Beiname' Koalemos ganz gewisserst der iilteren Kom/Sdiezu verdanken ist; vgl. Verf., Miltiades-Forschungen(1968) 21 f.). 65. Vgl. A. Lesky, Die tragischeDichtung der Hellenen3 (1972) 49 f. 66. S. Anm. 22. SelbstBerve(60) musseinr•umen, dasssicheineengeBeziehung

desPeisistratos zudionysischen Belangen in derL•berlieferung nichtbelegen lasse. 67. Es ist •brigensnicht ausgeschlossen, dassderartigeCharakterziigeseines iiltesten Sohnes dem Peisistratosnicht verborgen gebliebensind, under ihn m6glichstvom /Sffentlichen Leben fernzuhaltensuchte:es ist jedenfallsh/Schst auffallend,dassHippiaserstnachdemTod seinesVatersArchonwurde,undzwar, sobaider es nur konnte, worauf White (Anm. 53) jfingst mit Nachdruck und zu Rechthingewiesen hat; seinCharaktererscheintauchsonstin merkwfirdigemLicht durch seinefiber das gewohnteMass hinausgehendeGliiubigkeitan Orakel, Seher, Vorzeichenund Triiume (vgl. z.B. Berve 557 ffir Belegstellen). 68. Auch wenn wir die M/Sglichkeit,Peisistratoshatte sofort nach PalleneLand der geflfichtetenAdligenverteilt, und diesehiittendasnachihrer HeimkehralsFait accomplistillschweigend akzeptiert,nicht a priori verneinenk/Snnen,will diesdoch recht unwahrscheinlich

erscheinen.

69. SieheAnm. 13zur Datierung.Ihr wahresPotentialfreilichwfirdendie neuen Silberstficke erst ein Jahrhundert darnach erreichen, im 'Attischen Reich'.

70. Vgl. M.E. White (Anm. 53) 84. 71. Die Periodevom Tod desPeisistratosbis zur VerbannungseinerS/Shneist noch weitestgehendunerforscht--trotz der nahezu un•bersehbaren Literatur, besonderszur Tyrannenm6rderfrageund zum Exkurs des Thukydides(6,53,3-

ZUR •LTEREN

GRIECHISCHEN

TYRANNIS

45

60,1). EinigeAnregungenerfahrenwir bereitsvon Archfiologenseite, ausdenensich zu ergebenscheint,dasserst nachdem Tod desPeisistratos eineAusweitungder Bautfitigkeitstattfand,vgl. z.B.E. Kluwe, Die Tyrannisder Peisistratidenund ihr Niederschlagin der Kunst (Diss. Jena 1966);Material bei J.S. Boerstoa,Athenian buildingpolicyfrom 561/0 to 405/4 B.C. (1970) 11-27,undbes.seinKatalog123if.; kein Fortschritt bei J. Kleine, Untersuchungenzur Chronologieder attischen Kunst yon Peisistratos bis Themistokles(1973) 13-56, wegen der heillosen Vermengungvon arch•iologischer und historiographischer Evidenzreit Geschichte als solcher.Ans•itzewerdenangedeutetvon F.J. Frost, AJAH I (1976) 66 if.; vgl. noch White (Anm. 53). Die ForschungmusssichmeinesErachtenserst nochvom Zwang l/3sen,alles, was sich in jenen Jahren zutrug, den Peisistratidenzuzuschreiben; vgl. obenS. 34 f. reit Anm. 66. Das bertihmteFragmentderArchonliste (SEG 10,352;vgl. Bradeen,Hesperia32 (1963) 187ff.) scheintmir denbeginnenden Verfall zu markieren, im Gegensatzzur g•ingigenAnsicht, es dokumentiereeine eben erst erreichte Vers/3hnungnach dem Tod des Alten Tyrannen. 72. Vgl. oben Anm. 2. 73. Die

hier vertretenen

Gedanken

wurden

zuerst im Mai

1977 in Wien

vorgetragen.Dem Freundund einstigenLehrer, Fritz Schachermeyr,verdankeich wertvolle Bemerkungen.Summarischdanken m/3chteich ftir vielf•iltigepositive Kritik und Anregungenanl•isslichyon Vortr•igenin Bangor(Wales),Berlin(West), Frankfurt, Halle, Konstanz, Leeds, London, Ottawa (im Rahmen der Jahrestagung der Associationof Ancient Historians) und Prag.

THE

DATE

OF THE

FOUNDATION

OF ALEXANDRIA

PierreJouguet,four decadesago,pointedout that thedateof Tybi 25 given by Pseudo-Callisthenes (I 32.10)for thefoundationof Alexandriapointed to the julian date of 7 April. • He found reasonsto regard this date as uncomfortablefrom the point of view of the chronologyof Alexander's whereabouts in 33 l, but he had no doubt that the date was one transmitted

in Alexandrian tradition from the earliesttimes. This date, 7 April, was advancedby C. BradfordWellesas beingthe correctdate of the founding of the city; Wellesusedthe date asa powerfulargumentfor his thesisthat Alexanderfoundedhisname-cityafter the visitto the oracleof Ammon at Siwah, and not before that trip.2 Wellesremarked that Ulrich Wilcken had "oddly enough"arrived at a date of 20 January by usingthe equation betweenthe Egyptian and julian calendarscurrent in Egypt after the reform of the calendarby Augustus,which halted the lossof one day in every four yearsby adding a sixth epagomenalday in the julian year precedingthe julian leap year (i.e., bissextileyear).3 Jouguet's(and Welles')calculationof the date hasbeenacceptedmore recentlyby thosewhohavefollowedWelles'reconstruction of thefounding

of Alexandria? butrejected withoutserious thoughtbythosewhoarenot convinced follows: 5

of Welles' views on the foundation.

P.M.

Fraser writes as

... the explanationis surely that Wilcken, more cautiousin this respectthan hiscritic, assumedthe precisedategivenby Ps.-Call.to be of Roman origin.

A similarviewisexpressed by F. AltheimandR. Stiehl,whowrite:"dass das Datum erst nach dem unter Augustuseingeft•hrtenKalendar zu berechnen ist,demnachaufden20Januarfiel."6A morebelligerent version of thisviewis alsoexpressed in Robin Lane Fox'sbook on Alexander: 7

"[Welles'] argument thatthefoundation dateof 25 Tybimeans anApril date is easilyrefutedby the likelihoodthat Ps.-Call.usedthe Roman calendar,mistakenly,rather than the Ptolemaiccalendar,as Welles assumes." 46

FOUNDATION

OF ALEXANDRIA

47

This is not a trivial matter of arithmetical calculation. It involves, rather,

somebasicassumptions about chronologicalreckoningin antiquitywhich cannot be right. It doesnot seemthat thesescholarshaveexaminedclosely what procedurethey supposethe author of the Alexandrian Romanceto have followed.

The Romance

tells us that the date of the foundation

was

Tybi 25; the view of Jouguetand Wellesis that this date on the Egyptian calendar was known and remembered

as the actual date of the foundation

in 331. It is, indeed, evident that the citizens of Alexandria celebrated on that date a festival which in the writer's time was still regardedas being

connectedto the foundationof the city and that of the Heroon of Agathos Daimon (Ps.-Call. 132.10-13).8This date, Tybi 25, fell on 7 April in 331Be, but becausethe Egyptian year lacked a leap year in this period, it had moved back to 20 January by the time of Augustus'calendarreform in 26/5.9 The other view, that of Wilcken, Fraser et al., assumesthat the date was

calculated(as Altheim and Stiehl say explicitly) rather than remembered. This view necessitatesthe assumptionthat the date was figured in the Roman period. But what can thismean?Calculationmakessenseonly if the date wasknown in anothercalendarwhichprovidedan objectivepoint of reference--thejulian, in other words--and wasconvertedfrom thisinto the Egyptiancalendarafter Augustus'calendarreform. But this is manifestlyimpossible.It is preciselyin the Egyptiancalendarthat any date must have been transmitted; no one in Alexandria in 331 can have known the fixedjulian calendar,asit wasnot createduntil almostthreecenturieslater. It is clear that in later Roman times, when recensionsB and P of the

Alexander Romance were composed, their authors proceeded in the reverse direction and arrived--like their modern followers--at January. Jouguetquotesfrom F asfollows:to'r• r•67.•¾&••'r• •:•th• 6 AXE•0•8•o• x•OiSpur•vT•[3••'ro• 'I•vvupi•, VOU•'•vi •. Clearlythewriterknewthat the date was in Tybi (though recensionA's date of the 25th has by now been lost),and he equatedthisto January;suchwascommonpracticein Egypt, for the Egyptianmonthto beequatedto themostnearlycoincidingRoman one.• But it would defy all logic to supposethat an Alexandrian writer receivedthe tradition in the reverseform, a foundation in January, which he managed to equate to Tybi. The onlyotherway of savingJanuarywouldbeto imaginethat the writer calculated the number of days (77, actually) by. which the Egyptian calendarhad slippedagainstthe solaryear sincethe time of Alexander,and then changed the date to suit; in other words, that the author of the Romanceknew that the true date in 331 was Hathyr 8 and calculated77 days back to reach his Tybi 25; but the information about the festival makesthisimpossible.We know from the CanopusDecree(0GIS56)that the Egyptianswereonly too awareof the problemcausedby the regression of their calendar,and that in particularit causedthe principalfestivalsto

48

ROGER

S. BAGNALL

fall at the wrong times of year. But even though an effort (abortive, it seems)was madein Euergetes'reignto correctthis problem,no attempt is recordedto compensatefor the shiftsby altering datesto take accountof pastregression. •2Indeed,as Jouguetsays,we haveno reasonto think that ancient writers were capable of the complicated analysisnecessaryfor making such changes,•3 nor that a writer like the author of the Romance would have cared about such a technical detail. I am not aware of any instancein all of antiquity of the later recalculationof the date of an historicalevent becauseof calendaricproblems. It is evident, then, that the date of Tybi 25 is what washandeddown and taken by Pseudo-Callisthenes as the date of the foundingof Alexandria. The only relevantquestion,then, is on what day Tybi 25 fell in 331 BC.The answeris 7 April, and that is the Alexandrinedatefor the city'sfoundation. Whether it is correctis another matter; the partisansof January,confident in their erroneous remarks about the julian date of Tybi 25, have all seeminglyacceptedthe testimonyof the Romance,and indeedI think the Romance'sTybi 25 is correct.Jouguet,taking Arrian's remark (3.6.1) that the expeditionleft Egypt in early springas being a precisetime, found himselftroubledby the April foundationdate(aswereAltheim and Stiehl). I doubtthat Arrian's6•z0•-r• •p• 6rro•0d,•o,•v• is to be pushedfor precision, and Engels has shown that a departure for Syria in late April from Memphisis perfectlycompatiblewith the remainderof Alexander'smarch to Mesopotamia.24 It is of coursepossiblethat Alexandervisitedthe siteof Alexandria and formulated his plansbefore visiting the oraclefor authorization; but the Tybi 25 date in no way can be usedas evidencethat this previewtook placein January,for Tybi 25 in 331 BChasnothingto dowith January.•5 Whatever answerone arrives at in the end, it will not do to misusethe calendaricproblemto explain away the 7 April date of PseudoCallisthenes. 16

Roger S. Bagnail

Columbia University NOTES

1. REA 42 (1940) 192-97. 2. "The discoveryof Sarapisand the foundation of Alexandria", Historia 11 (1962) 271-98 at 284. 3. For Wilcken's views, seeSitzbBerlin 1928(30) 579 n. 3; Welles notesthat H. Riemann, RE 18 (1949) 1567-69 s.v. Parmenion, followed Wilcken. 4. E.g., Eugene Borza, "Alexander and the return from Siwah", Historia 16 (1967) 369 and in his note on p. 355 of the Norton paperbackedition (New York 1967) of Wilcken'sAlexander the Great (transl. G.C. Richards);Donald W. Engels, Alexander the Great and the logisticsof the Macedonianarmy (Berkeleyand Los Angeles1978) 63 n. 54, who for somereasongives8 April as the date. J. Seibert,

FOUNDATION

OF ALEXANDRIA

49

Alexander der Grosse(Darmstadt 1972) 113 reports Welles' date as 27 April, presumablyby a misprint. 5. PtolemaicAlexandria (Oxford 1972)II 3 n. 9; cf. Op. Ath. 7 (1967) 30 n. 27. Fraser quotes the passagefrom Ps.-Call. in extenso. If I understand Fraser correctly,he is not making the (unlikely) assertionthat the Egyptian date itselfwas invented in Roman times or by the author of the Romance. 6. Die Araber in der alten Welt II (Berlin 1965) 13. 7. Alexander the Great (New York 1973) 522. 8. Jouguet (supra n. 1) 192. Cf. the modern editions:L. Bergson,Der griechischeAlexanderroman. Rezension B (Acta Univ. Stockholm., Studia Graeca Stockholm. 3, Uppsala 1965) 50; U. yon Lauenstein,Der griechischeAlexanderroman, RezensionF, Buch 1 (Beitr. z. kl. Phil. 4, Meisenheim 1962) 112. 9. See Wilcken, Grundzfigeund Chrestomathieder PapyruskundeI 1 (Berlin 1912) lv and T.C. Skeat, The reignsof the Ptolemies (Mtinch. Beitr. 39, Munich 1954). 10. Cf. supra n. 8.

11. Cf. R.S. Bagnailand K.A. Worp, Chronologicalsystems of ByzantineEgypt (Stud. Amst. 8, Zutphen 1978) 22. Jouguet supposesthat the text is an error for Tybi 25 = January1,but that isunlikely,giventhenatureof suchequations,andhis attempts to calculate a date (p. 193) on this basisare not useful. The writer of F does, however, give Tybi 25 as the date of the festival a few lines later. 12. On this calendar change under Euergetes,cf. A.E. Samuel, Ptolemaic chronology(Mtinch. Beitr. 43, Munich 1962) 138. 13. Jouguet (supra n. 1) 193. 14. Engels (supra n. 4). Similar vague referencesto the start of spring are characteristicof Arrian'sdescriptionsof the startsof campaigningseasons and need not have any precisereference.Cf. K.J. Beloch,GriechischeGeschichte 2III (BerlinLeipzig 1923) 313-19 for details. 15. Arrian (no doubt following Ptolemy) putsthe foundation beforethe visit to Siwah;so far as the formal foundation goes,this is wrong. But if he is referringto the preliminary scoutingvisit rather than the actual foundation, he may well be right. He givesno date for this first visit. 16. Fraser (supra n. 4) cites without comment Skeat's referenceto the 'Horoscope of Alexandria' in the Catalogus Codicum Astrologorum IX 2: Codices Britannicos.... ed. S. Weinstock (Brussels1953) 178, where the editor writes: "Alexandria conditaerat anno 330 a.Chr.n.... atque thema nostrum,ut benevole me docet v.d. Otto Neugebauer,recte ad hunc annum ac diem 16 mensisAprilis refert..." Weinstockis wrong about Alexandria'shaving beenfounded in 330, but ProfessorNeugebauerkindly tellsme that he iscorrectlyquotedthat the horoscope doesrefer to 330. The month and day thuscomecloseto Pseudo-Callisthenes'date, but the year is wrong by one. One shouldnot pressthis sourcetoo far, because,as Weinstocknotes,the datesfor severalothercitiesarequitewrong(cf. pp. 176-177).

THE RE-ESTABLISHMENT

OF THE BOIOTARCHIA

(378 BC)

In 386 Agesilaosusedthe autonomyclauseof the King'sPeaceto justify the dissolutionof the Boiotian Confederacy. • Or. in the wordsof Xenophon. who washardlyan admirerof Thebes.the Spartans•6zo.6•o.

©•

z&• Bo•t,ziS• :zd7.z••:zo•¾.

One neednot, however,take

Xenophon's wording as evidenceagainst the existenceof a Boiotian Confederacybeforethe King'sPeace.SinceXenophonwasquitereluctant to acknowledgethefederalprinciplein Boiotianpolitics,heroutinelyrefers insteadto Thebandominationof Boiotia.2 Perhapsthe closesthecomesto admitting the existenceof federalismin Boiotia is in a speechwhich he ascribesto Kleigenesof Akanthos(Hell. 5.2.16), in which the ambassador in 383 remindedthe Spartansof their carenot to allow Boiotiato beunited. Xenophon's interpretation of the Boiotian Confederacyas a Theban hegemonyis tendentious,as is proven by the Hellenika Oxyrhynchia, whichgivesa lessbiasedand moredetailedpictureof the Boiotianfederal government and the state of political affairs in Boiotia before 386.3 Furthermore, the Oxyrhynchoshistorian indicatesthat the constitution that hedescribes no longerexistedwhenhecomposedhiswork, and thushe providesevidencenot only for the (temporary) end of Theban preeminence in Boiotia but also for the dissolution

of the Boiotian

federal

government.4

Naturally, the dissolutionof the Boiotian Confederacyin 386 entailed the abolition of the boiotarchia,the chief executivemagistracyof the federalgovernment.Thereafter,eachBoiotianpolis,throughtheagencyof its own magistrates,pursuedits own policy. The only local government about which anything is known is that of Thebes, but at least here Xenophon and Plutarch, the two principal sourcesfor this period of Boiotianhistory, are in agreement.Xenophonconsistentlyrefersto the executiveofficialsat Thebesin the years386-379 aspolernarchoiand a secretary. 5 He relatesthat the polemarchoihad police powerswhich includedthe right to arrestand to arraigncriminal suspects, and that they conductedbusiness in a publicbuilding,thepolemarcheion. 6In hisaccount

50

RE-ESTABLISHMENT

OF THE BOIOTARCHIA

51

of Phoibidas' seizure of the Kadmeia and its aftermath, he mentions that

Leontiadesand the pro-Spartanfactionmadeno changein Thebaninstitutions, but that they contentedthemselveswith selectinganotherpolernarchosto replaceHismenias,who was bound over for trial.? In his PelopidasPlutarch alsomaintainsthat the executivefunctionsof local Theban governmentwere in the hands of three polernarchoi and a secretaryin the yearsbetweenthe King's Peaceand the liberation of the Kadmeia in 379/8. 8 So too in the Agesilaos(24.2), where he statesthat althoughLeontiadesand Archiaswerepolernarchoiin name, they were tyrants in deed. This sentimentis also found in Xenophon'sHellenika (7.3.7), where Xenophon ascribesa speechto one of the assassins of the Sikyoniantyrant Euphron,in which the assassin accusedthe followersof Archias and Hypates as endeavoringto act as tyrants.9 In Pelopidas6.2 Plutarch states that as a result of Phoibidas' coup the Thebans were deprivedof their ancestralconstitution.Clearly,in the opinionof Plutarch and Xenophon, then, the ascendancyof Leontiades'factionentailedonly a changeof leadershipand not a changein the Theban constitutionor a curtailment of the polernarchia.•o Although there is virtually no evidenceto corroboratePlutarch'sassertion as to the nature of the Theban ancestral constitution before 382, the view that the polernarchiawas a venerableoffice has been defendedby H. Swoboda?• Cloch• (74), Schaefer (1110), and Roesch (162), all of whom freely admit that Theban polemarchoi are first mentionedonly in the fourth century.•2 Schaefersuggestedthat originally the polernarchoi were the king's assistants,entrustedwith the military leadershipof the citizenlevy,a hypothesis whichis at leastconsistent with politicaldevelopmentselsewherein the Greek world.•3 From their first appearancein the sourcesof the fourth centurydown to the imperialperiod,boardsof three polernarchoiand a secretaryare regularlyfound in the role of chief local executivemagistratesin Thebesand in other Boiotian cities.TMIn periods when the Boiotian Confederacywas in existence,the polernarchoi and secretaryof the individualcitieswere inferior in statusand powerto the federal boiotarchoi.•

In their accountsof the liberationof Thebesin 379/8, Xenophon(Hell. 5.4.2-8) and Plutarch (Pel. 8-11; cf. also Mor. 597A-598C) agree that Pelopidas and his companions,with the connivanceof the secretary Phillidas, assassinated the polernarchoiand others of their coterie. That done,the next stepwasto rally the Thebanpeoplebehindthe liberatorsand to expel the Spartangarrisonfrom the Kadmeia.Xenophon(Hell. 5.4.9) mentions that the liberators

called the Thebans to arms and that at dawn

hoplitesand horsemencameto their aid. Plutarchtoo records(Pel. 12.1-5; cf. alsoMor. 598C-E)the proclamationof Pelopidasandhisfollowers,and

52

JOHN

BUCKLER

states(12.6-7) that at dawn the liberators convenedan assemblyof the people. Plutarch relates(Pel. 13.l; cf. Ages. 24.6) that at this sessionthe Theban damoselectedPelopidas,Melon, and Charon to the boiotarchia, and at Pelopidas 14.2 he mentionsthat Gorgidas was boiotarchosthat same year. After the elections,the Thebans stormed the Kadmeia, and forced the Spartansto depart under a truce (Plut. Pel. 13.1-3;btor. 598F; cf. Xen. Hell. 5.4.10-12). Even thoughPlutarch,whosenarrativeagreessubstantiallywith that of Xenophon, has been careful of his terminology, and even though he consistently,both in the Pelopidasand in the Agesilaos,speaksofpolemarchoi before the liberation of Thebes and of boiotarchoi afterwards,

K.J. Beloch,•6 H. Swoboda,•7 and G.L. CawkwelP8 have rejected his testimony about the re-establishmentof the boiotarchia in 378.•9 They arguedinsteadthat all four men were local Theban officials,specifically that Pelopidas,Melon, and Charon werepolemarchoi and that Gorgidas was a hipparchos.Hence, they ignoredthe fact that the normal board of local magistrates(as noted above) consistedof threepolemarchoi and a secretary.They basedtheir opinion on two piecesof evidence.The first is Isokrates'statement(14 (Plataikos) 29) that after theliberationof their city the Thebanssentambassadors to Sparta to showtheir willingness$z•8• x•z• '• •p&rcpo••p6½0t6•rob½ •b•oXoy•o•. The secondis Polyainos' account (2.5.2) of a raid which Gorgidas led on Thespiai. Yet one shouldnot rely on Isokrates14.29withouthavingfirst takenthe entire passageinto consideration. Plataikos 27-29 purports to be an accountof Theban relationswith Spartaand Athensfrom the King'sPeace to the liberation

of Thebes. Written after the Theban destruction of Plataia

in 373, Isokrates'pamphletis patentlyand virulentlyanti-Theban in tone, and it contradicts the testimony of better sources. 2øIn brief, Isokrates claims(27) that after Athenshad put an endto the Corinthian War (sic),the Thebans allied themselveswith Sparta against Athens.2• He asserts(28) that the Thebanssworea solemnoathto follow SpartaagainstAthens.The godspunishedthe Thebansfor their perfidy by the lossof the Kadmeia (28). Evenafter the Athenianshad liberatedThebesfrom Spartandomination, the Thebanswerepreparedto becomeslavesof Sparta onceagainand

to remaintrue to the oathsthat theyhad sworn(29). Virtually no aspectof Isokrates'testimonywill bearcarefulscrutiny,and I havearguedelsewhere against it and the authenticity of a Theban-Spartanalliance in 386.22 Isokrates'entire accountis simplyanti-Thebanpropaganda,a rhetorical display-piece,written at a time whenAthenianangeroverThebanactivity in Boiotia

was white.-hot. 23

Cawkwell (loc. cit.), who, like Beloch, accepted the truth and the accuracy of Isokrates' allegations, nonethelessthought that Isokrates' allusion (14.29) to the Theban oaths was really a referenceto the King's

RE-ESTABLISHMENT

OF THE BOIOTARCHIA

53

Peace. This view cannot stand. The allegedoaths of 14.29 refer back to thoseof 14.27-28,which Isokratesclaimswereswornby theThebanswhen they becameallied with Sparta. They havenothingto do with the Theban oaths sworn in ratification of the King's Peace. For the basis of the Common Peacewas the conceptof the autonomy of all Greek states(no matter how muchthe ideal, as articulatedin the stipulationsof the treaty, might be ignoredin practice);and no treatyenjoinedon participantsthe requirements or the obligations of alliance with the prostates of the Peace.24Accordingly,the Thebans would have been no more obligedto follow Sparta than was Athensor any other signatoryof the peace.Thus, thereis no reasonto acceptCawkwell'sviewthat the oathswerein any way connectedwith the King's Peace. Furthermore, sincethere is no reasonto believeIsokrates'testimonyof a Theban-Spartanallianceafter the King's Peace, there is no reason to believe that after the liberation of Thebes the

Thebans wanted to renew it. Isokrates' account of theseyears must be dismissedas worthless,exceptas an exercisein rhetoric. Nor can the testimonyof Polyainosprove that Gorgidaswas a hipparchosrather than a boiotarchosin 378. Polyainos,whosesourcewas probably Ephoros,25statesthat Gorgidas,[rrrr• •Z•,• ©•0t•ov•, raided the territory of Thespiai, where he encountered stiff resistancefrom peltastsunderthe commandof Phoibidas,the Spartanharmost.Gorgidas orderedhiscommandto retreat,beingpursuedall the while by Phoibidas; but whentheThebansreachedan openplain, Gorgidasgavethesignalfor a counter-attack. Phoibidas' peltastsgave way before the Theban cavalry attack; and despite a lacuna in the text, it is clear that they suffered casualties.Nevertheless,Phoibidasand many survivorsreachedthe safety of Thespiai. Polyainoshasin fact garbledan incidentwhichXenophon(Hell. 5.4.4244) describesin greater detail. According to Xenophon, who does not mentionGorgidas, 26the Thebansadvancedon Thespiairrot•8•{, which of courseincludedhoplitesas well ascavalry.In Xenophon'sversionof the episode,Phoibidasand his force of peltasts,supportedby a phalanx of Thespianhoplites,interruptedthe Thebanswhile theywereplunderingthe countryside.The Theban foragerswere beingcoveredby a contingentof cavalry, which was deployedbeyondthe front of its own phalanx. When Phoibidaspressedhome his attack, the entire Theban force--foragers, hoplites,and cavalry--turned to flight. In the ensuingpanicof the Theban army, Phoibidasand hispeltastsout-distanced their heavy-armedsupport. When the Thebancavalry,obviouslyin the rear of the van, reacheda glen?? in desperationit faced about and attacked Phoibidas'men.28Phoibidas and someof his peltastswerekilled, the restfled to the protectionof the Thespianhoplites,and theentireforce,peltastsandThespiansalike,fledto Thespiai.

54

JOHN

BUCKLER

The differences betweentheseaccountsmightsuggest at firstglancethat Xenophonand Polyainosare depictingtwo differentengagements, which is not impossiblein view of Diodoros'testimonythat the Thebanstwice attacked Thespiai. Accordingto Diodoros (15.27.4), the Thebansmade theirfirst attemptafter Kleombrotoshad retiredfrom their territoryin the aftermath of the Theban recoveryof the Kadmeia. In the second,which Diodoros (15.33.6) treats with few details, the Thebans killed Phoibidas and routed his army, which agreessubstantiallywith the testimonyof Xenophon? Yet it is quite unlikelythat Gorgidas'exploit,asdescribedby Polyainos,took placein the earlier encounter.Diodorossaysspecifically that the Thebanswere unsuccessful in their first attack, whichcertainly doesnot correspondwith what one readsin Polyainos.Rather one should look to the secondTheban attack for the solutionof the problem. The great number of similaritiesbetweenthe two accounts(including Xenophon'semphasison Phoibidasand the Thebancavalry)suggests that Polyainoshassingledout part of a largerengagement for inclusionin his Strategemata.Since in the Hellenika the Theban heavy-armedtroops stayedentirelyin the background,there wasno reasonfor Xenophonto pay them muchattention,whichis consistentwith hisversionof the story. Routinely in raids of this sort, the phalanx, which was the leastmobile componentof the entireforce, laggedbehindthe cavalryand provideda line of safetyin the eventof a repulse(preciselythe thingthat theThespian phalanx failed to do).3øThe Theban cavalry,as customary,skirmishedin the openplain betweenits own phalanxand that of the enemy.In thiscase the cavalrydefeatedPhoibidasbeforethe Thebanhoplitescouldenterthe fray. In addition,Xenophondwellson the conductof Phoibidas'peltasts becauseof his own viewson thehandlingof peltasts.As he makesclearin the episodeof Teleutiasbefore Olynthos,Xenophonfeelsthat a peltast commandershouldnot attack rashlyand that peltastsshouldnot be overly bold in their pursuitof cavalry.3•Xenophonusesthe storyof Phoibidasto underlinehis pointsand to illustratethe wisdomof his views.Thus, since evenin Xenophon'sdetailedaccountof this incidentthe Thebanhoplites playedno significantrole in the skirmish,the focusbeinginsteadon the Thebancavalryand Phoibidas'peltasts,it iseasyto seewhytheun-military Ephoroswould have tendedto ignore them or to give them little space. Thesefactorsalsoexplainwhy Polyainos,who wasprimarilyinterestedin cleverruses,concentratedonly on this aspectof the action. Yet it remains to explain the differencesbetweenthe two accounts, especiallysincethereare seriousproblemswith both. Xenophonemphasizesthepanicof theThebanarmy,andclaimsthat thecavalryconfronted Phoibidaswhenterraingavethemno alternative.By that time, however, the restof the Thebanforcehad disappeared. Sinceonly the cavalryhad not yet crossedthe glen, it is obvious(despiteXenophon'ssilenceon the

RE-ESTABLISHMENT

OF THE BOIOTARCHIA

55

matter) that they had successfully coveredthe retreat of the phalanx and muleteers,and that their tactics had permitted the escapeof the slower elementsof the Thebanarmy. This itselfindicatesthat Xenophonhasgiven a distortedaccountof the incident. Far from beinga disorganizedrabble, the cavalryapparentlyretiredin goodorder, perhapsevenfeigningflight to lure Phoibidason. This further suggests that the"blunder"of the horsemen may well be an inventionof Xenophon, who wasfrequentlyreluctantto acknowledgethe military prowessof the Thebans? Polyainos'testimonyabout Gorgidas'signalis evidencethat the Theban counter-attackwas intentional,33but his accountsuffersfrom topographical problems.In the StrategemataGorgidasfacesabout only when his horsemenhad arrived•v •:3.ot-c•-c6p9 Xcop•, whichisimpossible, regardless

of whetheronewishe•to takethecomparative in thesenseof "broader"or "flatter". According to Xenophon, the Theban cavalry retired toward Thebes. The valley of the Kanavari from Thespiai to Thebes actually narrows instead of opening, and the surfaceremains slightly rolling the entire distance.This is a mistakefar more likely to have been made by Polyainosthan by his source.Beginningwith the innovationsof Philip II and Alexander the Great, cavalry played a much more aggressiveand prominentrole in warfare than it had in the classicalperiod.34Perhapsin view of Hellenisticand Roman use of cavalry, Polyainosdid not understandwhy Gorgidaswaiteduntil he reachedthe ravineto opposePhoibidas and therefore substitutedfor the glen a level area where Gorgidas, in accordancewith later military practice,could havedeployedhis horsemen for a concerted attack. 3•

Another difference between the two accounts is that in Polyainos Phoibidasescapedwith hislife, whileXenophonstatesthat theharmostfell with someof the foremostpeltasts.This, like the topographicalconfusion, is a mistake attributable to Polyainos. Diodoros (15.33.6), who also epitomized Ephoros' work, agreeswith Xenophon that Phoibidaswas killed in his rash attack on the Thebans.Thus, Ephoroscould not have made the mistake.Plutarch too (Pel. 15.6)alludesto the death of Phoibidas at Thespiai; and although he gives no date for the event, he placesit betweenSphodrias'raid on the Peiraieusof 378 and the battleof Tegyrain 375.

Earlier in 378 Gorgidashad commandedhoplites.Polyainostellshow, during Agesilaos'first invasionof Boiotia, Gorgidasand Chabriasshowed their scorn for Agesilaosby orderingtheir troops to meet the Spartan king'sshowof bravadowith a showof indifference. 36While a hipparchos could not command infantry, a boiotarchoscould and regularly did.37 Accordingly,at Leuktra the sevenboiotarchoicommandedtheentirelevy, mountedand foot, of the BoiotianConfederacy; 38as did the boiotarchoi during the first invasion of the Peloponnesos, 39and during the second

56

JOHN

BUCKLER

invasion?ø as did the boiotarchoi Hypatos and Kleomenesin Thessalyin 367? and as Epameinondasdid in the Mantineian campaign of 362? Therefore,it is only reasonableto concludethat Gorgidaswasnot simplya hipparchosin 378,but rathera boiotarchos,the only officerwho habitually commandedboth infantry and cavalry. Hence,not only doesPlutarchcarefullydistinguishbetweenthe offices of polemarchoi and boiotarchoi in his account of these years, but the testimonyof Isokratesand Polyainosfails to prove him wrong. After the re-establishmentof the boiotarchia, thepolemarchoi at Thebesrevertedto their traditional role: that of exercisingthe executivefunctions of local government,including the duty to maintain public securitywith its concomitant police powers.•3 They next reappear in connection with the assassination of Euphron of Sikyon at Thebesin 366 (Xen. Hell. 7.3.4-6). Xenophon mentionsthat the assassinsof Euphron were arraigned before

the•o•X'• by certain•p7?'r½c;. Swobodapointedoutthatthepolicepowers of thesearchontes,notablythe right to arrestan individualfor a crimeand to bind the felon over for trial, are normal prerogativesof the polemarchoi.• Added weightis givento Swoboda'sargumentsby the fact that there was no federal boule in the Boiotian Confederacy of the Theban hegemony,its place havingbeen taken by a primary federal assembly? Thus, the boule in question,as Swoboda also suggested,could only have beenthat of the city of Thebes.After the liberationof Thebes,neveragain in the fourth centuryare thepolemarchoiand the secretaryfound activein the broader

areas of inter-state

affairs.

The electionof four boiotarchoiin 378 admitsof an easyand simple explanation: it was the declaration of Theban intentions to restore the

BoiotianConfederacy?The obviousofficialsto bringThebandesigns to fruition were boiotarchoi, not polemarchoi. At first, the duties of the boiotarchoiwere principallymilitary, as they stroveto make the Theban declarationa politicalrealityby drivingSpartangarrisonsout of Boiotian citiesandby defendingBoiotiafrom Spartaninvasions. Asearlyas378the Thebansset out to implementtheir designs.In early 378, after the with-

drawal of Kleombrotosfrom Boiotia,the ThebansassaultedThespiai? Againin 378,after Agesilaos' firstinvasion,the ThebansraidedThespiai and managedto kill Phoibidas? Xenophon(Hell. 5.4.16) statesthat this

victorysoheartenedthe Thebansthat theyundertookoperations against other neighboring cities as well. Sometime between 377 and 375 the Thebansexpelledthe Spartangarrisonfrom Tanagra? In 375 Pelopidas raided Orchomenos?and Charonwon a cavalrybattle beforePlataia.• Xenophon assertsthat by 375 Thebeshad recoveredthe citiesaround it. •2 Indeed, by that time only three of Thebes'rivals still held out; and even

thoughOrchomenos did notcapitulateuntil371,53PlataiaandThespiaifell to Thebes before Leuktra. •

RE-ESTABLISHMENT

OF THE

BOIOTARCHIA

57

The restoredConfederacywhich the boiotarchoi led has recentlybeen the subjectof somedebate. Although Cawkwell (loc cit.) has claimedthat "once there were Bo[to'rol,there was, technically,no longera statecalled Thebes", he has misconstruedthe nature of the new Confederacy. 55 Diodoros calls the new governmentboth synteleia•6 and koinon?7 and Hypereideslikewise speaks of the Boiotian koinon?8 as do Boiotian

inscriptions? Perhapsthe most conclusiveevidenceagainstCawkwell's view is SEG XXV 553, an inscriptionof the Boiotian koinon (lines2-3), which dates to the mid-fourth century. In its final, damagedlines(13-15) the inscription lists the namesof boiotarchoi from Thespiai and Tanagra. Hence, the evidencefrom the fourth century,both literary and epigraphical, indicatesthat the Boiotian Confederacywas truly a federal government, i.e. one in which individual cities, even while their populations

enjoyedfederalcitizenship, 6øretainedtheir politicalidentityand the right of local jurisdiction. Thus, from 378 until Leuktra removedthe specterof Spartanintervention, Thebesworkedrelentlessly to rebuildthe BoiotianConfederacy.The tyrannyof Leontiadesand hisfollowers,supportedby a Spartangarrison, demonstratedto the Thebans what they could expect at the hands of Sparta, and it made them realize that only a Boiotia united under their leadershipoffered them any hope of security.The election of four boiotarchoi in 378, in the first exuberanceof victory and freedom, wasalso the first declaration

of Theban

ambitions

in Boiotia.

Universityof Illinois

John

Buckler

Appendix:Polemarchoiat Thermopylai?

R.J. Buck, CP 69 (1974) 47-48, usedthe ©•[•dto¾'f2po[(or 'r& •pXo,rr•tq6rro•,•t-r•t, in the words of Plut. Mor. 867A) of Aristophanes of Boiotia(FGrH 379)asevidencefor theexistenceof Thebanpolernarchoi beforethe fourth centuryin his discussionof the Boiotian commandersat Thermopylaiin 480.62Plutarch, who citesAristophanesas his authority, claims that Anaxandros--not Leontiades,as Herodotos (7.205.2, 233) reports--was strategosat Thermopylai. Buck interpretedAristophanesPlutarch to mean that Anaxandrosand an unnamedThespianservedas polemarchoi under the overall command of the boiotarchoi Leontiades and Demophilos. Yet Aristophanes'chronicleis itself fraught with difficulties,especiallywith regard to the identity of thesearchontesand their relationshipto the strategoi. Jacoby (FGrH IIIb Komm. Text, p. 160) argued that Aristophanes' archonteswere eponymouslocal Theban officials, but admitted (FGrH IIIb Noten, p. 109,n. 5) that he did not know exactlywhichofficialsthey

58

JOHN

BUCKLER

o o

o

.oO*

o

ß

RE-ESTABLISHMENT

OF THE

BOIOTARCHIA

59

were.Although Schaefer,RE Suppl. 8 (1956) 1109,claimedthat oneof the three Theban polemarchoi was eponymous,his evidence(an inference

from Plut. Pel.7.4:•:•¾•t•p• 'Apx•0t¾ x0t••3, t:r:ro¾...•ro3,•t0t•Xo6,rrco¾; cf. alsoBusolt-Swoboda,1417n. 3) ishardlycompelling.In Boiotiancitiesthe local eponymousmagistratewas the archon, as seenmost clearly in the inscriptionspublishedby P. Guillon (Les trdpiedsdu Ptoion I (Paris 1943) 54-57;cf. also Roesch,157-162).The functionsof the eponymousarchon were sacral(Plut. Mor. 597A-D), and Bussmann(13) likened the Theban archon to the Athenian

archon basileus.

Yet thereis no needto equatethe eponymousofficialsof Aristophanes' list with strategoi,as Buck recognized,no more than one needconclude, whenan Atthidographerrecordeda militaryvictoryin a particulararchonship, that the officesof archon and strategoswereone and the same.Like the Atthidographers,Aristophanesno doubt recordedthe namesof the commandersat Thermopylaiundertheeponymousmagistrateof that year, as Jacobypointed out (Arthis (Oxford 1949) 180, 297, n. 5). Buck'ssuggestionthat Anaxandrosand the Thespianat Thermopylai were polemarchoi contradicts what is known of both local and federal magistracies.Bussmann(15) had earlier argued that, their title notwithstanding,the polemarchoidid not commandthe military contingentof their city in the classicalperiod.Salmon(AC 22 (1953)348-349)similarly demonstrated that in the fifth century the boiotarchoi commanded the leviesfrom the individualcitiesor units,and that lochagoiservedastheir subordinateofficers.In view of theseconclusions,it seemsvery unlikely that the strategoiat Thermopylaiwerepolemarchoi,and one can reasonably ask whetherit is usefulto conflatethe testimonyof AristophanesPlutarch and Herodotos, as Buck has. Consequently,one must reject Buck's argumentsabout Theban polemarchoi in the Persian War. 62 NOTES

I. For the terms of the King's Peace, see Xen. Hell. 5.1.31, 35-36; Diod. 14.110.3;Plut. Artox. 21.5 (ed. Ziegler); Justin 6.6.1; H. Bengtson,Die $taatsvertr•ge desAltertums II 2 (Munich 1975),no. 242. The mostpertinentmodern

discussions areK.J. Beloch,Griechische Geschichte III 2 1(BerlinandLeipzig1922) 94-95; P. Cloch•, Thbbesde Bdotie(Namur, n.d. (1952)) 112 (hereafterCloch•); T.T.B. Ryder, Koine Eirene(Oxford 1965)41, cf. 48; P. Roesch,Thespies et la confdddration bdotienne(Paris 1965)36, 43-44(hereafterRoesch);J.A.O. Larsen, Greekfederal states(Oxford 1968) 175 (hereafterLarsen);H.M. Hack, AJP 99

(1978)210-216;R.K. S!nclair,Chiron8 (1978)37-43. 2. Cf. Hell. 4.8.15; 5.1.33; 5.2.34. Xenophon'suseof the word boiotarchoiat

3.4.4 (cf. 3.5.5)is uniquein his works:seeF.G. Sturtz,LexiconXenophonteum (Leipzig1801-1804),s.v.•o•&-r0•pXo•. Xenophonis equallyreluctantto acknowl-

60

JOHN BUCKLER

edgethe federalprincipleof the Boiotian Confederacyestablishedin the 370s(cf. Hell. 5.4.63; 6.1.1; 6.4.3), although other sources--numismatic(cf. n. 3 below), epigraphic,and literary (cf. pp. 00-00 above)--clearly attest it. 3. 16.2-17.5(ed. Bartoletti); 11.2-12.5in the edition of B.P. Grenfell and A.S. Hunt, The OxyrhynchusPapryi V (London 1908).Accordingto B.V. Head, On the chronologicalsequenceof thecoinsof Boeotia(London 1881)30-60, Boiotiancoins similarlypoint to the disruptionof the Confederacyat thistime. Nonetheless, one hesitatesto rely too heavilyon evidencedrawn from Boiotiannumismatics.Head's studysuffersfrom two defects:(1) Head madeno die-studyof the coins,and (2) he often indulged in circular argument: numismatic material to support literary sources and vice versa.

4. Note especiallythe introductorywordsof 16.2(Bartoletti; 11.2 Grenfell and

Hunt):•X•, 8• •:&•p•yy.o•:o• •:6•:•xo•[•:& •:•]•,Bo•:i•,

o6•:•, andthediscussion of

Grenfell and Hunt, 225 and I.A.F. Bruce, An historical commentary on the Hellenica Oxyrhynchia(Cambridge1967)103.SeealsoM.N. Tod, Greekhistorical inscriptionsII (Oxford 1949),no. 101;for other evidencefor the existenceof the Boiotian Confederacybefore 386 seeAndok. 3.24; Diod. 14.81.2. 5. Hell. 5.2.25, 32; 5.4.2, 7-8; cf. also Nep. Pel. 3. 6. Police powers:Hell. 5.2.30; Plut. Pel. 9.8; 10.1-4(ed. Ziegler); Mot. 598A; polemarcheion:Hell. 5.4.6; cf. also H. Schaefer, "Polemarchos",RE Suppl. 8 (1956) 1109-1110(hereafter Schaefer). 7. Hell.

5.2.25-32.

8. 7.4; 9.4, 8; 11.4; cf. Mot. 577B. 9. On this speech,seealso Arist. Rhet. 2.23.3. Moreover, Xenophon (Hell. 5.4.1) attributesto Leontiadesand his followersnothing more ideologicalthan a desire to rule as tyrants. 10. Schaefer (1108) argued that the only change in the polemarchia under Leontiadesand his coterie was an unprecedentedextensionof the powersof the office.

11. Originally in Festgabenzu Ehren Max Biidingers(Innsbruck 1898) 59, repeatedin K.F. Hermann and H. Swoboda,Lehrbuchder griechischen$taatsaltertiimer III (Tfibingen 1913) 253, and G. Busolt and H. Swoboda, Griechische $taatskundeII (Munich 1926) 1440, 1446 (hereafter Busolt-Swoboda). 12. For argumentsagainstthe presenceofpolemarchoiat Thermopylai,seethe Appendix.

13. On the situationat Athens:Arist..Ath.Pol. 3.2;V. Ehrenberg, TheGreek state (New York 1964) 66. 14. Polemarchoi in Hyettos: IG VII 2809-2832;Akraiphia: 2178-2180; Orchomenos:3179; Kopai: 2782-2789;Chaironeia: 3292-3293;Lebadeia:3070, 3072; Tanagra:$IG 3 1185;Halai: ,,1J,,119 (1915) 444, no. 3; Thespiai:Plut. Demet. 39.5; cf. also J.B. Bussmann,Die b6otischeVerfassung(Fulda 1912) 14-18 (hereafter Bussmann);Busolt-Swoboda, 1438; Roesch, 162-176. 15. Bussmann, 15, 38; C1och•,74; Roesch, 162; Schaefer, ll10. This is most clearlyseenduring the third centuryin the list compiledby M. Feyel, Polybeet l'histoirede B•otie au II1 • sibcle(Paris 1942) 307-308, which indicatesthe incompatibility of the two offices.During the Theban hegemony,the need for local

magistratesin addition to federalofficialsbecameall the more pressing,as the

RE-ESTABLISHMENT

OF THE BOIOTARCHIA

61

dutiesof the boiotarchoi(e.g.Epameinondas' extensivecampaigns in thePeloponnesosand Pelopidas'missionsin northernGreece)took thembeyondthe bordersof Boiotia.

16. Op. cit. (supra n. 1) 145 n. 2. 17. "Gorgidas", RE 7 (1912) 1620. 18. CQ 66 (1972) 275-276. 19. Clochr, 119n. 2, Schaefer,1110,and Roesch,44, 100,acceptedtheaccuracy of Plutarch'sstatementwithout, however,arguing the point or answeringthe objectionsof thosewho reject it. 20. For the date of the Plataikos,seeA. Momigliano, AthenaeumNS 14(1936) 27-32;and for Isokrates'persistent hatredof Thebes,P. Clochr, RH 193(1942)277296.

21. Cf. Bengtson,Staatsvertr?ige II 2, no. 243. 22. See my "The allegedTheban-Spartanalliance of 386 BE",forthcoming in Eranos, where a detailedcriticismof Isokrates'testimonyand a discussionof his value as an historical source can be found. One must remember that Isokrates was a

rhetorician,not an historian;and "rhetoric",as R. Meiggs, TheAthenianEmpire (Oxford 1972)9, hasremindedus,"wasmoreconcernedwith persuasionthan with truth".

23. Cf. Xen. Hell. 6.3.1, 5, 13. Althoughsometerms(&xo3.ou0e•,•, m•txtx•3(•q, and m•tx&Xot•),whichfigure prominentlyin Isokrates,recur in Xen. 6.3.7-9, all that Autoklessaysis that the Spartanshad usedthe autonomyclauseof the King's Peaceasa pretextfor leadingtheir allies(amongwhomhe doesnot numberThebes) againstindependentstates,one of which was Thebes. 24. For the termsof the King'sPeace,cf. n. I above.Eventhoselater CommonPeacetreatieswhichincludedenablingclausesaimedat discipliningrecalcitrant statesdid not entail alliance:cf. Ryder (supra n. 1) 71-74, 131-133. 25. Cf. J. Melber,Jahrbficherffirclassische PhilologieSuppl.14(1885)550-551. 26. As G.E. Underhill, A commentaryon the Hellenica of Xenophon (Oxford 1900) xxviii, long ago pointed out, Xenophon was not in the habit of naming Theban

commanders.

27. Since Xenophon statesthat the muleteerswere fleeingalong the road to Thebes,the ravinecan only be thebedof the Thespios(themodernKanavari)river, whichis often markedby steepbanks,and whichflowsalongthe road to Thebes. The Kanavarirunseastwardsfrom Thespiaialongthe foot of the southernlineof hills, but at one point towardsThebesit cutsnorthwardsat a point wherethe northernline of hills draws near the southern.Here today only a narrow bridge providesa crossing. 28. Cavalry was often expectedto cover a retreat: cf. Hdt. 9.68; Thuc. 6.70.3; Diod. 15.71.6; Plut. Pel. 17.3.

29. Although Diodorosplacesthesetwo raidsin differentarchon-years,they nonetheless belongin the samecampaigningseason,as is madeclear by Xenophon'saccount. In the Hellenika the sequenceof eventsis the following:the recoveryof the Kadmeia and the expeditionof Kleombrotosduring the winter (5.4.3-18); the raid of Sphodriasbefore the campaigningseason(5.4.19-33); Agesilaos'invasionof 378 (5.4.34-41); and the subsequentTheban raid, which resultedin Phoibidas'death,beforethe nextcampaigning season (5.4.42-47).The

62

JOHN

BUCKLER

archon-year,which Diodoros relied upon in part for his chronology,interrupted the campaigningseasonin mid-summer;cf. A.E. Samuel, Greekand Roman chronology (Munich 1972)64; H.W. Parke, Festivalsof the Athenians(London 1977) 29.

30. This is seen most clearly in Xenophon's narrative (Hell. 6.5.30-31) of Epameinondas'advanceinto the southernsuburbsof Sparta in 369. 31. Cf. Hell. 5.3.4; 5.4.54, and for a generaldiscussionof Xenophon'sviews,cf. J.G.P. Best, Thracianpeltasts and their influenceon Greek warfare (Groningen 1969) 110, 122-126. 32. This point has been most recentlystressedby G.L. Cawkwell in his introduction to Xenophon, A history of my times,transl. R. Warner2 (Penguin Books 1979) 37, 198, 343, 399. 33. Ephoros,who preferredto apply a singleformula to battles(cf. G.L. Barber, The historian Ephorus(Cambridge 1935) 140-144;B. Farrington, Head and hand in ancient Greece(London 1947) 58), is not likely to have invented the signal. 34. Cf. W.W. Tarn, Hellenisticmilitary and naval developments(Cambridge 1930)57-92;F.E. Adcock, The Greekand Macedonianart of war (Berkeleyand Los Angeles1962) 48-53. 35. Other examplesof Polyainos'carelessness include 2.3.1, a version of the liberation of the Kadmeia. Here Polyainos portrays Phoibidas, who guards the Kadmeia,beinglured throughlustfor Epameinondas'wife (sic)to attenda dinner at the houseof œpameinondas, and there beingstruckdown by a band of young conspirators.Compare2.4.3,anotherversionin whichPelopidasin thecompanyof some hetairai strikesdown an unnamed garrison commander in the Kadmeia, and the more reliable accounts of the episode by Xenophon, Hell. 5.4.2-10, and Plutarch,Pel. 11.At 2.1.14Polyainoshasconfusedthetopographyof Sparta,when he calls the Issoriona "sacredhill", whereasPlutarch, Ages. 32.6-10, describesthe spotas "well fencedand hard to overrun".A far more seriouserror occursin 2.1.19, in which Polyainosclaims that at the battle of Koroneia in 394, Agesilaosallowed the Thebanhoplitesto penetratehis phalanx in orderto fall upontheir rear.Xenophon, Hell. 4.3.18-20, however,in a vivid accountof the action,describeshow the Spartan king met the Thebans head-on. Still other examples could be easily adduced.

36. 2.1.2; cf. also Diod. 15.32.5;Nep. Chab. 1.2;J. Buckler,Hesperia41 (1972) 466-467.

37. Cf. P. Salmon, AC 22 (1953) 350. Nor is there any evidenceto suggestthat the lochagosof the SacredBandhad the right eitherto commandcavalryor to lead Thebanforcesx0•,83[xe•.. When Pelopidasled the SacredBandand a contingentof cavalryat Tegyra,he wasboth lochagosof the Band(Plut. Pel. 16-17;Diod. 15.37) and a boiotarchos(Diod. 15.81.2;so too at Pel. 16.1, where Plutarch speaksof Pelopidas'victoryat Tegyrasurpassing theexploitsof his•u•'rt:0['r3¾o[, wherebyhe means the other boiotarchoi; cf. Pel. 29.1, where he refers to the Theban commandersof 367, who were boiotarchoi (Paus. 9.15.1-2), as strategoi;cf. also Xen. Hell. 7.4.40; Diod. 15.82.3). 38. Xen. Hell. 6.4.6 ff.; Diod. 15.52.1 if.; Plut. Pel. 20.3 if.; Paus. 9.13.6 ff. 39. Diod. 15.62.4-66.5; Plut. Pel. 24-25; Paus. 9.14.5; cf. Xen. Hell. 6.5.22-52. 40. Diod. 15.68.1; Paus. 9.15.4.

RE-ESTABLISHMENT

OF THE

BOIOTARCHIA

63

41. Diod. 15.71.3; Paus. 9.15.1-2.

42. Xen. Hell. 7.5.4 adfin.; cf. also Polyainos 2.3.14, where Epameinondas ordered his hipparchosto launch a diversion. 43. Cf. n. 6 above.

44. Festgaben . . . Bfidingers(n. 11 above) 58-60, repeated in HermannSwoboda, Lehrbuch der griechischen$taatsaltertfimerIII 253,267 n. 4; accepted by R.J. Bonner and G. Smith, CP40 (1945) 20-21, and by Schaefer,1110. 45. IG VII 2407-2408; $EG I 101; Diod. 15.80.1-2; cf. also Larsen, 178. 46. Cf. J.H. Thiel, Mnemosyne54 (1926)21-22. Similarly, Larsen(31) suggested that at the time of the Persian Wars the boiotarchoi may have beennothing more than "Thebanofficialswhosetitle proclaimedthe desireto ruleall Boeotia".On the connectionbetweenthe numberof timesthat Pelopidaswasboiotarchosand the date of the re-establishmentof the boiotarchia, cf. G.M. Bersanetti,Athenaeum NS 27 (1949)89-93and the commentsof Cawkwell.AlthoughDiodoros( 15.81.4)states that Pelopidasheld the boiotarchiaeveryyear from the return of the exilesto his death in 364, he has forgotten that Pelopidas was not a boiotarchosin 371. Plutarch, on the other hand, states(Pel. 34.7) that Pelopidasdied in 364 during his thirteenth boiotarchia, and at 20.3 he mentionsthat Pelopidasdid not hold the boiotarchiain 371. From this evidence,Bersanettirightly concludedthat 378 is the only date possiblefor the revival of the office. 47. Diod. 15.27.4;cf. Xen. Hell. 5.4.14-18. Plutarch(Pel. 15.2)alsoalludesto numerousengagements in theseyears.Diodoros'chronologyof theseyearsisawry. This is obvious from his placing of the formation of the Second Athenian Confederacyunderoneyear(15.28,29.7-30; cf. alsoIG II 2 34/5, 40, 41,42, 43, 44; G.L. Cawkwell, CQ 67 (1973)480);hisplacingof Persianattemptsto recoverEgypt in one archon-year(15.29; cf. Beloch, GG III2 2.229-230);his confusionover the Common-Peacetreaties of 375 and 371 (15.38, 50.4-6; cf. S. Lauffer, Historia 8 (1959) 315-348);and his confusionover the unrestin the Peloponnesos (l 5.40;cf. J. Roy, Klio 55 (1973) 135-139;G.L. Cawkwell,CQ70 (1976) 77 n. 53). It istherefore preferableto follow Xenophon,who at leasttakesnoteof the campaigningseasons of the years378-375(378: Hell. 5.4.13-46;377:5.4.47-57;376:5.4.58-61;375:5.4.62-

66 (noteespecially 63:•r• 8• • r&• •o•

06x•pt{3•{31•qx6z• z&• •ol•&o• o6r' •

48. See above pp. 53-55. 49. Xen. Hell. 5.4.49; Isok. 14.9; Plut. Pel. 15.6; cf. IG VII 1903, 1904. 50. Diod. 15.37.1-2; Plut. Pel. 16-17.

51. Plut. Pel. 25.8; E. Meyer, Geschichte desAltertumsV (Stuttgartand Berlin 1902) 390-391. 52. Hell. 5.4.63; W. Judeich,RhM 76 (1927) 180-181. 53. Diod. 15.57.1; cf. J. Buckler, W$ 90 (1977) 77-78. 54. Xen. Hell. 6.3.1, 5; Paus. 9.2.8.

55. See Larsen, 175-180,who discussedthis questionat length. 56. 15.50.4, 70.2 57. 16.85.3; cf. 15.80.2', 16.25.1.

58. 1 (Against Demosthenes)fr. 4 col. 18. 59. Cf. J.P. Michaud, BCH 98 (1974) 644-645, which datesa little after 338.

AlthoughN.G.L.Hammond, A historyof Greece • (Oxford1967)570,hasclaimed

64

JOHN

BUCKLER

that Philip dissolvedthe Boiotian Confederacyafter Chaironeia, which would requirethe redatingof this inscription,more recentlyJ.R. Ellis, Philip H and Macedonianimperialism(London 1976)201,296 n. 92, and G.L. Cawkwell, Philip of Macedon (London and Boston 1978) 168, 205 n. 7, have demonstratedthat Philip left it standing. Cf. also P. Cloch6, Un fondateur d'empire: Philippe H (Saint-Etienne 1955) 266. 60.

Cf. n. 45 above.

61. Recently repeatedin his A history of Boeotia (Alberta 1979) 129-132.

62. My thanksto ProfessorsE. Badian,Editor of this Journal, and John M. Fossey,and the two anonymousreadersfor their valuablecomments.While I have not alwaysagreedwith their views,I haveinvariablyfound themhelpful.Theyare, of course,not responsiblefor any defectsof this piece.

COUNT GAiNAS AND COUNT SEBASTIAN

A murderous Goth and a Roman knight-errant seem to be unlikely materialfor a studyof parallellives.The sources,however,makeit difficult to examineseparatelythecareersof CountGainasand CountSebastian.In the presentpaperI proposeto investigatemoreintensivelythan beforethe sourcesfor the climax of Gainas'career,and thetestimonyand chronology of the life of Sebastian.I shall argue that the Gothic historian Jordanes mistakenlyassignedsomeof Sebastian'sdeedsto Gainas.Finally, I shall attempt to identify the sourceof Jordanes'confusion.•

In AD 400 the GothicCount Gainaslosthislife after posinga seriousthreat to Constantinopleand the Roman Empire in the East. In the previousyear he had securedappointmentas magisterutriusquemilitiae upon meeting with the Emperor Arcadiusat Chalcedon.He had then crossedthe straits and entered Constantinople.Immediatelyhe tried to occupythe capital with barbarian forces under his command. After the attempt failed, he withdrewfrom the city, intendingto havehistroopsseizeit on a signal.The plan miscarried.The angry inhabitantsmassacredthoseof his men who were in the capital. For Gainas and his survivingfollowers withdrawal becameflight. Finding Thrace well guarded,he tried to crossto Asia at the Hellespont. But another Goth, the magistermilitum Fravitta, blockedhis passage,and readied a fleet of liburnae while Gainas delayed. GaYnas' situation becamedesperate.He was running short of supplies,and thus commandedhis men to build rafts in orderto crossthe Hellespont.In the ensuingconflict Fravitta's liburnae sank most of the rafts. After watching the d6b•cle from the Chersoneseshore, Gainas began a retreat through Thrace. He reached and crossedthe Danube, but the Huns under a certain Uldin attacked and killed him. Uldin then had GaYnas' severed head sent to

Constantinople.2 The defeat and death of Count Gainas made a deep impressionon

65

66

FRANK

M. CLOVER

Eastern Romans. The Emperor Arcadius ordered a column commemoratingthe victoryto be constructedin the XerolophosDistrict of Constantinople's Region Twelve.3 In Region Four stood a marble liburna, a monument to Fravitta's victory on the Hellespont.4 Of the surviving literary accountsthe most vivid and circumstantialis that of Zosimus.5 Behind Zosimusstand Eunapiusof Sardis' Hypomnemata historica, the secondedition of which treated eventsfrom AD 270 to 404.6 Although Zosimus does not cite Eunapiusdirectly, he providesa glimpseof his distinguishedpredecessor's narrative of eventsat Historia nova 5.20.3-4:

[3] Now Fravitta wasoccupiedwith thesemattersin Asia, watching over his own campand observingthe enemy'smaneuversnight and day. And he took care also of his fleet. For he had ships sufficientfor a naval battle, thosewhich are calledliberna(pl.), named from some city situated in Italy, at which from the beginningthe model of theseboats was constructed. [4] Theseboats seemto sail fast, not slowerthan fifty-oared ships, but they are considerablyinferior to triremes.The making of these had lapsed earlier by many years, even if the historian Polybiusseemedto havesetforth the measurements of the ships with six banksof oars,with which the Romansand Carthaginians seemto havefought againstone anothermany times.7 A comparisonof this passagewith the pertinentfragmentsof Eunapius suggests that Zosimusherelet slip someof the extravagantoriginalwhich, according to Photius, he transcribed but rendered more concise.8 The fragmentsof Eunapiuspertainingto Gainassurvivein ConstantinePorphyrogenitus'Excerpta de sententiis,which accordingto the excerptors camefrom the secondedition? and in the $uda, which evidentlydrew its material from the Constantinian excerpts.•0 In Fragment 75 Gainas appearswith the Goth Tribigild, plotting a revolt in Asia beforehis meeting with Arcadiusat Chalcedon.• Fragments78 and 79 appearto dealwith the immediateaftermath of the massacreat Constantinople. n Fragment 82 notes that Arcadius rewarded the victorious Fravitta with the consulate,

and concludeswith another referenceto Polybius:•3 Phameas,called Imilco, causingcountlesstroublesfor the Romans, [but] then not beingable[to do so]while Scipiowasgeneral,on being asked the reason, proclaimed:"The flocks are the same, but the shepherdis more robust and more many-eyedthan Argus."

But Fragment81 may afford a more directconnectionbetweenthetextsof

GA•NAS AND SEBASTIAN

67

Zosimusand Eunapius.Here the Suda definesthe term liberna, and then quoteswithout naming the sourcethe following phrase: Having built fast-runningthirty-mangalleysin the styleof libernides. Boissonadeassignedthis phrase to Eunapius on the basis of Zosimus' mention of Fravitta's liburnae.TMThe attribution may be acceptable, although subsequentdiscussionin the presentpaper will show that it should be modified. The general impressioncreatedby the extant fragments is that Eunapius embellishedhis account of Gainas' deedswith pedantic referencesto wars of the past, taken largely from Polybius' history. Eunapiusproducedthe secondedition of his Hypornnernatahistoricain or after AD 414, the year in which PulcheriabecameAugusta;sheis so designatedin Fragment 87, which appearsto be an excursuson events occurringwhile the historianwrote.• It is possiblethat Eunapiushimself observedthe defeat of Gainas, or that he relied on eyewitnessaccounts. One wonders,however, whether he drew in addition on written testimony. The ecclesiastical historian SocratesScholasticusrecordstwo other early sourcesfor the war against GaYnas.After his narrative of the conflict

Socratesnotesthat a full accountmay be found in the Ga•'nia,a popular epicpoemin four bookswrittenby EusebiusScholasticus, a studentof the sophistTroilus and a witnessof the war. Furthermore,statesSocrates,a poet named Ammonius presentedan epic on the same subjectin the presenceof TheodosiusII, while the Eastern Emperor was sharing the consulatewith the WesternsenatorFaustus(in AD 438). For his repetition of the GaYnasepisodeAmmonius earned considerablefame.•6 Eunapius was born in AD 345/6. •? It is unlikely that he lived long enoughto read Ammonius'epic. On the otherhand it is possiblethat Eunapiusnumbered himselfamong the avid readersof EusebiusScholasticus'Ga•'nia.Indeed this may be suggested (if not fully proven)by a closerlook at the Suda, s.v. "Liberna",whichonceagainiscommonlyregardedasEunapius'Fragment 81. For the moment it will be usefulto treat the final glossas a line of poetry:•s

If oneacceptssyneresis of the firsttwo syllablesof thethird word,four and one-half dactylic hexameterfeet emergefrom the line: _ y •,1_

,. yl_

•1

, L-

•:•,•xFevo½(•poF•x(•x½ "rpt•xxov"r•pet½.

68

FRANK

M. CLOVER

Three iambic feet follow. It is possiblethat the entire glossis merelya prose sentence with

metrical

elements.

Yet Zosimus'

excursus

on Fravitta's

liburnaeand Eunapius'predilectionfor exoticwordsand poetrysuggesta simple modification of Boissonade'sattribution: Eunapius began citing Eusebius'circumlocutorydescriptionof Fravitta's fleet, and then abandonedan extensivequotation in favor of a short explanatoryphrase.•9 Whether or not one acceptsthis interpretationof Eunapius'Fragment 81, the possibilityremainsstrongthat Eunapiusrefreshedhis memoryof the eventsof AD 400 by readingEusebius'Ga•'nia.The historian'suseof a popular eyewitnessaccountis in no way surprising.What is surprising, however, is that a generationafter the defeat and death of Gainas the poet Ammonius shouldearn fame by composinga secondGa•'nia.A possible explanationcomesto mind. To judgefrom Socrates'remarks,Ammonius concludedhis poemwith an accountof the birth of Theodosiusin the very year that the victorious Fravitta assumedthe consulate(AD 401). Surely sucha coincidencewas an omen for a brilliant reign. Thirty-sevenyears later the youngEmperorinauguratedthe codewhichbearshis name.20He was in a festivemood, and might have beenpleasedto hear a poet remind him that he wasbornjust after the Empirehad crusheda dangerousenemy. Let thisexplanationof the timing of Ammonius'Ga•'niasufficefor now. A review of the career of Count Sebastianmay yield a further reason. II

Three chroniclers--ProsperTiro, Hydatius and Count Marcellinus--and the ecclesiastical polemicistVictor of Vita are the chief sourcesfor the life of Count Sebastian.Sebastianenteredimperial politicsin ^D 432, whenhis father-in-law Count Boniface replaced Flavius A•tius as first magister utriusque militiae, and then died fighting A•tius. Sebastian succeeded Bonifaceassupremecomeset magisterutriusquemilitiae,but in 433 Ai•tius returnedto powerand expelledhim from thepalace.In 434 Sebastian,now an exile and a fugitive,left Italy and took refugein the imperial palaceat Constantinople.Some time after his arrival there he got wind of a plot againsthim and fled westwardto Toulouse,wherethe Visigothicchieftain Theoderic

I maintained

his court. But Sebastian soon wore out Theoderic's

welcome.He left Toulouse,briefly seizedBarcelona,and on beingexpelled from there set sail for the Vandal leader Geiseric'scourt at Carthage. Geiseric made him an adviser, but later had him murdered.TM

The four principal sourcesoffer differing chronologiesfor Sebastian's odysseyafter hisarrival at Constantinople.Count Marcellinus,an eastern chronicler who usually dates eastern events with precision, places Sebastian'sflight from the easterncapital in 435? Hydatiusand Prosper

GAiNAS AND SEBASTIAN

69

Tiro offer some confirmation: Sebastianleft Italy for Constantinoplein 434, and he fled from Spain to Africa in 440.23 By this chronology Sebastianwas presentat Toulousesometimebetween435 and ca. 439. But Hydatius contradicts this dating: Sebastian left Constantinople for Toulouseand captured Barcelonain 444, he setsailfor Vandal Carthagein 445, and in 450 Geiserichad him killed.24Only the lastdate encountersno resistancein the other sources. 25 As a chronicler Hydatius was best informed on eventsin his native Gallaecia;his graspand dating of events occurringelsewherewasuneven.26Hydatiusrecordsthe followingeventsin the twenty-firstyear of the Emperor Valentinian III's reign (AD 445):27 [130] In the Gallaeciancity of Asturica the BishopsHydatius and ThoribiusinvestigatedclandestineManichaeansand reported their findingsto Antoninus, Bishop of Emerita. [131] The Vandals made a seaborne raid on Turonium, near Tude (Tuy) in Gallaecia.28 [132] Sebastian fled from Barcelona to the Vandals. [133] The Bishop of Rome ordered investigationsof Manichaeans throughout all the provinces.

Like most chroniclersHydatius groupedeventswhich he believedto be related? He therefore located in the sameyear the persecutionof Manichaeans in Asturica and Pope Leo's orders regarding these heretics. Possiblyhe sawthe Vandal raid on Turonium and Sebastian'sflight to the Vandals as related, and thus placed Sebastian'sjourneys from Constantinople to Carthageat the time of the raid. Most recentcriticsadopt Hydatius'datesfor Sebastian'sodyssey. 30For the reasonsstated above, I prefer the chronology of Prosper Tiro and Count Marcellinus. Two passagesin the works of Jordanes, hitherto unexaminedin this connection,will support this interpretation. III

In hisDe summatemporumvel origineactibusquegentisRomanorumthe Gothic epitomistJordanesreportsfirst the suppressionof the Moorish rebel Gildo (in AD 398) and then the Gainasepisode: 31 Gaina vero supranominatuscomisConstantinopolimcivilebellum commovenstotam urbem igni ferroque turbavit fugiensqueHellispontopiraticoritu vivebat.contraquemnavaliproeliodato multi Gothorumeiusextincti.ipsequoquebelloevadens moxtamencapite plectitur.

70

FRANK

M. CLOVER

Jordanes'account containsenoughverbal echoesof the chronicleof Count Marcellinus to suggestthat the chronicler was one of his sources.32But Marcellinusdoesnot statethat GainasharrassedConstantinoplewith fire and sword, nor does he indicate that after his flight the Gothic count

"beganliving a pirate's life on the Hellespont".33This information may come from the lost Historia Romana of Quintus Aurelius Memmius Symmachus,the celebratedvictim of the OstrogothTheoderic'swrath, or it may come from a two-part epitome of history which Eustathius of Epiphaneiacomposedaround AD 502.34In either (or any other) casethe information may derive ultimately from sourceswritten earlier in the fifth century.The mostcuriouspart of Jordanes'testimonyishisinsistencethat after he fled from ConstantinopleGaYnasbeganengagingin piracy on the Hellespont. All other testimonyabout Gainas indicatesthat after he left Constantinoplehis seafaringat the Hellespontwas inept and brief. Nor is thisthe only strangepieceof informationfrom Jordanes.The De origine actibusque Getarurn, that tendentiousapology which stemsfrom Cassiodorus' twelve-book Historia Gothorum and other sources,35men-

tions Gainasin the context of a long discussionof the Visigothsafter they separatedfrom the Ostrogothsin the late fourth century.36In succession, statesJordanes,Alaric, Athaulf and Wallia led their peopleon the march from the Danube Basin to southwesternGaul. Upon Wallia's death Theoderictook the leadershipof the Visigoths.37Jordanesthen leapspast the early years of Theoderic'sreign, which beganaround AD 418,38and notes that peace betweenthe Romans and the Visigothsended in the consulatesof Theodosius and Festus (439), when the Roman general A•tius and his lieutenant Litorius took up arms in Gaul.39 Jordanes presentsthe Gothic versionof a war which beganin 436, whenTheoderic attemptedto seizeNarbo, and endedthreeyearslater with A•tius' victory over the Visigothsand the renewal of the treaty betweenRomans and Visigoths? Jordanesgivesa most perplexingreasonfor the outbreak of hostilities

in 439?

Contra quem [Theodoricum]Theodosioet Festo consulibuspace rupta Romani Hunnis auxiliaribussecumiunctisin Galliis arma moverunt.turbaveratnamqueeosGothorumfoederatorummanus, qui cum Gaina comiteConstantinopolim efferasset. Romans--Jordanes

here means Western Romans--and

Hunnish auxili-

ariesunder AEtiusand Litorius took up arms in Gaul in 439. They did so because"a band of Gothicfoederati,whichwith Count Gainashad stirred

up Constantinople, haddisturbedthem".Thepluperfect subjunctive in the subordinateclauseof the secondsentenceindicatesaction prior to the pluperfectindicativeof the main clause,whoseactionprecedes that of the

GA•NAS AND SEBASTIAN

71

foregoing sentence.It is highly unlikely that Gainas' troops remained together for nearly forty years after their commander'sdeath. Jordanes means that the kind

of Gothic

band

with

which

Gainas

had set Con-

stantinopleon edgein 400 moved A•tius to action in 439. But wherewere such troops when they provoked Ai•tius?In Constantinople?In Gaul? What connectiondid they have with Theoderic? Jordanes'

account

of the defeat of Gainas is intertwined

with informa-

tion on piratic activity on the Hellespont,and with a recollectionof a band of Gothicfoederati which moved Ai•tius' army to attack the Visigothsin Gaul in 439. The $uda's glosson the Emperor TheodosiusII--a complicatedentry dividedinto threesections--providesthe missinglink. The first sectiondescribesTheodosiusas a cowardly emperor who cateredto the whim of his eunuchs? The Constantinian Excerpta de virtutibus et vitiis preserveone of the sourcesof thisunfavorabledepiction,a nearlyverbatim fragmentof the seventh-century historianJohn of Antioch? But the $uda contains precious information omitted in the Constantinian excerpt. Under the youngerTheodosiusRoman noblespaid out much gold "because of the greed of the eunuchsand [because]the piratic band of doryphoroi (spear-bearers)of Sebastiandisturbedthe Hellespontand the Propontis"? One of John of Antioch'ssourceswasthe rhetoricianPriscus of Panium,who wrotea Historyof Byzantiumand theaffairsat thetimeof /lttila in the early 470s? The compilersof the $uda may havehad accessto a lost portion of the Excerpta de virtutibus et vitiis which contained excerptsfrom Priscus? In any casethe $uda's glossrendersunderstandable the testimony of Jordaneson Gainas. Count Sebastian,not Count Gainas, "beganliving a pirate'slife on the Hellespont"after he fled from Constantinople.From the Hellespontand PropontisSebastianjourneyed with his doryphoroi, that is his buccellarii?? to the court of TheodericI. Jordanesrefersto Sebastian'stroopsasfoederati;in thefifth centurythere was no cleardistinctionbetweenfoederatiand buccellarii.48The arrival of Sebastianand his men at ToulousecausedAi•tiusto wagewar againstthe Visigoths. IV

The life of Count Sebastianmaynowbe placedon a largercanvas.Between 435 and 440 the armies of Ai•tiusfought in Gaul againstthe Bacaudae, Burgundians,Franks and Visigoths? In 439 the Visigothsdefeatedand captured his lieutenant Litorius.50 Developmentsin other parts of the MediterraneanconditionedAi•tius'next move.In the pasttwo yearsthere had beena surgeof piratic activity. In 437 and 438, statesProsperTiro, "barbarians,runawaysof thefoederati", had raidedand devastatedmany islands,especiallySicily.• ThesepirateswereprobablyVandals,whomthe

72

FRANK

M. CLOVER

WesternEmperorValentinianIII hadrecognizedasfederatedoccupantsof Hippo Regiusand vicinityin 435.52Count Marcellinusreportsthe capture and executionof a pirate named Contradisin 438.53Sebastian,an exile from the Easterncourt, displayedhis piratic talent in the Hellespontand the Propontis.Underthedirectionof the PrefectCyrustheworriedEastern Romanssoughtto protectConstantinoplefrom attack by extendingthe seafortificationsalongthe Propontisfrom theend of theterrestrialwall of Constantine to the end of the Theodosian Wall around the Golden Gate. 54 In 438 or 439 Sebastian and his band of Gothic bucellarii fled westward to

the court of Theoderic I. On 19 October 439 the Vandals captured Carthage.55Theystrucka brief alliancewith the Visigoths, 56and soughtto protecttheir prizecatchfurtherby raidingSicilythefollowingyear? Of all these developmentsthe most immediatelythreateningto the powerful AiStiuswere the defeat of Litorius and the arrival of the arch-enemy SebastianatToulouse. In 439 AiStiussworea public oath: just as he had subduedthe Visigothsin peaceasa boy hostagein Alaric'scamp,sowould he now defeatthemin war.58Somewherein Gaul hefoughta terrificbattle. After the fighting stopped,he and Theodericnegotiateda renewalof the treaty which since 418 had establishedthe Visigoths as foederati in

AquitanicaSecunda. 59Oneof the resultsof thetreatywastheexpulsionof Sebastian, who left Toulouse for Barcelona, and then set sail for Vandal

Carthagejust as Geisericwasconductingraidson Sicily.Geisericreturned to Carthage to greet this threatening newcomer. 60 Soon the Emperor Valentinianseparated the Vandalsfrom the Visigothsby givingtheformer a more attractive

alliance?

The presentreconstructionof Sebastian's careeris tentative.A problem arises:Why would Jordanesmistakenlyassignsomeof Count Sebastian's deeds to Count GaYnas?Further speculation about the GaYnia of Ammoniusis in order. The poet producedthis echo of Eusebius'epic in 438. Around the sametime (accordingto the chronologyof ProsperTiro and Count Marcellinus) a new Gainas, Count Sebastian,threatened the Hellespont.Ammoniusmay have referredto recentand paralleleventsin his poem. His allusionsmay standbehindthe confusionof Jordanes. Frank

University of Wisconsin

M. Clover

Madison NOTES

I. In 1977 I presentedan early version of this paper at the Chicago Late Classicaland ByzantineColloquium. At variousstagesof preparationI received helpful criticism from ProfessorsA. Momigliano, W.E. Kaegi, T.D. Barnes, F. Paschoud,J.P. Heironimus, E.L. Bennett,and two anonymousreadersfor the

GA•NAS AND SEBASTIAN

73

AmericanJournal of Ancient History. Prof. E.A. Thompsonsentme a photographiccopyof the life of Count Sebastian, asit will appearin Theprosopography of the Later Roman Empire, Volume II. In additionto abbreviations usedby L'Anndephilologique I haveemployedthe following: AE BHAC

CSHB EL

ES EV

FHG

L'Annde dpigraphique(Paris 1888- ) Bonner Historia-Augusta-Colloquium. Antiquitas, Reihe 4: Bei-

tr//ge zur Historia-Augusta-Forschung (Bonn 1963- ) Corpusscriptorumhistoriaebyzantinae.50 vols.(Bonn 1828-1897) ExcerptahistoricaiussuImp. ConstantiniPorphyrogeniticonfecta I: Excerptade legationibus,ed. C. de Boor (Berlin 1903) ExcerptahistoricaiussuImp. Constantini Porphyrogeniti confecta IV: Excerpta de sententiis,ed. U.P. Boissevain(Berlin 1906) ExcerptahistoricaiussuImp. ConstantiniPorphyrogeniticonfecta II: Excerpta de virtutibuset vitiis Part 1, ed. T. B[ittner-Wobst (Berlin 1906) Part 2, ed. A.G. Roos (Berlin 1910) C. M[iller (ed.), Fragmentahistoricorumgraecorum.5 vols.(Paris 1841-84)

HGM L. Dindoff (ed.),Historicigraeciminores. 2 vols.(Leipzig1870-71) MGH.'AA Monumenta Germaniaehistorica:Auctoresantiquissimi.15vols. (Berlin vat. dates, repr. 1961)

PLRE

A.H.M. Jones,E.A. Thompson,et al., Theprosopography of the

Later Roman Empire (Cambridge 1971- ) 2. Seethe summariesof GaYnas'careerin PLRE I 379-80;and H. Wolfram, Geschichte der Gotenvon der Anf•ngen bis zur Mitte dessechsten Jahrhunderts: Entwurf einer historischenEthnographie(Munich 1979) 175-178.

3. Cf. J. Kollwitz,Ostrdmische Plastikdertheodosianischen Zeit(Berlin1941) 17-68;andK. Weitzmann(ed.),Ageofspirituality:LateantiqueandearlyChristian art, third to seventhcentury (New York 1979) 79-81, no. 68.

4. Notitia urbisConstantinopolitanae 5.11(p. 232 Seeck).Despitethe hesitation of D. Kienast,Untersuchungen zu denKriegsflotten derr6mischenKaiserzeit (Bonn 1966)157,n..94, the literarysources' references to Fravitta'sliburnae(see below)makeclearthat this monumentcommemorated Fravitta'svictory. 5. Zosimus,Historianova, Book 5, passim,esp. 18-22.

6. Cf. Photius,Bibliotheca, Codd.77 and98 (I 158-160,II 65-66Henry);F. Paschoud,Cinqdtudessur Zosime(Paris 1975)100-183,207-212(with previous discussions cited);T.D. Barnes,Thesourcesof the HistoriaAugusta(Collection

Latomus155, Brussels 1978)114-23;F. Paschoud,"Quand parutla premiere •ditionde l'Histoired'Eunape?", BHAC 1977-78(forthcoming). 7. Cf. Polybius, Historiae, frg. 39 (IV 519 B-W); and F.W. Walbank, A historicalcommentaryon PolybiusIII (Oxford 1979)746. 8. Seeagain Photius,Bibliotheca,Cod. 98 (II 66 H). 9. ES, Praefatio (p. 71). 10. Cf. C. de Boor,"Suidasunddie Konstantinsche Exzerptsammlung", ByzZ XXIII (1914-19)1-127,esp. 30-31;A.D.E. Cameron,"An allegedfragmentof Eunapius", CQ 2 XIII (1963) 232-36, at 235.

74

FRANK

M. CLOVER

11. Eunapius,Hypomnematahistorica,frg. 75 (FHG IV 47-48)=ES64-71 (pp. 94-96).

12. Eunapius,Hypomnematahistorica,frg. 78 (FHG IV 49)=ES72-73(pp. 9697); and frg. 79 (FHG IV 49)=Suda,s.vv."Hellanodikai" and "Polyandrion". Cf. B. Baldwin,"'Perses':A mysteriousprefectin Eunapius",ByzantionXLVI (1976) 5-8.

13. Eunapius,Hypomnematahistorica,frg. 82 (FHG IV 49-50)=ES74-75(pp. 97-99).Cf. Polybius,Historiae36.8.1-4;Appian,Libyca97,459if., esp.100,471if.; and seeWalbank, Historical commentaryon Polybius III 660-63. 14. Eunapii Sardiani Vitas sophistarumet fragmenta historiarum recensuit notisqueillustravitI. F. BoissonadeII (Amsterdam1822)525, frg. 44. Cf. Mtiller, FHG IV 49, and Dindoff, HGM I 264--both ad Eunapius,frg. 81. 15. Eunapius,Hypomnematahistorica,frg. 87 (FHG IV 52-53)=ES79-80 (pp. 100-102). Cf. Paschoud, Cinq •tudes 170-75. 16. SocratesScholasticus,Historia ecclesiastica 6.6 (pp. 258-59 Bright). Two versesof Ammonius' epic appear to survivein Codex A of the Etymologicum genuinum. Cf. R. Reitzenstein,Geschichteder griechischenEtymologika:Ein Beitragzur Geschichte der Philologiein Alexandriaund Byzanz(Leipzig1897)28789.

17. Cf. PLREI

296.

18. For the metrical length of the secondsyllableof the secondword and the third syllableof the fourth word of thosemarked with quantities,seeLSJ9, s.vv.

19. For shortquotationsof poetryby Eunapiussee,for example,Hypomnemata historica,frgg. 24, 60, 87 (FHG IV 24, 41, 52)=ES 27, 79 (pp. 82, 101) and EL 7 (p. 598).

20. Cf. A. Lippold, "Theodosius"(11), RE Suppl. XIII (1973) 961-1044,at 962, 988-89.

21. ProsperTiro, Epitoma chronicon1310and 1342(MGH:AA IX 473-4, 478), s.aa. 432, 440; Marcellinus Comes, Chronicon, s.aa. 432, 435 (ibid. XI 78-9); Hydatius, Chronicon 99, 104, 129, 132 and 144, s.aa. 432, 434, 444, 445 and 450 respectively;Victor Vitensis,Historiapersecutionisafricanaeprovinciae 1.19-21. Other informationon Sebastianappearsin SidoniusApollinaris,Carmen9.277-88 and the Suda, s.v. "Theodosios";the latter sourcewill receiveattention in Part III of this paper. All datesin Hydatius'chronicleare here in accordwith the new edition

of A. Tranoy, Hydace: Chronique(Sourceschr•tiennes,nos.218,219; Paris 1974). 22. Marcellinus Comes, Chronicon, s.a. 435 (MGH:AA XI 79). One awaits a publishedversionof the Oxford D. Phil. thesisof Brian F. Croke: "The Chronicle of Count Marcellinusin its contemporaryand historiographicalcontext".

23. ProsperTiro, Epitoma chronicon1342 (MGH:AA IX 478); Hydatius, Chronicon 99, 104. 24. Hydatius, Chronicon 129, 132, 144.

25. CountMarcellinusand ProsperTiro (ascitedabove,nn. 22 and23) record Sebastian'sdeath under the years 435 and 440 respectively.But sincetheir chroniclescontainonly oneentry on Sebastian,they meanby thesedatesthat Sebastiandiedafter hefledfrom Constantinople in 435(soMarcellinus)andafter heleft Spainfor Carthagein 440(soProsper).In hisecclesiastical polemicVictorof

GAiNAS AND SEBASTIAN

75

Vita gives dramatic rather than absolute dates for events. Sebastian'smurder occursafter the Vandal captureof Carthage(^D 439) and beforethe ordinationof Deogratiasas Bishopof Carthage(454). SeeHistoria persecutionis1.12-27;and cf.

C. Courtois, VictordeVitaetsonoeuvre : t•tudecritique (Algiers 1954)58,n.284,et passim.

26. See now E.A. Thompson, "The end of Roman Spain I", Nottingham Medieval Studies XX (1976) 3-28, at 4-18. 27. Cf. Tranoy, Hydace I 140. 28. On the location of Turonium see Parochiale $ueuum 12.9 (Corpus

Christianorum:Serieslatina CLXXV 419); and E.A. Thompson,"The end of Roman Spain II", Nottingham Medieval StudiesXXI (1977) 3-31, at 24, n. 98. 29. The bestobserverof this phenomenonis 0. Holder-Egger,"Untersuchungen fiber einige annalistischenQuellen zur Geschichtedesffinften und sechstenJahrhunderts",NeuesArchiv der Gesellschaft far a'lteredeutscheGeschichtskunde I (1876) 13-120, 213-368; II (1877) 47-109. 30. See,for example, Courtois, Victor de Vita 53, n. 224; Wolfram, Geschichte der Goten 213-14; PLRE II 983-84. Only J.L.M. de Lepper, De rebus gestis Bonifatii comitisAfricae et magistrimilitum (Breda 1941) 107-115,hasarguedin favor of the chronologyof ProsperTiro and Count Marcellinus. 31. Jordanes, Romana 320.

32. Compare Marcellinus Comes, Chronicon,s.aa. 398-400(MGH:AA XI 6566) and Jordanes,Romana 320. Cf. L. Vfirady, "Jordanes-Studien:Jordanesund das 'Chronicon' des Marcellinus Comes--Die Selbst/indigkeitdes Jordanes", Chiron VI (1976) 441-87, esp. 452-53 and 483-84. 33. Cf. W. Ensslin,Des SymmachusHistoria Romanaals Quelleft2r Jordanes (SBAW 1948, 3, Munich 1949)66-67. 34. Cf. Ensslin, Symmachus,passim; M.A. Wes, Das Ende des Kaisertumsim Westendes R6mischenReichs(The Hague 1967),passim;and Vfirady, Chiron VI (1976) 441-87.

35. Cf. Jordanes, Getica1-3;andJ.J.O'DonneH, Cassiodorus (Berkeley1979) 43-54.

36. Jordanes, Getica 131-245. 37. Jordanes, Getica 131-176.

38. Cf. E. Stein, Histoiredu Bas-Empire,ed. J.-R. Palanque,12([Bruges]1959) 267-68.

39. Jordanes, Getica 176-177.

40. Cf. Stein, Bas-Empire 12322-24. 41. Jordanes, Getica 176.

42. $uda, s.v. "Theodosios"(II 694-95 Adler). 43. JoannesAntiochenus,Historia chronica,frg. 194(FHG IV 612)=EV72 (pp. 204-205). 44. $uda, s.v. "Theodosios"(II 694 Adler, lines 34-36). 45. Cf. G. Moravcsik, ByzantinoturcicaI :, BerlinerByzantinistischeArbeiten X

(Berlin1958),313-15,479-'88. 46. Cf. C. de Boor, "Zu JohannesAntiochenus",HermesXX (1885) 321-30,at 328-29;and "Suidasunddie Konstantinsche Exzerptsammlung", ByzZ XXI (1912) 381-424,at 400, and ibid. XXIII (1914-19) 19-20.

76

FRANK

M.

CLOVER

47. Cf. H.-J. Diesner, "Das Buccellariertum von Stilcho und Sarus bis auf Aetius (454/455)", Klio LIV (1972) 321-50, at 322-23. 48. Jordanes,Getica 176: "Gothorum foederatorum manus." Onfoederati and buccellarii see Olympiodorus Thebaeus, Historiae, sections7 and 11 (FHG IV 59)--Photius,Bibliotheca, Cod. 80 (I 168-69 H.). 49. Cf. Stein, Bas-Empire 12322-24. 50. See,for example,ProsperTiro, Epitomachronicon1335(MGH:AA IX 476) and Hydatius, Chronicon 116. 51. Prosper Tiro, Epitoma chronicon 1330, 1332 (MGH:AA IX 476). 52. See especiallyProsperTiro, Epitoma chronicon 132! (MGH:AA IX 474). 53. Marcellinus Comes, Chronicon, s.a. 438 (MGH:AA XI 79). 54. Chroniconpaschale,s.a.439 (CSHB XV1583). G. Dagron, Naissanced'une capitale: Constantinopleet sesinstitutionsde330 d 451 (Paris 1974)111-12and 26872, correctlydeterminestheextentof constructionsupervised by Cyrus,but follows the PaschalChroniclein associatingthis extensionof the wallswith Vandal piracy. But the Vandals raided only as far east as Sicily. Sebastian, not the Vandals, threatenedConstantinople. 55. See especiallyProsperTiro, Epitoma chronicon 1339(MGH:AA IX 477). 56. This followsfrom Jordanes,Getica184. Cf. A. Gitti, Ricerchesui rapporti tra i Vandali e l'Impero romano (Bari 1953) 49-65. 57. See especiallyProsperTiro, Epitoma chronicon 1342(MGH:AA IX 478); Hydatius, Chronicon 120. 58. AE 1950,no. 30;and seeMerobaudes,Panegyric2.127-42(esp.130-32),with Carmen4.42-46. Cf. T. Olajos,"L' inscriptiondela statued'A•tius et Merobaudes", Acta of theFifth InternationalCongress of Greekand Latin Epigraphy,Cambridge 1967 (Oxford 1971) 469-72; F.M. Clover, Flavius Merobaudes:A translationand historical commentary (Philadelphia 1971) 39. 59. Cf. Clover, Merobaudes 58-59; Wolfram, Geschichteder Goren 212-13. 60. Cf. ProsperTiro, Epitoma chronicon 1342(MGH.'AA IX 478). 61. Cf. Gitti, Rapporti 13-65.

QUALIFICATION

OF NUMERALS

IN THUCYDIDES

•

Numbers constitutea large and important part of the detailedinformation supplied by ancient historians. For both ancient and modern scholars, moreover,they havebeen,for variousreasons,a focusof frequentcontroversy; and in modern times scholarshave tended, perhaps,to regard numericaldetails as more objectivelypreciseand lesslikely to be coloured by emotion than other aspects of historical narrative. It is therefore important to pay particularattentionto a factor whichaffectsthe precision and emotional content of numerical information. Many numbers are statednot absolutelybut with somequalification attached,which is clearly intended to influence their interpretation. In a contemporary context everyoneis familiar with this mode of expression,and interpretsit usually without difficulty. Thus, if a reportersendsa preliminaryreport of a plane crashstatingthat "about 50 peoplehavebeeninjured" or "at least24 have died", we easily infer in the former casea generaldegreeof uncertainty about the number of those injured, and in the latter the fear, or the likelihood, of more bodies being discoveredas the wreckage is more thoroughlysearched.Similarly, no one is surprisedto hearspokesmen for two opposingpolitical parties using qualifters emotionally to favour a partisaninterpretationof somefact. Thus if the cost-of-livingindex has gone up by .9% over a certain period, Governmentministersare likely to refer to this as "a rise of lessthan 1%", while membersof the Opposition will speakrather of"this largejump of closeto 1%".The nuanceconveyed by a qualifiednumeralin the text of an ancienthistorianis often harder to interpret,becausewe lack both the quickinstinctof a nativein interpreting AncientGreekor Latin and mostof the backgroundinformationwhichwe useto aid us in our own time. But it shouldbe clearthat unlesstheattempt

is madeto interpretthe implicationsof qualifyingexpressions, an important aspectof any historicalnarrativeis beingneglected. I presentheresomeof the resultsof a studyof howThucydidesqualified numerals.Many scholarshavecommented,of course,oncertainaspectsof this question?and discussed the possibleimplicationsof the qualification attached to certain numerals. 3 But no one seems to have collected and

77

78

CATHERINE

REID

RUBINCAM

studied all the qualified numerals in Thucydides; and all too often the presenceof a qualifying word with a number has been ignored unlessit occasionedsomespecialdifficulty in senseor syntax. It is the intention of this paperto demonstratethat qualificationis not a negligiblefactor, but one that deservesand repaysseriousand systematicstudy,sinceit can shed new light both on many specificproblemsand on somegeneralquestions regarding how Thucydides investigated,thought about, and wrote his history.

Qualifying expressionsseemto fall into three main types:(i) "approximating" expressions--thosethat indicate some unspecifieddegree of uncertaintyand/or imprecisionfelt by the writer or speakerconcerningthe number mentioned;4 (ii) "comparative" expressions--thosethat imply a comparisonbetweena realfigure,whichisnot specified,and anotherfigure which does not claim to be exact but is said to lie at some distance above or

below the real figure; and (iii) "emphatic"expressions--thosethat simply emphasizehow large or smallthe statedfigureis. Thucydides'vocabulary of qualifierscan be classifiedon this schemeas follows:approximating expressions--tz•X•'r•,d)•,•zpl, 'r•, . .

o•ov, otov; comparativeexpressions--o6x 6l•y• •lz[ou, •lz[ou, oG•lz•ou, •l•ou, G•p, oG•oll• z•v•El•aov, OGS•,•p•Z6 R=•l•=ov;6 emphaticexpressions--•6vo•.7 In addition, a few expressions(&g •=• z6 =o•6, and =lz•oog, •az• ff• •lRaaoo) are used specificallyto indicatean average.8Thesereallyconstitutea sub-typeunder

"approximating expressions" (seebelowpp. 79f.). The word•g, usedby

itselfor in various combinations (&... ?•l•az•, &... z•,&g,61•yoo •g), seemsto fall somewhere betweenthe two majortypesof approximating and comparative words:in its basicmeaningit isa limitingword,andas suchhasmuchin commonwith comparativequalifters,but it issometimes usedrather in an approximatingsense. 9 Emphatic qualiftersare the leastcommon and most straightforwardof thesethree typesof qualifying expressions.VI 46.1 is a typical example: .

u•ecXovzo , vp•xov• There •zr• stresses the discrepancybetweenthe Rhegians'vaguepromises to contribute largely towards the Athenian interventionin Sicily (VI 6.2 and 8.2) and the mere 30 talents(enoughto pay their fleet for only 15 days--see 8.1) which the Atheniansfound to be available when they

actuallyreachedRhegion.20 It isimportantto distinguish betweenexpressions(like •.6•o•) whichare purelyand basicallyemphatic,and are always

NUMERALS

IN THUCYDIDES

79

so used,and expressions(like o6xk•,dooo•) which are basicallycomparative, but sometimesservean emphaticpurpose(asis arguedbelowp. 85). In the former casetherecan be no doubt about the purposeof the qualifier;

•6•o• alwaysemphasizes the relativesmallness of itsnumber.Thepossible implicationsof a comparativequalifier are, however,more numerousand complex, so that it can alwaysbe debatedwhetheremphasisis intendedin any given case. Both approximating and comparativequalifiers most commonly generalize in someway about a number: that is to say, the qualifier marks the number qualified as being somethingother than the exact figure that appliedor appliesor will apply in oneparticularcaseof the phenomenonin question.• Approximating qualifiersname a figure as beingin the general neighbourhoodof the actual figure in any singlecase,and indicate some degreeof imprecisionand/or uncertaintyregardingthe figure. Comparative qualifiersnamea figure as lyingaboveor belowthe actualfigurein any singlecase.They are usedfrom a desireto delimit, more preciselythan would an approximating qualifier, the range within which a particular figurefalls or shouldfall, and/orto emphasizethe smallnessor largenessof the figure mentioned. Somethingof the variety of motivationsthat can be detectedin Thucydides'use of qualifiersof thesetwo major typeswill be seenfrom the following selectionof examples.I shalldiscussfirst approximating, and then comparative, qualification. One obvioususeof approximationis to refernot to the actualfigurethat obtained in any one instanceof a phenomenon,but rather to an averageof severalfigures.At IV 32.3 Thucydidesdescribeshow the assaultforce led

by Demosthenes againstthe Lakedaimonians on SphakteriaIsland •r• x•r& 8•-•xo•lo•q•r• x• r•>,•1o•,•r•- 8',• i>,0[•o•q;the qualifying phrasespecifiesthat the figure200 may not applyexactlyto any onegroup of men,but isthe approximateaveragefor all thegroups.II 13.3and I197.3 both give averagefiguresof tribute payment over a number of years.The former passagereports Perikles' assertion,in his summary of Athens' resources for war in 43 l, that "600talentsof tributecomesin &• ird •r• everyyear from the allies".]2I197.3 statesthat the tribute collectedby King Seuthes So much of Thrace for approximation amountedto zezp•xoalo•v as a meansof z•X0ivzo•v indicating 0•Oy•Oiou the averageof severalfigures.More commonlyan approximatingqualifier is attachedto a figure that does apply to a single instance of some phenomenon. Sometimesthis is the writer'sway of givingan estimatewhichhe believesis closeto correct, perhapsbecausehe or his informant felt uncertainabout the figure. At V 74.3, having given the casualtyfigures in the battle of Mantineia for three different sectionsof the Quadruple Alliance absolutely, and notingthat the alliesof the Lakedaimonianssufferednegligible

80

CATHERINE

REID

RUBINCAM

losses,Thucydidessaysof the Lakedaimoniansthemselves, Xo•Xer•6• •zkviv

•r•v 0[X•0•0tvrru0•00t•, •:X,•yov.o 8•:r•p[ .p•otxoc•(.ou• •rroOotv•v. The figurecited is an evenhundred;thereforealmostcertainlyroundedto some degree.It is possiblethat the historianhimselfroundeda precisefigure of (say) 286 up to 300. But sincehe emphasizes (at 68.2) the frustrationof his researchesby Spartan secrecy,and statesthe casualtyfigure in such a hesitatingway,13one would naturally infer that he took particular pains with hisenquiries,and that he probablysetdownthe mostprecisefigurehe could get. Strictly interpreted,the languagehere suggests that the qualifying rre0[ was part of the report that Thucydidesreceived;but, given the fluidity with which qualifters are employed, it is doubtful whether his phraseologyshouldbe sopressed.Thucydideshimselfmay haveput in the rre0tas well as the indirectformulation to expresshis doubt about the accuracyof the best figure he could get. At VII 72.3 two approximateestimatesare givenin a parallelconstruction, the former with an approximatingqualifier (doq),the latter with a comparative(•X0•ouq). Theseare estimatesof the numberof shipson each side still capableof fighting after the last battle in the harbour at Syracuse:

In theconfusionof thismomentit seemsunlikelythat eventhegenerals on either sidecould have givenany more precisefiguresfor the shipsstill seaworthy,let alonethat Thucydidesshouldhaveobtainedsuchfigures.14 The differencein the qualifters("about 60" Athenianships;"lessthan 50" Syracusan)reflectsthe tone of Demosthenes' pleato Nikias (thisstatement about the numbers,though not formally part of Demosthenes'speech, providessupportfor his remarks):he contendedthat the Athenianshad a

significantadvantagein the numberof shipsstill seaworthy; and the comparativequalifier with the Syracusanfigure servesto magnify by implicationthe gap betweenthe two figures.]5 In thesecases,then, Thucydidesusedan approximatingqualifier to mark a figure as imprecisebecauseof uncertainty(either his own or his informant's,or both). Elsewherehe sometimesusedthis samemeansof indicatingthat he wasbeingdeliberatelylessprecisethan he couldhave been.

At I 118.2hesumsup the periodembracedby hisexcursuson thegrowth

of Athenianpoweras "50 yearstx0•Xmr0t". He givesthe terminiof this period here as "the retreat of Xerxes" and "the beginningof this war",

NUMERALS

IN THUCYDIDES

81

which would mean autumn 480 to spring431, i.e., 48 •,•au-ro• (or solar yearsbeginningin early spring)or 49 archonyears(480/79-432/1). Here no uncertaintywas involved,but the contextdemandeda generalization; and soThucydidesroundedthefigureout to 50 yearsandadded•z0tX•a'ra. •6 VIII 68.4,the statementthat the oligarchicleadersweretryingto put an end

to Atheniandemocracy•' •zz• •x•:o•:•o [z•X•:• •rcz•Sq• o[ -r6p•vvo• x•:zXGO•a•v is somewhatanalogous.This passagehas troubled commentatorsbecause,while they naturallyfelt that this figure ought to be a round approximation,in fact, by archon-yearreckoning,countinginclusively,as isusuallydonewith ordinalnumbers,l?the yearof the oligarchic revolutionat Athens,412/11, is exactlythe 100thyear after that in which Hippiaswas expelled,511/10. Hencea preference,not alwaysexplicitly labelledassuch,for assumingeitherthat Thucydideswasnot reckoningin archonyears,or that oneof his terrniniwasrather vague,sothat thefigure works out to somethingunder or over 100.•8 I think another sort of explanationis required--one that will explainwhythisfigureisan ordinal. The patternof useof ordinals,ascomparedwith cardinals,by Thucydides to designateintervalsof time suggests that he tendedto useordinalswhere he wishedto be especiallyprecise.•9This ordinal is very unusualin being qualified,somethingthat happensonly four timesaltogether.20If ordinals serveessentiallyto fix an interval of time more preciselythan cardinals, whereasapproximatingqualifiersgenerallyindicatesomeuncertaintyor imprecision,then a qualifiedordinal is a kind of oxymoron.Why did this happenhere?For the point that Thucydidesis making (that the oligarchs faceda difficult task in trying to overthrowthe democracyat Athens,since it had beenin existencefor so long and becomeso firmly rooted) a general statementis all that is required.In thesecircumstances onewould expectto find, asat I 118.2,a round(cardinal)numberqualifiedby •z0tX•a-r•. On the other hand, not only Thucydidesbut also many othersmusthavenoticed the strikingcoincidencethat the oligarchicrevolutionoccurredexactlya century after the end of the tyranny. To registerthis kind of coincidence ordinal reckoning would naturally be used? But the man in the street would not necessarilycountthe archons'nameson the list with the exacting care of a historian, nor take thought for the differencebetweencardinal and ordinal reckoning. And Thucydides,being unable (becausehe was in exile when he wrote this?)or simply unwilling (becauseit seemedunimportant) to take the extra trouble, to checkthe archonlist for himself,was not sure whether the popular date was correct. Thucydideswill then have taken over the ordinal numeral from the popular tradition, and addedthe •z0tX•a-r• to indicatethat he could not vouchfor the figure as exact,22and that in any casehis purposeof the moment(the assertionthat the Athenian democracyhad becomefirmly establishedover a long period so that it was hard to overthrow) demanded only a large round figure. Finally, there are caseswhere the approximatingqualifter expresses

82

CATHERINE

REID

RUBINCAM

doubt about the certainty, but not about the precision,of a figure. This must be the explanationfor the use of [z&X[a-r0t with the figure for the number of shipsbrought to Sicily by Demosthenesin 413 (VII 42.1). Editors have noticed that the several constituents of this fleet enumerated

by Thucydidesadd up, in fact, to exactly73 ships.And somehavewanted, accordingly,to excisethe •.•X,.,'r0t,especiallysincesomeMSS omit it.23 Another reasonfor editorial disquiet, which has not usuallybeenarticulated, is that 73 so obviously cannot be a round number, such as one expectsto appear with an approximatingqualifier. In fact, of 75 qualified figuresabove 20, denotinggroupsof men or shipsinvolvedin battle, thisis the only one which is not a multiple of ten. Since•X,.•z0t here is not marking this figure as consciouslyrounded(as at I 118.2) nor as beingan approximateestimate(as r•zp{did at V 74.3), Dover rightlyglosses it "not 'about' but 'I calculate' (sc. 'assumingthe correctnessof the data from which the total is calculated'...)-.24 If this is correct, then it is natural that the figuresgivenfor the differentconstituentsof thisfleet (at VII 16.2,20.2, 31.3, 31.5, and 33.5) add up to exactly73, for Thucydideshassimplysetout in his narrative

the information

from which he calculated

his own total.

Why he lacked perfect confidencein the result of his calculationswe can only speculate? but he expressedthat lack of confidenceby adding an approximatingqualifierto thefigure.Thus,whereasin mostcasesapproximatingqualifierssignifysomedegreeof imprecision,and in a goodmanya mixtureof imprecisionanduncertainty,here•.•X,.az0tseemsto indicatenot imprecisionbut only uncertainty.This'islogicallyodd,becausethe qualification does not involve any generalization,as do most casesof approximatingqualification.But it is an easyand natural extensionof the useof a qualifier, for which parallelscan readily be found in colloquialEnglish.26 The different usesof comparativequalifierscan be sortedon a similar plan to that appliedto approximatingqualifiers.First of all, the qualification sometimesindicatesthat the figure is not onethat actuallyobtainedor will obtain in any onecase,but ratherdelimitstherangewithin whichsome quantity will fall in any one of the number of cases. Comparative qualifiers are naturally used for this purpose in legal documentsor similarcontexts,wherea prescriptionis madeto applyto all casesof a certaintype,a situationwhereeitherapproximationor absolute specificationwould be inappropriate(the former beingtoo vague,and the latter too specific).Two examplesof this are IV 118.5,a clausefrom the trucemadein 422, laying down a restrictionon thesizeof the shipsthat the

Peloponnesians are permittedto use(A•x•8•lxo,Aou• •.•'rp•), and VIII 65.3, one of the planksin the political platform of the oligarchicconspirators in 411(X6¾oq ze

NUMERALS

IN THUCYDIDES

83

In both thesecasesthe qualifier marks the figure cited as the maximum permissible. The secondpassage is interesting asbeingthefirst mentionin Thucydides of thatgroupof 5000citizenswhoseexistence andresponsibilitiesbecamesucha point of disputeunderthe governmentof the 400. To judgefrom this passage, the organizersof the revolutionregarded5000as the maximum number of citizens who would have a say in governing Athensunder the oligarchy.And this accordsperfectlywith the deliberate vagueness concerningthe precisestatusof the5000whichThucydideslater attributesto the oligarchicleaders(VIII 92.11; cf. also 86.2). The phraseology of their opponents,on the other hand, consistentlyshowstheir feelingthat "the 5000" oughtto be a definitelyconstitutedgroupof that precisenumber(cf. VIII 86.6, 89.2, and 93.2).28 Justasapproximatingqualifiersoftendesignatean average,socomparativequalifiersmay specifyin similarsituationsthehighestor lowestfigure that was, or will be, reached.At VII 28.3, discoursingon the astonishing staying power displayedby Athens, Thucydidesassertsthat at the beginning of the war some thought she might hold out a year, others thoughttwo years,but no one gaveher more than threeyears,whereasin fact sheso far confoundedthesepredictionsas not only to carry on the original struggle, but actually to open a new and no less demanding operation in Sicily in the 17th year of the war. The basicreasonfor the comparativequalifier here is to sum up a number of different predictions by namingthe highestof them. But the contextsuggests that Thucydides used a comparative,rather than an approximating,qualifier partly, at least,so as to emphasizehow largewasthe gap (r6,• 7:o,.•lo•,.... ) betweenthe predictionsand what actually happened. VI 25.2 showsan extensionof comparativequalificationto specifylimits within which a figure will fall in a recommendationfor the future:

This is Nikias' estimatebeforethe Athenianassemblyof the forcesneeded for the proposedexpeditionto Sicily. An expert askedto advisein such circumstancesmost naturally calculateswhat is the smallestnumber that could do thejob, and citesthat figure as a minimum limit with a comparative qualifier, as Nikias doeshere. A subsidiarymotivation,however,was undoubtedlyNikias' pessimismand lack of enthusiasmfor the project. ThucydidessaysNikias hoped,by magnifyingthe difficultieslikely to be encounteredby the expedition,and settinghis estimateof the necessary forces high, to wean the Athenians from their commitment to it. The

84

CATHERINE

REID RUBINCAM

comparativequalifierscontributesignificantlyto this purpose,while the additional injunction, "... and if possible,let them send even more troops", emphasizes explicitly that the figures cited are minimum estimates.

Wherea figureis appliedto a single,actualinstanceof somephenomenon, a writer may usea comparativequalifier to indicatethe limits of his uncertainty,when giving an estimateof a figure not known precisely. Thucydides' statement (IV 72.2) that Brasidashad "not lessthan 6000" hoplitesat Megara in 424 well exemplifiesthis. He citesabsolutefigures(at IV 70.1 and 72.2) for all but oneof the hoplitecontingents amongBrasidas' forces,which add up to 5900, the exceptionbeinga body of troopsthat Brasidashad collectedfor his forthcomingexpeditionto Thrace. Clearly, he roundedthe 5900 up to 6000, and then addedthe comparativequalifier to showthat the group for which he could get no figure numberedat least 100, and probably more. An approximating qualifier would not have indicated so preciselythe limits of his knowledge? This passageshowscomparativequalificationemployedwith a figure containing an element of uncertainty. In other caseswhere there is no uncertaintythis type of qualificationmay serveto showdeliberateimprecisionby the author, the comparativequalifier specifyingnot only that the figure has beendeliberatelyrounded,but whetherthe roundingwasup or down. At 1 54.2 the numberof prisonerscapturedby the Korinthiansin the Battle of Sybotais givenas "not lessthan 1000".But at 1 55.1 Thucydides enumeratestwo groups of prisoners--800 slaves and 250 substantial Kerkyraiancitizens--whichadd up to 1050.Why did he not givethismore precisefigure in the former passage? The context there almostdemandsa measure of generalization,for the figures are cited in virtual indirect discourse,as part of Thucydides'descriptionof the stateof mind that led the Korinthians to claim victory: "they set up a trophy becausethey had had the bestof the battle up until nightfall, sothat they took possession of most of the wrecksand corpses,and becausetheyhad captureda verylarge number of menand ships".What matteredto them,and to him, wasnot the exact numberof men captured,but just that it wasvery high;and 1000was an impressive-sounding round number closeenoughto the true figure.3ø Why "not lessthan 1000"insteadof"about 1000"?First, no doubt, because it was more accurate:he knew the actual figure to be not just "in the neighbourhoodof 1000", but specifically"above 1000". But secondly,I suspect,becauseof the emphaticovertoneof the comparativequalifier' "not lessthan 1000" implies "somewhatmore than 1000", but leavesit to the reader whether he thinks of 1001 or (say) 1100;and the tendencywill always be to suggesta figure quite a lot higher than the thresholdnamed. In severalof the casesjust discussed a desireto emphasizethe magnitude of the figure hasbeensuggested as part of the reasonfor the comparative

NUMERALS

IN THUCYDIDES

85

qualifter.There are somecases,however,wherecomparativequalification is usedonly, or primarily, for emphasis.The clearestexampleis II 31.1-2, where Thucydides enumeratesthe citizen constituent of the Athenian hopliteforcethat invadedthe Megarid late in 431 as"not lessthan 10,000". If we disregardthe qualifying o6x •X0[•ou•, the figuresgiven here for citizen hoplites--10,000 at Megara and 3000 at Poteidaia--add up to preciselythe 13,000mentionedabsolutelyat II 13.6 as Athens'total active forceof citizenhoplitesin 431. Why, then,the qualificationat 31.2?There is no reasonto assumeeither (as at IV 72.2) that he had exact information only up to a point and then took a calculated guessat the unknown quantity, or (as at I 54.2) that he was deliberatelyrounding down some more precisefigureabove10,000.31Thucydidescitesthesefigurestojustify theimmediatelyprecedingstatementthat thisMegarianexpeditionwasthe largestmountedby Athens,"sincethe city wasthen still at the heightof her strength,having not yet been strickenby the plague".The comparative qualifiersare usedpurelyfor emphasis,to magnifyby implicationthe size of the figures? This is a particularlyclearcaseof purelyemphaticcomparativequalification,becausewecanbeprettysurethat Thucydideswasin no uncertainty about thesefigures.There are otherpassages wherea considerable element of uncertaintyprobablyexistedfor Thucydidesor his informants,but the stronglyhyperboliccontext in which the figuresare cited, like that at II 31.2, makes it likely that the primary reasonfor the qualification was emphasis. At VII 27.5, Thucydidesmentionsthat one of the most seriousinjuries Athens sufferedfrom the enemyoccupationof Dekeleia wasthe desertion of"more than 20,000slaves".It isdifficult to imaginewhat kind of detailed informationunderliesthisfigure;but we canbeprettycertainthatnothing like an accurateand preciserecord existed at Athens of the total slave populationbeforethe desertionsbegan,let alone any tally of how many had run awayin anygivenperiod.Any figurefor thenumberof desertions, therefore,would have had to be the mostapproximateof estimates. 33But

giventhecontextin whichthefigurestands, asoneof thedetailssupporting the generalstatementthat the occupationof Dekeleia "hurt the Athenians very much, and becauseof the destructionof property and lossof manpower that it caused, was one of the chief reasonsfor the decline of Athenian power", it seemsmost likely that Thucydidessaid "more than 20,000" rather than "about 20,000" not becausehe had reason to believe

thisfigurewasthe lowestconceivable (i.e.,thatit wasa trueminimum),but becausehe wanted to emphasizehow large the figure was.34 Similarly, when he assertsthat the crowd that streamed out of the

Athenian camp near Syracusefor the final hopelessretreat by land numbered"not lessthan 40,000" (VII 75.5), I cannot believethat this is

86

CATHERINE

REID

RUBINCAM

anything more than the roughestof estimates.If, as seemsprobable,the group enumeratedincludednot only fightingtroops,but hangers-onand slaves,it would be surprisingif even Nikias knew how many they were.35 Why, then, a comparative rather than an approximating qualifier?The context here is highly rhetorical,the whole chapterdevotedto the description of the retreat being written so as to evoke strongemotion, in the style of 'tragic history'. And the figure is cited to justify the statementthat the retreat resembleda processionof refugeesfleeing from a capturedcity, "... and a large[city] at that". This hyperbolictonein the passage suggests that Thucydideswrote "more than 40,000" instead of "about 40,000" chieflyfrom a desireto emphasizethe magnitudeof thefigure.That it isno true minimum is confirmedby its largeness,both by Thucydides'standards and by thoseof modernscholars:it isthe secondlargestnumberappliedby Thucydidesto a group of people,and one of only four numbersfrom 20,000 up?6 and many modernscholarshavefelt that evenif 40,000 were taken as an absolutefigure, or an outsidemaximum, it is probablystill too high.37 This discussionhas focussedon 17 casesof qualification which exemplify the major usesof approximating,comparative,and emphaticqualitiers in Thucydides.These caseswere chosenbecausethey appeared to offer, for variousreasons,the bestchanceof interpretingthe historian's motivation for qualifyinghis numbers.I have suggested in footnotessome other passages that seemto me to resemblesignificantlytheseparadigm cases.There are, of course,many caseswhere, in spite of the guidance providedby a systematicstudyof Thucydides'practiceof qualification,we still lack sufficientinformation to interpret his mind with any confidence. But a betterunderstandingof evena minority of casessufficientlyjustifies the undertaking.38 Numeral qualification is surely a factor that significantlyaffects our interpretationof the numbersin any historicalnarrative,and therefore merits more than the incidental and sporadic attention it has generally received.I hope to have demonstratedhere the kind of useful insights

yieldedby a systematic studyof thisfactorinjust onehistorian.It isclear that Thucydidesusedqualifiersfor a widevariety of purposes.That his motivationwassometimesrhetoricalshouldsurpriseno onewho giveshim creditfor beingan effectivestylistaswell asa painstakingresearcher. 39But it needsto be emphasized that hereisa rhetoricalfactorwhichaffectsour interpretationof a part of his historicaldata, namelynumbers,whichone

mightexpectto befreefrom the distortionsof stylisticartistry? I do not considerthat thisdetractsseriouslyfrom thereputationThucydidesenjoys of beingan unusuallycarefulandaccuratehistorian.To writehistorywell necessarily involvesa constantbalancingof the demandsof rhetoricand

NUMERALS

IN THUCYDIDES

87

objectivity.The largejustificationfor the kind of detailedphilological studyheresetforth istheilluminationit supplies of thatdelicatetensionin the mind of Thucydides.4• Catherine

Erindale College University of Toronto

Reid Rubincam

NOTES

1. The following works are cited by author's name alone: Thukydides,erklfirt von J. Classen,bearbeitet von J. Steup, Berlin 1919-1922;A.W. Gomme, A. Andrewes, and K.J. Dover, A historical commentary on Thucydides,4 vols., Oxford 1945-1970(continuing).The data that form the basisof this studywere

compiled directly fromthetextofThucydides, since it isimpossible toextractfrom' the availablelexicaall the occurrences of suchwordsasnumeralsand prepositions. 2. Clagsen-Steup frequentlynoticequalifyingwords,and cite severalspecific workson the meaningof particularqualifters:see,e.g.,their noteson Thuc.I 13.3-4

(y.dO,•-r00,54.2 (06x •),0•o•) III 68.3 (•), IV 124.1 (6),•.¾o••). GommeAndrewes-Doveralsomake perceptivecommentson many casesof qualification: see,e.g.,thenoteon II 19.1(y.0•),•a'r00; andespecially Dover'ssuggestion (on VI 2-5) that the appearanceof the qualifiers•y¾6gand •¾¾6-r0•-r0• uniquelyin Thucydides' brief sketchof early Sicilianhistory(•y¾6gat VI 2.5and 5.2, •y¾6-r0•-r0• at VI 4.4 and 5.3), togetherwith certain other stylisticpeculiaritiesof thesechapters,reflects Thucydides'use here of a literary sourcewritten in the Ionic dialect, namely Antiochosof Syracuse'sbook on Sicilianhistory..R. van Compernolle,"L'emploi de yA),•-r0•,de •¾¾6•et de •y¾6-r0•-r0• avecdesnomsde nombre,chezThucydide", AC 27 (1958) 5-12, is a rare attemptto analyzethe nuancesconveyedby threeof Thucydides'qualifyingwords. G.M. Kirkwood's discussionof litotesin Thucydides,in "Thucydides'judgmentof the constitutionof the 5000 (VIII, 97,2)", AJP93 (1972) 92-103 (a referencewhich I owe to one of the journal's referees),has some bearing on the interpretation of expressionssuchas 06x •),0•aaovg,although Kirkwood dealsonly briefly with comparativelitotes(art. cit. 97 and 100),and noneof the exampleshe discusses involvesa numeral attachedto the comparativeexpression. I agree in general with Kirkwood's conclusionthat expressionsinvolving litotes,whetherpositive,comparative,or superlative,do not necessarily imply their

opposites. I wouldmaintain,however,that in certainpassages (discussed above pp. 83 if.) thecontextjustifiestheinterpretation of 06x•),0•o• asimplyingnot just "not lessthan" (i.e., "as manyas"), but "morethan"the figurementioned. 3. For example, almost all commentatorsdevote somespaceto 1 74.1, where problemsof both syntaxand substance havebeenfound. In fact,the preposition•g in this passageis not acting as a numeral qualifier, in spite of the numeral that follows it. Seemy article,"Thucydides1.74.1and the useof •g with numerals",CP 74 (1979) 327-337. Similarly, IV 124.1 has been much discussedbecauseof the occurrence there of the unusualqualifyingexpression 6•.•¾ov•g. Scholarswriting on historical(and particularlychronological)problemsthat involveevidencefrom Thucydidessometimeshave occasionto discussthe implicationsof the qualifi-

88

CATHERINE

REID

RUBINCAM

cation attachedto crucial figures:see,e.g., the large corpusof work on Peisistratid chronology, conveniently collected by P.J. Rhodes, "Peisistratid chronology again", Phoenix 30 (1976) 219, n. 1, especiallythe articlesby N.G.L. Hammond (Historia 4 (1955) 371-411) and G.V. Sumner (CQ N.S. 11 (1961) 31-54). 4. Uncertainty and imprecision, though not identical, are often hard to distinguishin practice.Peopletendto uselessprecise(i.e., oftenround)figuresequally when they are guessingand when they lack confidencein the accuracyof some information or calculation. See the discussionabove(p. 82) of Thuc. VII 42.1, arguingthat this is a rare casewhereuncertaintywaspresentwithout imprecision.

5. The numberof occurrences of eachis as follows:y.&),Lazot: 42 cases(on y.&•,La•:• asa numeralqualifier,seethearticleof van Compernolle,aboven. 2), &•: 19, 'rr•p•:10, ztv•: 3 (on •r• as a numeralqualifter,seeClassen-Steupon IV 68.3), •ep•... •.d•az•: 2, k¾¾Og: 2, •'VFOz•z•:2 (on thesetwo qualifiersseeDoveron VI 2-5, cited above n. 2), 6•ov: 2, olov: 1.

6. The figuresare: oOx•.$•ou: 18, 6•.t¾• •.$•ou: 5, 6•.•¾• rr•.•tou: 4, rrk•.ou•:4, o6 r•k•.ou½: 3, •k•ou•: 2, 6rr•p: l, o6 r•okk• •rtvtEk•ov: l, o68& 1,' •p•Xb •k•rrov: 1. I make no distinctionhere betweenthe neuterform of these comparativeadjectives(e.g., obxEX•o•), usedadverbially,and the variousforms of the adjective (obx •X•ou•). B•tant (Lexicon Thucydideurn,Hildesheim 1961 [=Genf 1843]) groupsadverbial and adjectivaluses separately(s.v. Ek•ou• and Ek•ov). C. Preibisch, De cornparativi curn cornparata re coniuncti usu Thucydideo(Bratislava 1869), contendedthat there was a fundamentaldifference of meaningbetweenthe adverbialand the adjectivalexpressions; but I havenot found any suchdistinctionto be consistentlyobserved. 7. •6vo• occursthree times as a numeral qualifier in Thucydides.See above pp. 78 f.

8. These"averaging"expressions occuronceeachas qualifiers. 9. • occurs19timesasa numeralqualifter,•... •k•'r• 3 times,•... 'r•&• twice,and 6•,•¾o•& onceonly.For a discussion of •g asa numeralqualifier,seethe article cited above n. 3.

10. Only two other examplesoccur of a numeralthus qualifiedby a purely emphaticexpression:III 69.2 and VIII 72.1 (on the latter cf. the discussionof VIII 65.3, anotherpassageconcerningthe 5000wherethe qualificationis crucial,above p. 83).

11. Emphaticqualifters,on the other hand,do not generalize;and the figures qualifiedby (e.g.) •.6vo•are real and exact,insteadof approximations. 12. Gomme comments:"The meaning of d• •rr[ 'r6 rro•6 must be 'as a rule', 'generally','by and large' (opposedhere to changesin detail from one year to another)."

13. Classen-Steup(on 68.2) point out the connectionbetweenThucydides' statement(68.2) that the numberof Lakedaimonianswho went into the battle could

not beprecisely ascertained because of thesecrecy of theirpoliticalsystem, andthe qualificationattached(74.3) to the numberof Lakedaimoniandead.

14. Thucydides notes(72.2)thattheAthenians in theirdespondency madeno move to recovertheir dead and their wreckedships.Sinceno more naval battles

werefought,neithersidehadoccasion to repairtheirdamagedvessels at onceand preparefor furtherconflicta fleetthatcouldbecounted.ThatThucydides qualified

NUMERALS

IN THUCYDIDES

89

the figuresfor the numberof shipsthat wentinto the final battleon eitherside(60.4 and 70.4) may indicatethat he couldnot getthesefiguresin precise,absoluteform. 15. Seeabovep. 85, for discussion of this useof comparativequalifters. 16. Note that this deliberategeneralizationby meansof a round-figureapproximation saved Thucydides from having to make a decisionwhich has troubled many of his readers,namely,wheredoesthe excursuson the Pentekontaetiareally begin.There are, notoriously,threeapparentbeginnings--89.1(withdrawal of the Persiansafter Mykale; autumn 479), 94.1 (beginning of Pausanias'campaign; spring 478), and 97.1-2 (first activities of Athens and the newly formed Delian League;spring477); and at 97.1 Thucydidesdefinesthe limits of the excursusas "this war"and "the Medic War", thislatter beinga vaguerexpressionthan that used at 118.2("the retreat of Xerxes") (pace N.G.L. Hammond, CR 7 (1957) 100-101;I cannot believe that Thucydides was as perfectly consistent in his use of the expressionszb M•&x6v, z0• M•&x0•, and 6 M•&xb• •:6•.zy.o•as Hammond suggests;but even if he is right, there is still a discrepancybetweenthe termini of 97.1 and 118.2). Classen-Steupsaw a contradiction between97.1 and 118.2, and proposedtherefore to exciseas a glossthe defining phraseat 118.2. I think this showsa misguideddesireto force Thucydidesto be more precisethan he choseto be. His round approximatefigure("about 50 years")for the periodcoveredby the excursuswas accurateenough no matter which event he took as its starting-point (or what kind of reckoninghe used--archonyearsor •v•zo•, countedinclusively or exclusively).That more than one starting-pointwas possibleresultedfrom the naturaluntidinessof historicalevents:the PersianWarsdid not comecleanlyto an end at Plataia or Mykale; and the Athenianhegemonygrewgraduallyout of the anti-Persian

alliance of 480.

17. P.J. Rhodes(aboven. 3, pp. 220-221)hasrecentlymadethispoint about ordinalreckoning withparticularemphasis. I wouldagreethatordinalreckoning in the Constitutionof Athensand other Greekhistoricalworksseemsusuallyto be inclusive.But Rhodesdoesnot noticethat the qualificationof someof the ordinal numeralsin the accountof the Peisistratidtyrannythat he is examining(at Ath. Pol. 15.1 and 19.2) makes it hard to believe that the use of ordinals in time reckoningwasas straightforwardand totally unambiguousas he maintains.G.V. Sumner(above n. 3, pp. 3740) doesoffer someexplanationfor the qualified

ordinalsin thesepassages. For a freer,and in my opinionmorerealistic,interpretation of ordinal numbers, see Gomme's comment on Thuc. I 18.2.

18. If Thucydideswas countingin •o•,•:o•, the 100th year from Hippias' expulsionbeganat the equivalentpointin thearchonyear412/11 to that at which theexpulsionhadoccurredin 511/ 10.Thusif Hippiaswasexpelledat anytimelater than the monthof Thargelion,in whichthe400seizedpower(cf. Ath. Pol. 33.1), then the latter eventactuallytook placeonly in the99th year(•o•v•:6q)after the former.Gomme(on Thuc.I 18.2)suggests thisasonepossible explanationfor the qualificationat VIII 68.4.(Hammond(aboven. 3, pp. 381-385and389-390)also believesthat Thucydideswascountingby•o•v•:o{, althoughhisconvictionthatthe expulsionoccurredin 510/09 hastheconsequence of makingevena reckoning by archonyearscomeout to 99.) Alternatively,Gommesuggested that Thucydides

had in mind no precisebeginningfor the periodin question,but was"thinking generallyof the period of 103 years514411".

90

CATHERINE

REID

RUBINCAM

19. This can be inferredfrom the following facts:(1) ordinalsare usedto describe intervalsof time in Thucydidesonly about half as oftenascardinals(43 ordinals;93 cardinals);(2) ordinalsare muchlessfrequentlyrounded(of numbersbetween11 and 100, 53% of ordinalsas against71% of cardinalsare multiplesof 5, and 41% of ordinalsas against68% of cardinalsare multiplesof 10; in addition, ordinalsare spreadmuch more evenlyacrossthe scale,84% of them appearingonly once,as comparedwith only 49% of cardinals);(3) the rate of qualificationfor ordinalsis lessthan half that for cardinals(9% of ordinalsare qualified;25% of cardinals). 20. The other casesare: II 19.1 (the Peloponnesians advancedinto Attika from

Oinoeon the 80th day ?.0t•ta'r0t after the eventsin Plataia), III 76.1 (the Peloponnesian fleet arrived at strife-torn Kerkyra on the fourth or fifth day after the transportof a large group of oligarchsto an islandfor their protectionfrom the democrats--this use of alternative figures is clearly a kind of qualification), and VIII 64.4 (the Thasiansbeganfortifying their city in the secondmonth ?.0t•v0• after the departure of the Athenian generalDieitrephes). 21. In other words,I think it no accidentthat the figure specifiedhereis" 100th" rather than "83rd" or "98th". A decimal numeral systemgeneratesa natural tendencyto think in intervalsof five, ten, and 100yearsand multiplesthereof.This showsup in the habit of rounding odd numbersup or down to the nearestmultiple of thesenumbers,as well as in the periodsfor whichpeopleenter into agreements

(seeAppendix,p. 95).In threeotherplacesThucydides employswhatappearto be rounded ordinal numerals(tenth, 20th, and 50th--all nodal points of the decimal numeral system),albeit without qualification, to refer to intervalswhich musthave been subjectsof popular talk. (1) At I 18.2Xerxes'invasionisput in the tenthyearafter thebattle of Marathon. On the natural interpretation, that the yearsmeant are those of the major battles, viz. respectively480/79 and 490/89, the exact figure would be 1l th, not tenth. "Tenth" would be exactly correct, of course,if Thucydideswas counting•¾•0t•ro•, insteadof archons(as Hammond contends,above n. 3), or if his secondterminus was the beginningof Xerxes' invasion,in 481/0, rather than the battle of Salamis (Gomme considersthis possible),but it is equally likely that the figure is a round one.

(2) At VI 59.4 Hippias is saidto havereturnedto Marathon with the Persiansin the 20th year after he had fled to their court from Athens. If the interval meant is that from Hippias'expulsion(511/ 10)to the battle of Marathon (490/89), thenthe exact figure would be 22nd. Other interpretationshave beensuggested (by Sumner and Hammond (aboven. 3) among others)to vindicate20th as an exactfigure,but they are very strained(seethe commentsof Gomme (on I 18.2) and Dover (on V1 59.4) to this effect). (3) At IIl 116.2 Thucydidesrecordsthat the eruption of Etna at the end of the seventhyear of the PeloponnesianWar"was saidto haveoccurred"inthe 50th year after the previous eruption. Both Gomme and Classen-Steupnote that by this wording Thucydideswas passingon the figure as he had receivedit, without vouchingfor its accuracy.They alsoremarkthat, if hisfigureis takenasexact,the date here given for the previouseruption (474) does not agree with that of the Marmor Parium (479). Classen-Steupsuggestthat either Thucydides'figure was rounded, or it should be emended to "55th". Gomme seemscontent to take 50th as

NUMERALS IN THUCYDIDES

91

meantto be exactin spiteof the discrepancy,on the groundthat "Thucydidesisnot vouchingfor the accuracyof his date". I think that the ordinal numeral was includedin the popularreport whichThucydidesheard,and that thosewho framed it were somewhatlessprecisein their calculationsthan he would have been. 22. Cf. the discussionof III 116.2 (above n. 21). 23. See Classen-Steupand Dover, ad loc.

24. R. van Compernolle,in his study of the usesof •.0tXm'r0• in Thucydides (above n. 2, p. 10), singlesout this passageas unique, and cites approvingly A. Croiset'scomment(Thucydide,Histoire de la guerredu Pgloponn•se(Paris 1886)

164,ad I 13.3)thathere"lz0tXm'r0• n'indique'qu'unel•g•re r•servedansl'affirmation (sauf erreur, si je ne me trompe)'". 25. Steup wonderedif Thucydideswas in doubt about Eurymedon'scontribution to the fleet. Other possibilitiescould also be suggested.It may havebeenjust that he couldget no precisetotal figurefrom any trustworthyinformant. Compare the reconstructionsketchedby Gomme, on V 68.3, of the stagesin Thucydides' enquiryinto how many Lakedaimoniansfought at Mantineia. In that casehe gave no total figure, but simply set out the information he had obtained on the organization of the Lakedaimonian army, leaving his readersto calculatethe total on that basis.At VII 42. l the difficulty will have beenno suchdeliberatesecrecyas he had met at Sparta, but perhapsthat, sincethe Athenian generalsdid not long survivetheir surrender,and Thucydidesprobablyhad difficultyanyway,becauseof his banishment,interrogatingthose Athenians who did come home, he could find no one who knew exactly the figure in question. 26. I recently overhearda friend telephonea colleagueand ask, "Am I right in thinking that we have approximately $147.80 in the Visiting LecturersFund?" When I askedthe speakerwhy shehad qualifiedsoprecise-sounding a figure(which could not be rounded), she replied, "Well, I wasn't sure if I had rememberedthe amount right". This is an exactparallelto what I think happenedat ThucydidesVII 42.1, except that the reason for the speaker'suncertainty was not the same. Similarly,"perhaps"in Englishand•c•toq in Greekare both usedasapproximating qualifters(•otogis so usedby Polybios,though not by Thucydides).Thesewords

indicatebasicallyuncertaintyratherthanimprecision,whereaswordslike•eO•and "about" specifysomedegreeof imprecision;but in neitherlanguageare thesetwo differenttypesof qualifterskept strictlyseparatein practice. 27. In the former passagethe qualifter is •q, which is not strictlya comparative word, but functionsoften in the sameway; seeabovep. 78 and the article citedin n. 3.

28. Classen-Steup'scommentson VIII 67.3, 93.2, and 97.5 suggestthat they would similarly emphasizethe comparativequalifterat 65.3 as contributingto the confusion over the identity and responsibilitiesof the 50•. Aristotle, A th. Pol. 29.5, reportsthe oligarchs •programmeusinga qualifterof theoppositeimport:

•6Xz•oq • I shouldlike to believethat thisisa reformulationgovernedby whatthe populationat largewishedto hearratherthan what the oligarchsactuallyintended. In fact, Aristotle goes further than Thucydides in not only referring to •zv•xmX[Xm• thereafter(29.5, 30.1, 31.2, 32.1) as if they werea body of precisely

92

CATHERINE

REID

RUBINCAM

5000 peoplewhich had somereal existence,but alsoascribingvariousactionsto them, while he nevertheless insists(32.3) that o[ ... rreYr0•x•aX•.3.•o• X6¾• •z6¾o¾ ,•060•a0•¾(cf. Thuc. VIII 92.11). 29. Classen-Steup commenton ro6½•ze0' 0•6ro•6•o• •8• •u¾•X•¾•z•¾o• •,• (70.2): "etwa 100 an der Zahl, wie sichbei VergleichungunsererSt[elle] mit c. 72,1 u. 2 ergiebt." This seemsto imply that Thucydidesknew how many were in this

troop,butsimply didnotheregive•thefigure.I thinkthisinterpretation gives too little weightto the comparativequalifierat 72.2, whichsurelysuggests rather, in the circumstances, that Thucydideswastaking a calculatedguessat the one unknown constituent figure, and therefore at the overall total. 30. Cf. VIII 68.4, discussed abovep. 81, whereit wasarguedthat the approximating qualifier •z&•,•a'r•was motivated partly, at least, by the demandsof a similarlygeneralizingcontext. In the presentpassagethe Korinthiansare alsosaid

to havecaptured.r:e[:•. [•38o•t•xo¾'r0• shipsfrom the Kerkyraians. Why thisfigure shouldbe qualified by an approximatingrather than a comparativeexpressionwe can only guess:perhapsthis wasthe mostprecisefigureThucydidescouldobtain. 31. On eitherof theseassumptions, Thucydidesmusthavebelievedthat Athens had more citizenhopliteson activeserviceat the end of the campaigningseasonof

431 than the 13,000whichhe mentionedin hislistof herresources in the springof that year. This seemsunlikely. 32. Althoughno suchclearevidencerulesout otherpossiblemotivationsfor the qualificationof the meticfigureat 31.2, both analogyand the generalargument from the moodof the passage suggestthat thistoo is purelyemphatic.It hasbeen suggestedto me that perhaps a more prosaic explanation could account for

Thucydides'useof the qualifier"not lessthan" with troop numbers:giventhat in ancient,as in modern, armiesunits going into battle were rarely up to paperstrength,might Thucydideshave usedthis kind of qualificationsimplyto mark figuresthat he knewto be moreexactthan merepaper-totals?If thiswerecorrect, one wouldexpectthat in at leasta fair numberof casessuchqualificationof troop numberswould appear unmotivatedby any rhetorical factors in the immediate context.In fact, of the 220figuresgivenby Thucydides for troopsof mengoinginto battleor on expeditions,only sevenare qualifiedby o6x•3.&aaou½. Two of theseare the figuresat II 31.2, under discussionhere. Three more are also discussedin this paperascaseswherethe contextclearlysuggests a particularreason,at leastpartly

rhetoricalemphasis, for thequalification (VI 25.2onp. 83;VII 75.5onpp.85f.;II 98.3 in n. 34). In each of the remainingtwo passages(IV 72.2--the forcesof the Peloponnesianalliance at Delion includedo6x [X•aao,• 6000 hoplites;and VI 67.2--the Syracusanshad o6x •3.•aao,•1200cavalryin the first battle againstthe Athenian invadingforce), emphasison the size of the numberby meansof the qualifier would equally well suit the tendencyof the context. 33. Cf. the commentsof M.I. Finley, Theancienteconomy(Berkeleyand Los Angeles1973)24, on thisfigure.' Theremay havebeena recordof the numberof slavesownedbythecity,astheseconstituted an importantpartof Athens'property, and perhapsalsoof thoseemployedin the silver-minesat Laureion.But thiswould stillleavemuchof the slavepopulationunaccounted for. And wouldThucydides (if he wasin Athens)or anyoneelsehavetakenthetroubleto goaroundandaddup the lossesreportedby eachownerof slavesduringtheperiodwhenthe Peloponnesians were occupyingDekeleia?

NUMERALS

IN THUCYDIDES

93

34. It was suggestedabove(p. 80) that the comparativequalifter attachedto the numberof Syracusanshipsmentionedat VII 72.3 wasmotivatedby a combinationof uncertaintyand a desireto emphasizethe differencebetweenthe numberof shipson eitherside.This caseis thereforeanalogousto VII 27.5and VII 75.5. Other passages wherethe rhetoricaland sometimeshyperbolictone strongly suggests thatcomparativequalifiersarebeingusedat leastpartlyfor emphasis are:I 74.1 (6My• •X0•ou½--the proportionof the Greekfleetof 480 suppliedby Athens; cf. thearticlecitedaboven. 3); II 13.4(o6x•X0mao,•--thevalueof uncoinedgoldand silverin the templesof Athensin 431,accordingto Perikles);I198.3 (o6x the sizeof the Thracian army led into Macedoniaby Sitalkes;cf. n. 36 below);III 87.3 (o6x •X0•aaou½--losses from the ranks of Athens'hoplitesand cavalrydue to the plague);III 113.4(•X•o,•--a despairingestimateof the numberof despoiled

suitsof armourfrom an overwhelming defeat);VI 1.2(06 •oXX• 'r•,• lengthof thevoyageroundSicily);VII 70.4 ([3p0tX6 0•x•X[xo,•--thetotal numberof shipsinvolvedin the last battle at Syracuse);VII 87.4 (06x •X•aaou½--thenumber of prisonerstaken after the defeat at Syracuse). 35. Both the generaltermsof his description and hisuseof thephrase6 6XXo ½suggest that Thucydideswascountingnot only hoplitesand cavalrybut also sailorsand light-armedtroops, and possiblyeven slaves.Dover and G. Busolt (GriechischeGeschichte,III.ii (Gotha 1904) 1370, n. 3) agree that this figure includesslavesand other supernumeraries.Dover's inferencethat, having first calculatedhow many sailorsand fightingtroopsthe expeditionshe reports(at VI 43, VII 16.2and42.1) originallytransportedto Sicily,Thucydides thenappliedthat samefigurehereto thenumberof fightingtroops,sailors,andslavesthatsetout on the retreat,"on the assumption that thenumberof slavesleft wasroughlyequalto the number of soldiersand sailorskilled, left behind, or missing",puzzlesme for several reasons.First of all, Dover seemsnot to set out for inspectionthe calculationby whichhe believesThucydidesarrivedat 40,000asthe total numberof fightingtroopsand sailorssentto Sicilyby Athens;and I havenot foundit easyto duplicate.Even if this figure is correct,however--and Busoltoffersa rival estimate, plausiblydocumented, of 50,000--thereis nothingto showthatThucydides made any suchcalculation.And it hardlyhelpshiscreditasa historianto assumefirst that he laboriously worked out, but neglectedto tell his readers, that Athens sent altogether40,000 sailorsand soldiersto Sicily, and then that he so arbitrarily decidedthat thenumberof slaveswhojoinedthe retreatcompensated moreor less exactlyfor the lossesamongthe fightingtroops.I find it easierto believethat the figurewasgivento Thucydidesby oneof the informantswho suppliedsomeof the vivid details in the descriptionof the retreat, and that it was the product of emotionalrather than rational processes. 36. The otherthreeare:II 98.3--the army of Thraciansthat invadedMacedonia

underSitalkesX•y•-r0t• o6xEX0t•o,•x•,rr•x0t•8•x0t •u•0tSto,•y•,•00tt; V 63.2--the Lakedaimonians proposedto fine King Agis8•x0t•upt0tm8p0tX•&,• for misconductingmilitary operationsagainstArgosand Orchomenos;VIII 27.5--the "more than 20,000 slaves"that fled from Athensto Dekeleia(discussed abovep. 85). 37. On the accuracyof thefigureDoverdoesnotcommithimselfunequivocally, but hisnotesonVII 82.3and87.4suggest that hefoundit hardto believein sohigha rate of losson theretreatasThucydidesimplies(leavingonly"not lessthan7000"of the original40,000to be takenprisonerat theend--97.4). Busolt(aboven. 35) after

94

CATHERINE

REID

RUBINCAM

discussingthesefiguresat lengthin the light of his estimatesof the total Athenian strengthin Sicily (50,000sailorsand soldiersaltogether),and of the lossesfrom this total mentioned or implied by Thucydides,calculatesthat no more than 25,000 combatants can have survived the final battle, so that one would have to add an

improbable15,000supernumeraries to make up Thucydides'40,000. He therefore concludes, "Die Zahl 40,000 beruht h6chst wahrscheinlich auf einer zu hohen

Sch•itzung,aberjeden•ihereBerechnungschwebtin der Luft". With thisconclusion I would agree. Similarly, J.H. Finley remarks(Thucydides (Cambridge,Mass. 1942) 244) that Thucydides'"no lessthan forty thousand ... [is] seeminglyan excessive figure in view of the previouslosses".Equallysignificant,to my mind, is the manipulationof qualifierscarriedout by P. Green(Armadafrom Athens(New York 1970) 318-319) and B.W. Henderson(The Great War betweenAthens and Sparta (London 1927)) to minimize Thucydides'figure: the former writes, "The whole force numberedsomeforty thousandmen ..." (my emphasis),the latter simply, "Forty thousand men had begun the retreat". 38. A scepticmight objectthat in many,or evenmost,of thecasesherediscussed onecannotbe absolutelysurethat the qualificationwasattachedto thenumeralby Thucydidesrather than by his informant(s). This is quite true. It seemsto me, however, that the closeintegration of a qualifier with the generalsenseand the rhetorical emphasisof a piece of narrative, such as is found in many of these passages,suggeststhat the samepersonwasresponsiblefor both. And mostreaders would surelybe reluctantto believethat Thucydidesmadea practiceof recording his informants' testimony verbatim and unedited. 39. For a recent general treatment of some of the more emotional aspectsof Thucydides'work see J.R. Grant, "Toward knowing Thucydides",Phoenix 28 (1974) 81-94. 40. A rhetoricalinfluenceon someof Thucydides'figureshasoccasionallybeen suggested,as generallyby Grant (seelast note), and more specificallyby S. Dow, "Thucydidesand the number ofAcharnian Hoplitai", TAPA 92 (1961) 66-80, who arguedthat the incrediblylargefigure of 3000 givenby Thucydidesfor the number of hoplitesfrom Acharnai (II 20.4) was best explainedas an exaggerationmade under the spell of rhetoric. 41. The researchon whichthis paper is basedwascarriedout while the author wasenjoyingthesupportof a CanadaCouncilLeaveFellowship.An earlierversion of this paperwaspresentedto the ClassicalAssociationof Canadaat its meetingin Fredericton, New Brunswick, in June 1977. I am grateful to all those who were presenton that occasionand offered commentsand criticism.In addition, I thank the journal's refereesfor somesuggestions whichhaveimprovedthe argumentin several places. Finally, I owe special thanks to ProfessorJ.M. Bigwood for helpfullydiscussing a numberof passages from Thucydides,and to ProfessorsJ.R. Grant and M.B. Wallacefor readingand criticizinga secondversionof this paper. They should not be taken to agree with all the opinions expressedhere, nor regardedas sharingresponsibilityfor any shortcomingsin this final version.

NUMERALS IN THUCYDIDES

APPENDIX

TO THE

95

NOTES

Evidenceof Rounding of Numbers in Thucydides Figuresfor troopsof men going into battle or on expeditions(i.e., excluding casualtyfigures)number220in all, and215of these(=98%)aremultiplesof ! 0. The highernumbersalsoshowroundingto othernodalpointson thedecimalscale,viz. 50, 100,500, and 1000.The detailsare as follows. Of the 20 numbersfrom 1 to 100, 17 (=85%) are multiplesof 10; of the 80 numbersfrom 101 to 500, 78 (=97%) are

multiplesof 10, 72 (=90%)are multiplesof 50, and 66 (=82%)are multiplesof 100. All of the 60 numbersfrom 501 to 1000are multiplesof 50, and 59 of them(=98%) are multiplesof 100.All of the60 numbersabove1000aremultiplesof 100,while45 of them (:75%) are multiplesof 500, and 36 (:60%) are multiplesof 1000. Truces,treaties,and alliancesmentionedby Thucydidesare for periodsof: two days(IV 114.2),ten days(V 26.3; VI 7.4), four months(V 60.1), oneyear(IV 117.1), five years(I ! 12.1), 30 years(I 23.4; V 14.4), 50 years(V 18.3;V 41.2; V 79.1), and 100 years (III!

14.3; V 47.1).

[For the use of qualifterswith numeralsin the Athenaion Politeia, seenow my article "Qualification of numerals in the Constitution of Athens'; Phoenix 33 (1979) 293-307, discussingin detail matters touchedupon in n. 17 above.]