Studies in the Ancient Greek Polis 3515067590, 9783515067591

Contents: F. de Polignac: Repenser la „cité“? Rituels et société en Grèce archaïque ― M. H. Hansen: The „Autonomous City

140 62 7MB

English Pages 219 [226] Year 1995

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
CONTENTS
François de Polignac, Repenser la “cité”? Rituels et société en Grèce archaïque
Mogens Herman Hansen, The “Autonomous City-State”. Ancient Fact or Modern Fiction?
Mogens Herman Hansen, Kome. A Study in How the Greeks Designated and Classified Settlements which were not Poleis
Thomas Heine Nielsen, Was Eutaia a Polis? A Note on Xenophon’s Use of the Term Polis in the Hellenika
Pernille Flensted-Jensen, The Bottiaians and their Poleis
Stephen G. Miller, Old Metroon and Old Bouleuterion in the Classical Agora of Athens
T. Leslie Shear, Jr., Bouleuterion, Metroon and the Archives at Athens
Alexandru Avram, Poleis und Nicht-Poleis im Ersten und Zweiten Attischen Seebund
Walter Burkert, Greek Poleis and Civic Cults: Some Further Thoughts
Lene Rubinstein, Pausanias as a Source for the Classical Greek Polis
Recommend Papers

Studies in the Ancient Greek Polis
 3515067590, 9783515067591

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

Mogens Herman Hansen and Kurt Raaflaub (Eds.)

Studies in the Ancient Greek Polis HISTORIA Einzelschriften 95

Franz Steiner Verlag Stuttgart

MOGENS HERMAN HANSEN

AND KURT RAAFLAUB (EDS.) STUDIES IN THE ANCIENT GREEK POLIS

CONTENTS François

dePolignac, Repenser la “cité” ? Rituels et société enGrèce archaïque

Mogens Herman Hansen, Modern Fiction?

The“Autonomous City-State” . Ancient Fact or

Mogens Herman Hansen, Kome. A Study inHowtheGreeks Designated and Classified Settlements which werenotPoleis Thomas Heine Nielsen, WasEutaia

TermPolis intheHellenika

a Polis? A NoteonXenophon’s Useof the

7

21 45 83

Pernille Flensted-Jensen, The Bottiaians andtheir Poleis

103

Stephen G. Miller, OldMetroon andOldBouleuterion Athens

133

in the Classical Agora of

T. Leslie Shear, Jr., Bouleuterion, Metroon andtheArchives atAthens Alexandru Avram, Poleis undNicht-Poleis Attischen Seebund

imErsten undZweiten

157

191

Walter Burkert, Greek Poleis andCivic Cults: SomeFurther Thoughts

201

Lene Rubinstein, Pausanias asa Source fortheClassical Greek Polis

211

REPENSER LA “ CITÉ” ? RITUELS ET SOCIÉTÉ EN GRÈCE ARCHAÏQUE* par FRANÇOIS DE POLIGNAC L’étude du monde grec de la fin de l’âge du bronze aux premiers temps de l’archaïsme est sans conteste undesdomaines qui, dans les sciences del’Antiquité, a été le plus profondément renouvelé depuis unevingtaine d’années, aupoint que l’appellation traditionnelle de“ siècles obscurs”pourcette époque n’est plus guère utilisée. Les grandes synthèses archéologiques desannées 70, essentiellement anglosaxonnes1, ontouvert la voie à toute unegénération derecherches qui, combinées à de multiples découvertes, ont fait de la Grèce des débuts de l’âge dufer le lieu non seulement d’un enrichissement notable en données nouvelles, mais aussi d’une réflexion thématique et méthodologique qui a fini parremettre en cause certaines desidées les plus communément admises, celles-là mêmes quiavaient fourni à ces recherches leur point de départ. Une de ces idées, et l’une des plus importantes dans les orientations de la recherche, est quela plupart deschangements observables dans la société grecque dansladeuxième moitié del’époque géométrique etaudébut del’époque archaïque, soit dela fin duIXe auVIIe siècles, pouvaient être misen rapport d’unefaçon ou d’uneautre avec unphénomène historique majeur: la naissance dela cité, celle-ci entendue comme un type de société et d’organisation politique appréhendé par référence aumodèle dela cité classique et desesinstitutions. Cette vuea cependant été vivement critiquée deplusieurs parts, encore quedefaçon contradictoire. D’un la cité” , dumoins d’unecertaine forme de côté, il estclair quel’existence sinon de“ polis présentant quelques uns des traits censés caractériser la cité (un temple, un conseil, des assemblées), peut être décelée dans les poèmes homériques. Certes, ni l’Iliade nil’Odyssée nesont desimples documents historiques ousociologiques, et leurs évocations présentent souvent des contradictions, des traits fictifs ou des réminiscences d’autres temps ou d’autres sociétés; néanmoins, le tableau de communautés distinctes possédant uneorganisation collective, mêmerudimentaire,

* 1

Cetexte estcelui d’uneconférence présentée à l’Académie royale dessciences etdeslettres du Danemark le 14 novembre 1993, à l’invitation du Professeur Mogens H. Hansen, que je remercie vivement, ainsi quelesparticipants quiontbienvoulu mefaire partdeleurs observations. V. R. D’A. Desborough, TheGreek DarkAges, London 1972; A. Snodgrass, TheDarkAgeof

Greece, Edimbourg 1971; Archeology and the Rise of the Greek State, Cambridge 1977; Archaic Greece. Theage of experiment, Cambridge 1980; N.Coldstream, Geometric Greece, Cambridge 1977; et plus récemment, A. Snodgrass, AnArchaeology of Greece. Thepresent state andfuture scope ofa discipline, Berkeley/Londres 1987 (Sather Classical Lectures 53).

François dePolignac

8

et exhibant

leurs premières réalisations monumentales, n’aurait

puêtre composé si

ces éléments avaient été totalement inconnus dans la Grèce géométrique2. Certains vont même plus loin et considèrent que ces communautés étaient en fait les héritières des“ cités”del’âgeduBronze quimenaient leur existence propre à côté

ouendessous del’organisation palatiale minoenne oumycénienne, conçue comme unesuperstructure parasitaire dontladestruction aurait enfaitlibéré cescommunautés

et leur aurait permis dedévelopper graduellement leurs potentialités toutaulong de l’âgedufer3. Del’autre côté, onpeut aussi bien montrer quele niveau d’intégration

sociale et politique caractéristique de la cité classique n’a généralement pas été atteint avant le VIe siècle: démocratique ounon, le modèle delapolis tel quesel’est forgé l’historiographie moderne présente un degré d’élaboration formelle et d’abstraction inconcevable sans les réformes de l’époque archaïque. La cité parachevée apparaît alors comme unphénomène plus tardif. A considérer ces différents points devue, onenarrive à la conclusion qu’il n’y ajamais eudegenèse, ounaissance, ouformation dela cité quipuisse être placée à uneépoque donnée, quelle qu’elle soit, et quela cité classique est le résultat d’un très long processus évolutif continu, sans seuil marquant ou décisif4. Cette interprétation, poussée jusqu’aubout de sa logique, finit parrendre toute analyse historique impossible: si on substitue sans cesse le principe d’héritage auprincipe de causalité, l’explication n’est plus que renvoi à un état originel toujours plus reculé, donc pour finir inaccessible. Elle peut néanmoins constituer uneutile mise en garde contre une vision trop radicale de la rupture entre la Grèce de l’âge du bronze et la Grèce archaïque, oucontre les systématisations excessives auxquelles peutmener l’abusdela notion decité. Ainsi, recourir à unehypothétique “idéologie , marquée parle concept d’isonomie, pourexpliquer leschangements de delapolis” pratiques collectives à la fin de la période géométrique, illustre bien le type de raisonnement circulaire auquel peutmener l’usage denotions toutes faites5: comment, eneffet, plaquer surlapériode oùla cité estcensée prendre forme uneidéologie de l’isonomie elle-même considérée comme le résultat dudéveloppement historique dela cité? Les sociétés archaïques évoluaient enremodelant leur propre passé, et nonentendant vers unidéal abstrait présent à leur esprit comme le butà atteindre. Cen’estcependant làqu’unaspect particulier d’unproblème beaucoup plus général sous-jacent à tout usage duterme polis: quand l’historien étudie la société grecque d’époque géométrique et archaïque, peut-il se passer desoutils intellectuels forgés paret pourl’analyse dela cité classique, peut-il s’interroger surles premiers temps dela cité sans se référer presque automatiquement auxconcepts et représentations

2 3 4

5

19; K. Raaflaub, “Homer to , AION2, 1980, 7– Ithaque oula naissance dela cité” “ , TheAncient Greek City State (Acts of the Solon: the Rise of the Polis. The written sources” 59. Copenhagen Polis Centre 1),éd.M.H.Hansen, Copenhague 1993, 46– H. vanEffenterre, La cité grecque. Des origines à la défaite deMarathon, Paris 1985. , Greece between East and Introduction” Tel est le point de vue, parexemple, de S. Morris (“ West, 10th– 8thcent. BC, éd.G. Kopcke, I. Tokumaru, Mainz 1992, xvii). C’est le défaut despremiers travaux deI. Morris: Burial andAncient Society. The rise of the , Antiquity 62, Tomb Cult and the Greek Renaissance” Greek city-state, Cambridge 1987; “ 761. 1988, 750–

Cl. Mossé,

Repenser la “cité”?

9

attachées au modèle de la polis classique dans l’historiographie moderne6? La vraie question est alors de savoir si, plutôt que de chercher à identifier l’époque et le lieu où il est possible de tracer une ligne entre un “avant la cité” et le temps de la cité, il ne faudrait pas se débarrasser de la notion même de polis qui, quelque effort que l’on fasse, serait davantage une étiquette trompeuse qu’un concept heuristique. Faut-il, autrement dit, oublier la cité pour penser la société ? Le problème, on le sait, résulte en partie de l’ambiguïté du terme qui, chez les théoriciens antiques comme chez les historiens modernes, recouvre à la fois une réalité politique, institutionnelle, et une entité sociale, dont l’articulation mutuelle est source de longs débats et fréquents malentendus selon que l’on entend le “politique” dans une acception restreinte - la définition de la citoyenneté au sens institutionnel - ou dans une acception large - la délimitation du koinon, de ce qui définit la communauté7. Toute analyse qui privilégie un de ces aspects ne peut rendre compte ni de la globalité des phénomènes historiques, ni de la façon dont varie l’agencement des différentes composantes de la vie sociale. Il faut donc, pour échapper au dilemne, déterminer le type d’analyse susceptible de dévoiler comment les diverses composantes de la vie collective se déterminent l’une par rapport à l’autre. L’étude des pratiques rituelles ou ritualisées me semble une des plus aptes à répondre à ces exigences, d’autant plus qu’elle a considérablement évolué ces dernières années: sa relation avec la question de la “cité” peut être désormais examinée en termes plus précis. L’évolution des pratiques rituelles à la fin de l’époque géométrique: essor des sanctuaires et de leurs offrandes, accompagné du déclin ou de la disparition des offrandes funéraires, qui donne l’impression d’un transfert d’un domaine à l’autre, a longtemps été présentée comme l’indice du passage d’une société centrée sur l’expression individuelle à une société dominée par l’expression communautaire, celle de la cité; cette vue très générale a maintenant subi quelques retouches8. Il apparaît de plus en plus clairement que les pratiques funéraires et cultuelles entretenaient parfois un rapport de complémentarité plus que d’opposition, en fonction de leur contexte. En effet, le milieu de réception des premières était plutôt local ou régional, dans la mesure où les rites funéraires mobilisaient avant tout les réseaux proches de parenté, d’alliance et de pouvoir de l’élite et demeuraient fortement enracinées dans des traditions culturelles locales bien définies; inversement, l’essor des pratiques cultuelles concerne en premier lieu, aux Xe et IXe siècles, des sanctuaires de contact 6

7

8

W. Gawantka, Die sogenannte Polis. Entstehung, Geschichte und Kritik der modernen althistorischen Grundbegriffe: der griechische Staat, die griechische Staatsidee, die Polis, Stuttgart 1985; E. Lévy, “La cité grecque: invention moderne ou réalité antique?”, Du pouvoir dans l’Antiquité. Mots et réalités, ed. Cl. Nicolet, Genève 1990, 53-67. Voir à ce sujet les analyses d’O. Murray, “Cities o f reason”, The Greek City from Homer to Alexander, éd. O. Murray, S. Price (Oxford 1990), 1-25; M. H. Hansen, “The Polis as CitizenState”, The Ancient Greek City State, 1993 (n. 2), 7-29; J. Ober, “The Polis as a Society”, ibidem, 129-160. Snodgrass, Archaic Greece (n. 1), 52-54; I. Morris, Death Ritual and Social Structure in Classical Antiquity, Cambridge 1992, en particulier 25-28; F. de Polignac, “Entre les dieux et les morts. Statut individuel et rites collectifs dans la cité archaïque”, The Role of Religion in the Early Polis (colloque de l’Institut suédois d’Athènes, octobre 1992) (à paraître).

10

François dePolignac

et de rassemblement à vocation régionale ou interrégionale où rites et offrandes s’inscrivaient dans un contexte élargi, établissant une médiation entre diverses ’ lympie, le plus riche en offrandes de communautés ou groupes sociaux. Celui dO prestige de l’époque géométrique, faisait ainsi office de lieu de reconnaissance et d’auto-identification pour l’élite des basileis péloponnésiens9. De plus, dans le cadre strictement régional, l’articulation logique et chronologique entre l’évolution despratiques funéraires et cultuelles auVille siècle varie d’uncas à l’autre. Si la coïncidence est réelle en Corinthie entre l’appauvrissement des tombes et le réaménagement (ainsi quel’enrichissement) dessanctuaires vers le milieu duVille siècle, enArgolide aucontraire l’apparition descélèbres tombes à armes ouarmures et obeloi estcontemporaine del’apparition d’offrandes deprestige (par exemple de chaudrons à trépied monumentaux) à l’Héraion deProsymna dansla seconde moitié dusiècle. Dansce dernier cas, l’évolution desoffrandes dusanctuaire et l’évolution desrites et offrandes funéraires tendent toutes deux à renforcer les distinctions au sein de la société, avant que toute forme de distinction funéraire ne disparaisse quand, à partir de 700, l’usage des ensevelissements en pithos, sans offrande, se généralisa10. L’Attique présente untableau encore plus particulier; l’apparition de 700, bronzes monumentaux surl’Acropole s’insère en effet brièvement, vers 740– redeviennent, puis restent, funéraires une les rites où des entre deux périodes principales formes d’expression ritualisée del’élite athénienne. Mais ces rites sont eux-mêmes divers et fluctuants. Deuxmodèles lesdominent encore danslapremière moitié duVIIIe siècle. Dans le premier, où la priorité est donnée à la richesse du contenu dela tombe, la fonction d’expression dustatut dudéfunt est concentrée sur les rites accompagnant l’offrande et la déposition des objets; dans le second au contraire, attesté principalement auCéramique, la priorité est donnée à la visibilité extérieure dela tombe parle biais d’unmarqueur monumental, emblème et témoin de la mémoire collective réservée à certains défunts. A partir de la fin dusiècle, après unephase de transition oùplus aucune règle précise ne semble encadrer les pratiques funéraires attiques, unnouveau rituel exclusif tend à combiner les deux tendances précédentes: les offrandes, souvent de la poterie de qualité évoquant l’univers dubanquet, sont alors en règle générale exposées puis brûlées à côté de la tombe, dans les tranchées aménagées à cet effet (les Opferrinnen), tandis que la unmarqueur (vase, stèle)11. tombe reste signalée paruntumulus souvent surmonté d’ 9 C. Morgan, Athletes andOracles. Thetransformation of Olympia andDelphi intheeighth cent. BC,Cambridge 1990; F. dePolignac, “Mediation, competition andsovereignty. Theevolution , Placing the Gods. Sanctuary andSacred Space in of rural sanctuaries in geometric Greece” 18. ancient Greece, éd. S. Alcock et R. Osborne, Oxford 1994, 3– Theevolution of a ‘sacral landscape’: 10 Corinthe: I. Morris, Death ritual (n. 8), 25; C. Morgan, “ 142, enpart. 128– , Placing the Gods (n. 9), 105– Isthmia, Perachora andthe Corinthian Gulf” , The 132. Argos: R. Hägg, “Burial customs andsocial differenciation in 8thcentury Argos” 31; F. dePolignac, Greek Renaissance intheEighth Century, éd.R. Hägg, Stockholm 1983, 27– , Argos et l’Argolide. Topographie et urbanisme, éds. A. Cité et territoire: unmodèle argien?” “ Pariente et G.Touchais (EFA, Recherches franco-helléniques 3) (sous presse). 172; S. Houby11 J. Whitley, Style and Society in Dark Age Greece, Cambridge 1991, 137– Nielsen, “ Interactions between Chieftains and Citizens? 7th cent. BC Burial Customs in 374. , Acta Hyperborea 4,1992, 343– Athens”

Repenser

la “cité” ?

11

L’évolution desrites funéraires est sans nuldoute étroitement liée à leur plus ou moins grande diffusion dansla société, comme l’a bien montré IanMorris à propos de l’Attique: l’usage discriminatoire de rites et de catégories d’offrandes très spécifiques (les broches et pièces d’armement dans l’A rgolide duVIIIe siècle, les vases de style orientalisant protoattique dans l’Athènes duVIIe siècle) correspond à dessociétés oùl’élaboration formelle desfunérailles et dessépultures était réservée àuneélite restreinte, ouencore objet deconflits d’appropriation, tandis quel’extension du“droit à la sépulture formelle”se traduit soit parunedésorganisation dusystème symbolique, comme dans l’Attique de la seconde moitié du VIIIe siècle, soit par l’homogénéité des nécropoles et l’abandon (apparent) des pratiques et offrandes ostentatoires (Argolide, Corinthie)12. Doit-on pour autant y reconnaître la victoire, l’idée decité” dans le dernier cas, oula défaite dans le premier, de “ , à la suite de conflits entre agathoi et kakoi qui auraient contraint les aristocraties argiennes et corinthiennes à se couler dans le moule de l’isonomie mais exclu les non nobles attiques detoute forme d’expression ritualisée13? Cette interprétation estévidemment trop simpliste pour rendre compte de phénomènes dont les facettes sont plus y paraît aupremier abord. nombreuses et les implications plus diverses qu’il n’ L’exemple athénien attire enpremier lieu l’attention surl’erreur deperspective qui consisterait à confondre systématiquement l’absence d’offrandes deprix dans les sépultures avec la disparition de toute forme d’expression du statut et de compétition dans le domaine funéraire. Les pratiques ostentatoires del’aristocratie attique ont laissé d’abondants vestiges matériels à côté des tombes; mais quid de celles qui,parnature, nelaissent aucune trace archéologique: chants funéraires, jeux et autres formes decommémoration? La disparition desoffrandes peut impliquer undéplacement desmodes d’expression dustatut, pasnécessairement leur abolition radicale. Il est donc impératif d’élargir autant que possible l’horizon de l’analyse afin de faire entrer en ligne de compte tous les changements concomitants, aulieu de les considérer isolément dans chaque domaine: toute interprétation doit alors aider à unsecteur particulier. comprendre l’agencement del’ensemble, etnonl’évolution d’ Or,leschangements constatés danslespratiques funéraires coïncident généralement non pas tant avec un “essor”quantitatif des pratiques cultuelles, comme on l’a souvent écrit14, essor quisouvent les précède, qu’avec deprofondes mutations dans la conception et la nature même descultes. Ainsi, enArgolide, la deuxième moitié du VIIIe siècle est aussi la période où se répand une forme d’appropriation des tombes helladiques, soit parleurréemploi, soit pardesmanifestations devénération. Sans constituer unvéritable culte durable, cespratiques, attestées à Argos, Mycènes, Prosymna (Héraion), Dendra et Berbati, semblent revendiquer quelque lien privilégié avec le passé “héroïque”local à l’appui d’unstatut particulier: elles pourraient donc défi” lancé parles funérailles représenter, à unniveau plusmodeste, uneréponse au“ ostentatoires et héroïsantes des basileis d’Argos. A Prosymna, elles pourraient 172. 12 Morris, Burial andAncient Society (n. 5); Whitley, Style andSociety (n. 11), 162– 210. 13 Morris, o.c., 205– 27. 14 Ainsi queje l’avais fait moi-même dansLa naissance dela cité grecque, Paris 1984, 26–

12

François dePolignac

même constituer une tentative d’appropriation symbolique du site de l’Héraion, devenu le théatre et l’enjeu de la compétition ritualisée des élites de la plaine d’Argolide, poury défendre uneposition menacée parl’affirmation deplus enplus

dela prééminence argienne15. Ces pratiques néanmoins disparaissent vers la fin du siècle, peu après la tombe à panoplie”argienne, ce quitendrait à confirmer l’interaction entre dernière “ les deux phénomènes; et leur disparition coïncide avec la naissance de nouveaux cultes qui furent particulièrement florissants au VIIe siècle: celui de l’ Agamemnoneion”deMycènes, à unkilomètre ausuddela ville, près delaroute menant “ à l’Héraion, et celui du“sanctuaire secondaire”del’Héraion, surunepetite terrasse à moins d’unkilomètre aunord-ouest dusanctuaire principal, près de la route de Mycènes16. Ces deux cultes ont bien des points communs: s’il n’est pas certain qu’ils aient été descultes dehéros ausens strict, enrevanche il estindéniable qu’ils étaient associés d’une façon ou d’une autre au souvenir des anciennes dynasties locales17. L’unet l’autre comportaient desrepas rituels; leur apparition simultanée, leuressor parallèle àl’époque archaïque, etleurface à face topographique traduisent larivalité entre lesaristocraties argienne etmycénienne enconflit pourl’appropriation des mythes de souveraineté susceptibles de légitimer une hégémonie régionale (pour les Argiens) oule refus decette hégémonie (pour les Mycéniens): exemple de guerre mythologique”dontle conflit entre Argos et Sicyone autour desSept contre “ Thèbes fournit uneautre illustration environ unsiècle plus tard18. En Attique, l’apparition des Opferrinnen coïncide de la même façon avec l’essor de deux types de culte: les cultes à connotation funéraire et héroïque, représentés parplusieurs dépôts votifs trouvés surl’aire dela future agora et parles cultes pratiqués sur les tombes à tholos mycéniennes (Menidi, Thoricos); et les cultes desommet deZeus, auparavant attestés seulement surle Parnès et l’Hymette (depuis la fin du Xe siècle), qui se répandent alors sur de nombreux monts ou collines attiques, enparticulier danslamésogée del’A ttique oriental19. Laprospérité

évidente

15 Cesphénomènes ontrécemment suscité denombreuses études: J. Whitley, “Early States and 182; C.Antonaccio, “Terraces, Tombs andthe , JHS 108, 1988, 173– Herocults: a reappraisal” 105; idem, AnArchaeology ofAncestors: Hero , Hesperia 61, 1992, 85– Early Argive Heraion” , TheArchaeology ofAncestors” andTombCultinEarly Greece, Londres, 1993, chap. 2; idem, “ 70, en Cultural Poetics inArchaic Greece, éds. C. Dougherty, L. Kurke, Cambridge 1993, 46– 61; Polignac, “Cité et territoire”(n. 10); idem, Cults, Territory and the Rise of the part. 59– 162. 2, et La naissance dela cité grecque2 (1995), 161– Greek City-State, Chicago 1995, 141– , Mélanges A. Keramopoullou, Athènes The cult of Agamemnon at Mycenae” 16 J. M. Cook, “ , BSA48, 30 sq. C. W. 1952. TheAgamemnoneion” 118; idem, “Mycenae 1939– 1953, 112– 444.L’interprétation , AJA43, 1939, 410– Prosymna: Remains ofPost-Mycenaean Date” Blegen, “ , 101) decette petite terrasse comme “premier Terraces, Tombs” donnée parC. Antonaccio (“ d’Héranemeparaît pasdutout convaincante. sanctuaire” 17 La nature exacte deces cultes et deleurs destinataires reste incertaine, mais la mention (plus tardive) d’Agamemnon dansle sanctuaire deMycènes rendvraisemblable l’association duculte ausouvenir duroi, d’unefaçon oud’uneautre. Le monument argien des ‘Sept contre Thèbes’” 18 A. Pariente, “ , Polydipsion Argos. Argos dela fin despalais mycéniens à la constitution del’Etat classique, éd.M. Piérart, Athènes-Fribourg 229. 1992, 195– 19 K. M. Langdon, TheSanctuary of Zeus on Mount Hymettos (Hesperia Suppl. 16), Princeton

Repenser

la “cité” ?

13

decesdeuxformes deculte auVIIe siècle contraste vigoureusement avec l’effacement grands” apparent des“ cultes del’acropole d’Athènes durant la mêmepériode. Mais tandis queles premiers présentent denombreuses analogies formelles avec les rites funéraires aristocratiques, en particulier la destruction somptuaire par le feu des mêmes types d’offrandes, dont la poterie proto-attique quiprovient pourl’essentiel decesdeuxcontextes, les seconds endiffèrent radicalement à lafois parlaqualité et la typologie de la poterie: de moindre qualité (“ sub-géométrique” ) et moins ostentatoire, elle évoque plutôt le partage rituel de la boisson et de la nourriture frairies”rurales20. caractéristique dessimples “ Cesconcomitances sont d’autant plus frappantes qu’elles sereproduisent, mais ensens inverse, unpeuplus d’unsiècle plustard. L’extrême attention accordée aux changements de la fin de l’époque géométrique a en effet longtemps occulté les changements similaires, et tout aussi spectaculaires, qui se produisent dans les mêmes domaines vers la fin duVIIe ou le début duVIe siècle; ceux-ci n’ont été pleinement misenlumière quepardesétudes récentes21. EnArgolide ainsi, le culte du“sanctuaire secondaire”de l’Héraion semble avoir décliné vers la fin duVIIe siècle, alors mêmequedesoffrandes réapparaissaient dansles tombes argiennes; et les tombes à ciste, dont l’usage avait été délaissé enArgolide pendant unsiècle au profit de l’ensevelissement en pithos, réapparaissent également au début du VIe siècle. Le cas athénien estplus explicite encore. EnAttique eneffet, le parallélisme est frappant dans la première moitié duVIe siècle entre le déclin de l’usage des Opferrinnen et des rites qui leur étaient associés, celui de la plupart des cultes héroïques (en particulier le plus aristocratique de tous, celui de Menidi), et enfin celui descultes desommet22. Cesderniers permettent depréciser lesmodalités dece déclin: la dévitalisation progressive, puis l’abandon presque total, est en effet un trait commun à la plupart des cultes de sommet dans l’aire de la Grèce des cités (Attique, Corinthie, Argolide, Laconie, Cyclades) auVIe siècle, suite àune intégration plus poussée descantons périphériques auxterritoires et à l’organisation descités; 1976. H. Lauter, Der Kultplatz auf demTurkovouni (AM-Beiheft 12, 1985). H. Lauter, H. , Festschrift Werner Böser, Karlsruhe Lauter-Bufe, “ Einattisches Höhenheiligtum beiVarkiza” 305. J. Whitley, “ The Monument that stood before Marathon: Tomb Cult andHero 1986, 289– 230; F. de Polignac, “Sanctuaires et société en , AJA98, 1994, 213– Cult in Archaic Attica” , Culture et société: l’a vènement d’Athènes à l’époque Attique géométrique et archaïque” archaïque, A. Verbanck-Piérard, D. Viviers éds., Bruxelles 1995. 20 Houby-Nielsen, “Interactions”(n. 11), 354; J. Whitley, “Protoattic Pottery: a contextual approach” , Classical Greece: modern histories andmodern archaeologies, éd.I. Morris, Cambridge , Anthropologie de la 70. Surles frairies, l’article deL. Gernet, “Frairies antiques” 1994, 51– 61, reste essentiel. Grèce antique, Paris 1968, 21– , A Crisis in archaeological history? The Seventh Century in Attica” 21 L’article deR. Osborne, “ 322, avait attiré l’attention sur l’oubli du VIIe siècle, écartelé entre le BSA 84, 1989, 297– tropisme géométrique”desarchéologues et le confinement deshistoriens auVIe siècle; les “ études deS. Houby-Nielsen etJ. Whitley mentionnées ci-dessus comblent engrande partie cette lacune. 600 BC. Anarchaeological survey, Göteborg 1988, 48– 22 Argolide: A. Foley, TheArgolid 800– 363, 51. Attique: Whitley, “Marathon”(n. 19), 218; Houby-Nielsen, “Interactions”(n. 11), 362–

fig. 8.

14

François dePolignac

mais ce quidisparaît d’abord, et très rapidement, dansles cultes attiques, est ce qui constituait leur trait distinctif par rapport aux cultes péloponnésiens: l’usage de graffiti sur les vases, qui cesse dès les débuts duVIe siècle, avant même que la fréquentation deces hauts lieux nebaisse defaçon significative23. Detoute évidence, cesensembles demutations appellent uneinterprétation plus précise quele simple recours à l’idée toute faite etgénéralisante de“ formation dela . Plusieurs facteurs peuvent amener la modification oula cessation d’unrituel. cité” La disparition dugroupe social quis’identifiait à unecoutume enest un;la fin des tombes à panoplie argienne pourrait signifier ainsi la disparition (aumoins en tant quegroupe distinct) desbasileis d’Argos. Il n’estcependant passûrquel’explication par un changement de structure sociale soit la plus fiable: elle n’aide guère à comprendre comment certaines pratiques réapparaissent après unephase delatence, et aboutit à unevision très heurtée del’histoire sociale. Unautre facteur possible est l’exercice d’une pression qui contraint un groupe soit à accentuer les pratiques distinctives qui manifestent sa supériorité, soit au contraire à les abandonner. En Attique, le déclin desrites funéraires ostentatoires associés auxOpferrinnen s’insère ainsi pour l’essentiel dans la période troublée qui va de l’archontat de Solon à la y voir la conséquence des conflits qui tyrannie de Pisistrate. Il est donc tentant d’ opposaient les“ Eupatrides” auxautres Athéniens, etplusspécifiquement desréformes attribuées à Solon tendant àlimiter l’ampleur etle luxedesmanifestations funéraires24. Néanmoins, l’évolution despratiques funéraires fut progressive; auCéramique, le déclin del’usage destranchées à offrandes s’était mêmeamorcé avant l’intervention solonienne, vers la fin duVIIe siècle; et l’abandon d’uncertain type de pratique somptuaire ne signifiait pasle renoncement à toute forme d’expression d’unstatut particulier en contexte funéraire, comme le montre la multiplication des kouroi, korai et stèles figurées dans les nécropoles duVIe siècle25. Unemesure ponctuelle, politique, n’explique donc pastout: elle nesecomprend, enfait, quesi onla replace dans un processus plus large qui la déborde, l’explique, et qu’elle ne fait que renforcer. Un changement dans les rites distinctifs d’un groupe social peut en effet provenir aussi del’intérieur même dece groupe, dèslors qu’unenouvelle pratique semble mieux à même que la tradition précédente d’exprimer son identité et son unité symboliques. Ainsi, le rejet decoutumes funéraires excessivement ostentatoires a puse faire jour nonseulement chez ceux quienétaient detoute façon exclus, le démos, maisaussi etpeut-être mêmed’abord dansunefraction del’élite quis’y était reconnue maiss’entrouvait progressivement écartée parunecompétition somptuaire poussée deplusenplusloin. Unsystème dereconnaissance mutuelle d’appartenance à l’aristocratie reposant sur la destruction ritualisée de la richesse fonctionne correctement tant qu’uneréciprocité approximative est possible dans l’apport des offrandes et l’hospitalité aux divers repas rituels. Dès lors qu’une compétition 23 Polignac, “Sanctuaires et société”(n. 19). 8; 21, 6. 24 Plutarque, Solon 12, 7– 25 Houby-Nielsen, “Interactions”(n. 11), 363, fig. 8; A.-M. D’Onofrio, “Aspetti e problemi del 96. , AION 10, 1988, 83– monumento funerario arcaico”

Repenser

la “cité” ?

15

croissante engendre uneinflation demunificence, la truphè, l’équilibre del’échange est rompu et les membres de l’élite qui ne peuvent rendre autant qu’ils recoivent sont rejetés dans uneposition inférieure: la cohésion del’aristocratie, et dureste de la société avec elle, estalors menacée, etaffaiblie laconscience quecette aristocratie s’était formée de son identité collective26. Or, la fin du VIIe siècle fut de toute évidence unepériode d’instabilité etdedissension auseindel’aristocratie athénienne, comme le montrent la tentative tyrannique deCylon etlesrèglements decompte qui s’ensuivirent. La nature même des prescriptions soloniennes, en particulier la réduction dunombre desacrifices et l’interdiction desacrifier desbœ ufs–victime parexcellence desgrands sacrifices auxdieux –auxfunérailles, atteste queles rites funéraires dont les Opferrinnen ontgardé les vestiges avaient été pervertis parune compétition exacerbée etparl’hybris queSolon dénonça et quise manifestaient par des sacrifices disproportionnés, unehospitalité “princière”aux repas funéraires et l’exhibition ostentatoire des offrandes de prix prêtes à être détruites; elles apparaissaient désormais comme un facteur de division et d’exclusion au sein de l’élite athénienne, et parsuite d’instabilité detoute la société. Le plaidoyer deSolon, lui même eupatride, en faveur del’e unomia avait donc toute chance derencontrer les préoccupations d’unefraction del’aristocratie marginalisée parl’évolution des rapports depouvoir entre les grandes familles athéniennes27. Lareprésentation sociale, autrement ditlarecherche del’expression symbolique correspondant le mieux à l’image qu’un groupe entend donner de lui-même, est donc à mêmedemodifier lescomportements rituels sans qu’intervienne nécessairementunepression extérieure. C’estdanscette perspective quel’onpeutcomprendre la relation entre l’évolution despratiques funéraires et celle despratiques cultuelles auVIesiècle. Ledéclin progressif d’unculte surtombe mycénienne comme celui de Menidi n’est évidemment qu’unindice supplémentaire de la crise dusystème de représentation “ héroïsante”del’élite attique dont il faisait partie aumêmetitre que les rites funéraires. La dévitalisation des cultes de sommet qui, dispersés dans la mésogée, réunissaient sans doute les populations desdèmes ruraux avoisinants, en estl’exacte contrepartie. Lamodestie générale desoffrandes enexclut laparticipation aristocratique, mais l’usage des graffiti sur vases montre que ces cultes étaient fréquentés au VIIe siècle par l’ élite”de la société paysanne attique, ces agroikoi laclasse hoplitique athénienne28. Ladisparition des aisés quiconstituaient le grosde“ graffiti audébut duVIe siècle coïncide donc avec l’irruption massive dece groupe social dans l’espace politique athénien et avec les réformes de Solon qui, prenant acte de cette irruption, organisaient le mode de participation des hoplites à la vie publique; il est significatif quel’usage privé, cultuel et dispersé del’écriture aitpris fin aumoment où la rédaction des “lois”soloniennes sur les kurbeis déposées au

26 P. Schmitt Pantel, La cité aubanquet. Histoire desrepas publics dans les cités grecques, Rome, 59; K. Fagerström, “Wealth Destruction as a Sign of Iron Age Ecole française, 1992, 57– 57; E. Scheid-Tissinier, Lesusages Political Strife” , Current Swedish Archaeology, 1, 1993, 49– dudonchez Homère, Nancy 1994. 27 Cette analyse rejoint celle de M. Stahl, Aristokraten und Tyrannen im archaischen Athen, Stuttgart 1987, 73. 28 Polignac, “Sanctuaires et société”(n. 19).

François dePolignac

16

prytanée conférait à l’écriture, aucentre dela cité, le statut public quigarantissait le nouvel agencement politique29. Unautre aspect deces réformes pourrait avoir joué unrôle dans le déclin de ces cultes locaux. Onattribue en effet à Solon diverses prescriptions règlementant laparticipation auxrepas communs, lasitesis endemosioi, et élargissant le système de commensalité déléguée qui permettait aux citoyens vivants dans les dèmes d’être tour à tourparasitoi dans les repas publics se tenant à Athènes même, y compris, naturellement, les repas desfêtes religieuses, auxquels toute la population rurale nepouvait affluer30. Enfavorisant uneconception élargie etplusabstraite delacommensalité, cesmesures établissaient lesrègles quifaisaient d’Athènes le centre effectif delaviesociale, cultuelle etpolitique detoute l’A ttique: elles ôtaient donc deleur attrait auxrassemblements locaux quiavaient jusqu’alors constitué la forme principale deviepublique d’unegrande partie des“ citoyens”des dèmes. A tous points de vue donc, le déclin des cultes de sommet paraît refléter l’accès dela paysannerie hoplitique à denouvelles formes d’expression publique, à uneconscience plus directement civique desonidentité. Ce type d’analyse éclaire également les mutations antérieures de la fin de l’époque géométrique. Les tombes argiennes à pièces d’armement et obeloi traduisent clairement la revendication d’unefraction del’aristocratie à unstatut exceptionnel. La dernière (vers 720/710) et la plus connue, la “tombe à la panoplie” , est la plus complète et la plus significative; les objets qui y furent déposés correspondent à trois activités biendéterminées: l’armure pourlaguerre, lavaisselle pourle banquet, les broches et les chenêts pour griller la viande, symbole du sacrifice31. Cette trilogie évoque les trois cadeaux quelesjeunes nobles crétois, à l’époque classique, recevaient de leur éraste au moment où ils accédaient à la citoyenneté après leur initiation: la panoplie, la coupe, le bœ ufpourle sacrifice32. Dans l’unet l’autre cas, la guerre, le banquet et le sacrifice sont donc les trois composantes fondamentales dustatut aristocratique. Mais enCrète, comme dans les autres cités aristocratiques oùcette coutume est attestée, ces trois dimensions dela vie sociale constituaient un privilège partagé par tout le groupe des citoyens. En Argolide au contraire, elles paraissent constituer une distinction éminemment personnelle, puisqu’exprimée dans le rite funéraire33. L’exclusivité de la relation ainsi affichée au sacrifice (et à

29

Solon et la voix de Plutarque, Solon, 25,1. Le rapport à l’écriture est analysé parN.Loraux, “ , Les savoirs de l’écriture. En Grèce ancienne, s. dir. deM. Detienne, Lille 1988, 95– l’écrit”

129.

99. 30 Schmitt Pantel, La cité aubanquet (n. 26), 97– 386. La signification des , BCH 81, 1957, 322– 31 P. Courbin, “Tombe géométrique d’Argos” Obeloi of Pre- or obeloi placés dans les tombes a fait l’objet de nombreux débats; I. Strøm, “ , Economics of Cult in the Ancient Greek Proto-monetary Value in the Greek Sanctuaries” 51, en donne une synthèse World (Boreas 21), éds. T. Linders, B. Alroth, Uppsala 1992, 41– Grecs récente. C’estégalement unesignification sacrificielle queleurattribue B. D’Agostino, “ , L’archéologie aujourd’hui, éd. A. et indigènes sur la côte tyrrhénienne au VIIe siècle” 226. Schnapp, Paris 1981, 207– d. Schmitt Pantel, La cité aubanquet (n. 32 Ephore, ap.Strabon X, 4, 16sq.; Athénée, IV, 143a– 81. 26), 78– 33 L’âgedudéfunt reste cependant uneinconnue, et l’absence dedonnées surlacorrélation entre l’âge et le type desymbolisme funéraire limite les possibilités d’interprétation: les offrandes . peuvent représenter unstatut réel ou“ virtuel”

Repenser

la“cité” ?

17

l’hospitalité) pourrait avoir été la raison mêmedesadisparition quand se concrétisa la nécessité ou la possibilité d’uncontrôle symbolique élargi et plus efficace des cultes dela région. La destruction d’Asiné vers 710 avait fait disparaître le principal obstacle à la réalisation del’hégémonie régionale argienne. Le recours à la guerre et aux symboles guerriers devenait dès lors moins essentiel que le contrôle des relations régionales qui se nouaient autour du sanctuaire central, l’Héraion de Prosymna. Mais l’appropriation symbolique dusanctuaire semble alors passer d’un niveau individuel, celui des basileis du VIIIe siècle dont l’influence se bâtissait et s’exprimait par les riches sacrifices, l’hospitalité somptueuse et les offrandes de prix, à unniveau collectif: ladisparition desbroches dessépultures (etleur apparition fréquente dans les sanctuaires où,mêmeoffertes parunindividu, elles font office de bien commun)34, et l’essor descultes aristocratiques dela “terrasse secondaire”de Prosymna et de l’A gamemnoneion de Mycènes montrent que les sacrifices, les repas et banquets cultuels étaient désormais les pratiques constitutives de groupes réunissant une fraction plus ou moins grande, sinon la totalité de l’élite de la cité, groupes dont l’image est donnée par les associations cultuelles des platioinoi d’Athéna et d’Héraclès, connues auVIIe siècle dansla cité voisine deTirynthe35. La distinction principale n’était plus entre un basileus dominant, revendiquant une autorité particulière sur les activités rituelles, et le reste de l’élite, mais entre les membres deces associations oùla commensalité impliquait unerelative égalité de statut et demoyens, et les autres. Ces phénomènes divers sont donc en fait autant de signes de la constitution d’unearistocratie ausens premier duterme, et cela nous ramène auproblème dela “ cité” . L’aristocratie identifiée parses rites spécifiques est aussi celle despremiers citoyens qui forment le conseil et se partagent les archai. Les cultes qui lui sont particuliers sont aussi d’importants marqueurs dansle processus deterritorialisation en cours à la fin de l’époque géométrique et pendant le haut archaïsme: au VIIIe siècle, le “théatre social”des pratiques rituelles, qu’il s’agisse des funérailles princières d’Argos ou des sacrifices à l’Héraion, était l’Argéia (plaine d’Argos) entière, tandis que les deux sanctuaires à connotation héroïque de Prosymna et de Mycènes balisent sinon desfrontières territoriales stricto sensu, dumoins leslimites de l’influence des aristocraties respectivement argienne et mycénienne. L’unité culturelle de l’Argolide n’est pas totalement brisée, mais deux centres politiques distincts s’y disputent visiblement l’appropriation desgrands mythes desouveraineté et descultes unificateurs dela région. Si l’onpeut hésiter à appeler formation de la cité ce processus complexe deréorganisation politique et sociale quiredéfinit tout à la fois le champ dukoinon, la détermination de l’identité collective et la façon d’y exercer l’autorité sous sesdiverses formes, c’estessentiellement pournepastomber dans le piège des mots: en tant que bourgades distinctes, dotées de leurs propres dirigeants etdeleurs propres cultes, Mycènes, Tirynthe, Nauplie ouArgos pouvaient 34 I. Strøm, o.c. (n.31). , AE 35 N. Verdelis, M. Jameson, P. Papachristopoulou, “Archaïkai Epigraphai en Tirynthos” 205; Sophocle évoque aussi un“festin d’Agamemnon”qui passait pour avoir été 1975, 150– 285). fondé parClytemnestre à Mycènes (Electre, 277–

18

François

dePolignac

eneffet êtrequalifiées depoleis àl’époque archaïque comme àl’époque géométrique. Néanmoins, leur passage à un degré supérieur d’intégration régionale, avec les résistances et fractures que cette évolution suscite au fur et à mesure qu’elle se consolide, illustre bien la façon dont se constituèrent les grandes cités archaïques, même si le processus d’unification nefutpaspoussé aussi loin enArgolide qu’il le futenLaconie ouenCorinthie. Athènes, danscette perspective, illustre nonpasun“ échec dela cité” , mais une autre variante dece processus. Le retour del’élite athénienne duhaut archaïsme à des pratiques funéraires exclusives, retour facilité parl’ancienneté des structures sociales et politiques attiques, nepeut être assimilé trop rapidement à unretour en arrière empêchant la formation de la “ cité” . Il s’explique en effet par le fait que l’unité politique del’A ttique était sans doute effective depuis longtemps, et quela constitution d’une grande entité territoriale ne nécessitait pas, comme dans les autres régions, l’appropriation symbolique par l’aristocratie d’un culte commun articulant l’ensemble desrelations sociales. Cependant, mêmeorienté vers le monde des morts plustôt que vers le monde des dieux comme dans la plupart des autres cités, le rite est là aussi l’élément unificateur de l’aristocratie athénienne des Eupatrides dont l’identité collective est définie parréférence auxhéros locaux. Le rapprochement sciemment opéré entre lespratiques funéraires etle monde deshéros dont les tombes, objets d’unculte, jalonnent le territoire, pourrait être la première traduction de l’idée d’autochtonie en tant que mythe de légitimation politique, et représente donc unchangement dans l’idéologie aristocratique pour lequel il n’est pasabsolument nécessaire depostuler unconflit ouvert avec le reste dela société36. D’un autre côté, le rôle très particulier des cultes de sommet témoigne de l’élaboration parallèle d’une identité rituelle par l’ élite”des non nobles, exclus de la athénienne provient donc de formalisation des pratiques funéraires. La spécificité “ l’absence depoint derencontre entre ces deux modes deconstruction del’identité collective jusqu’auVIesiècle: Athènes représente nonpasl’absence ouladéfaite de lacité, maisuneévolution delacité à partir d’unautre typed’articulation entre unité

et unité culturelle d’unerégion. Ces observations apportent quelques éléments de réponse à la question posée surla pertinence del’oubli dela “cité” . Si l’on seréfère à unmodèle canonique dela polis en fonction duquel les sociétés sont classées en “ non-cités”ou“ cités cités”et “ inachevées”sans que soit clarifié le rapport entre la cité-institution et la cité-société, oule temps divisé entre un“avant lapolis”et le temps delapolis aumoyen d’une ligne de démarcation à la datation aléatoire et discutable, il vaut mieux en effet oublier la cité pour penser la société. En revanche, l’étude des pratiques rituelles montre queleschangements quiles affectèrent à la findel’époque géométrique et à l’époque archaïque mettaient enjeuunereformulation desidentités collectives oùle réaménagement desformes del’autorité était associé à laréorganisation desrelations sociales dansunespace défini; le nouvel agencement quienrésultait était symbolisé par des rites qui manifestaient tout à la fois la nature et l’étendue des pratiques communes et le type de pouvoir qui s’y exerçait. La diversité des agencements et politique

29. 36 Houby-Nielsen, “Interactions”(n. 20), 344; Morris, Death ritual, 27–

Repenser

la “cité” ?

19

desrites quiles matérialisaient n’estpasle signe d’unprogrès plus oumoins rapide surla voie d’uneévolution linéaire: elle reflète plutôt ladiversité descontextes dans lequel ce processus se déroula. Assurément, les phénomènes étudiés concernent davantage les grandes cités quelespetites communautés à l’évolution plus simple et plus graduelle, encore qu’uneétude attentive pourrait montrer qu’ils se sont également produits à petite échelle37; mais ce sont précisément les grandes cités qui, par les problèmes qu’elles affrontaient, peuvent nourrir la réflexion surla constitution des sociétés. Les mutations dela fin de l’époque géométrique signalent donc bien l’émergence de nouvelles conceptions de la communauté caractérisées par l’intégration de plus en plus poussée des différentes composantes de la vie collective, sans laquelle lapolis ne serait pasdevenue lapoliteia auxmultiples facettes. François dePolignac Centre Louis Gernet, EHESS/CNRS, Paris

37 Les problèmes d’intégration et de hiérarchisation analysés auplan régional pour les grandes cités ontpueneffet sereproduire auplanlocal, lors del’émergence depetites cités auseind’un groupe dekomai autonomes.

THE »AUTONOMOUS CITY-STATE«. ANCIENT FACT OR MODERN FICTION? by MOGENS HERMAN HANSEN

I. The Orthodox View

In their

descriptions of the ancient Greek city-state many modern historians take autonomy (autonomia) to be a defining characteristic of thepolis in the archaic and classical periods. One example is the editors’preface to The Greek City-State from Homer toAlexander: »Our focus hasbeen the autonomous Greek city-state orpolis from its origins in the ‘Dark Age’until the point at which it wastransformed into a basis for world civilization by the conquests of Alexander the Great.«1 Andin the grande finale of that valuable volume W.G. Runciman takes autonomy to be the most important defining characteristic of thepolis: »But what is a polis? ... First, a

1

O.Murray & S. Price (eds.), TheGreek City-State FromHomer toAlexander (Oxford 1990) vii. See also: J. Gaudemet, Institutions del’a ntiquité (Paris 1967) 147: »Ainsi entendue la cité doit jouir d’une double indépendance. L’une, économique (»autarcie«) ... L’autre, politique (»autonomie«). L’indépendance vis-à-vis d’autres cités oud’autres peuples, constitue l’essentiel. TheAncient Là oùil n’y a plus d’indépendance, il n’y a plus vraiment decité.«; M.I. Finley, “ CityfromFustel deCoulanges toMaxWeber andBeyond,”inEconomy andSociety inAncient Greece (London 1989) 4– 5: »Aristotle ... waswriting about the autonomous city-state, thepolis

inGreek ... (5) ... Theancient city wassoontolose itsautonomy. Theprocess began soonafter

The Greek Polis,”in R. Griffeth & C.G. Thomas (eds.), The Aristotle died.«; C.G. Thomas, “ City-State inFive Cultures (Santa Barbara 1981) 40: »Itwaswhenallpretense of autonomy and independence wasabandoned that thepolis became a city retaining only its vestigal form and notits substance.«; cf. xiii, xv;W.Gawantka, Diesogenannte Polis (Stuttgart 1985) 9 n. 1: »Da einun-autonomer Stadtstaat evident einWiderspruch insich wäre ...« L. Bruit Zaidman & P. Schmitt Pantel, Religion in theAncient Greek City, English edn. translated by P. Cartledge (Cambridge 1992) 7: »a“ isanautonomous andindependent political unit.«; G.Clemente, city” Concluding Reflections,”in A. Molho, K. Raaflaub, J. Emlen (eds.), City-States in Classical “ 3: »The most generic, but most efficient Antiquity and Medieval Italy (Stuttgart 1991) 642– definition of a city-state would seem to be this: a political entity that wasautonomous, and therefore capable of controlling a decision-making process.«; cf. also L.H. Jeffery, Archaic Greece (London 1976) 39; K.-W. Welwei, Die griechische Polis (Stuttgart 1983) 10; H. van 5; H. Bruhns, “ La cité antique deMaxWeber,” Effenterre, La cité grecque (Paris 1985) 24– 9) 323 (Abstract); G. Forrest in J. Boardman, J. Griffin & O. Murray (eds.), 8 (1987– Opus 6– The Oxford History of the Classical World (Oxford 1986) 19; P. Green, From Alexander to Actium (Berkeley andLos Angeles 1990) 53: »the collapse of the city-state«, »the loss of 8 etc. –For the explicit or implicit idenpolitical autonomy«, cf. 23, 56, 80, 155, 164, 196– tification of autonomia with autonomy andthus withtheconcept of thepolis see infra note 9.

22

Mogens Herman Hansen

polis must be juridically autonomous in the sense of holding a monopoly of the means of coercion within the territory to which its laws apply ... Thepoleis which survived andindeed flourished in the Hellenistic andeven Roman periods were, therefore, poleis in name only: they were urban communities with a life of their own, butnot“citizen-states”in the sociological sense.«2 Especially among Anglophone scholars the link between the concept of autonomia andthe concept ofpolis is so widespread that it can reasonably be called the orthodox view. But it is not universal. Some scholars have expressed their reservations.3 I will come back to that later, after discussing what still seems to be the orthodox view. ςπ μ Thepolis is bydefinition autonomos; butwhatis anα ο ό λ τό ο ις? A ready ὐ ν answer is either torender thetermby»the independent city (orcity-state)« ortouse ςand speak of the »autonomous city-state«, or μ ο ν ο the English equivalent of α τό ὐ to combine the twoterms anddefine thepolis as anautonomous andindependent political unit. Both answers, however, areproblematical, andoneis notbetter off by combining them. (1) To define thepolis as anindependent political unit4 runs counter to thefact that manypoleis were dependencies andthat, bytheearly 4thcentury B.C., even a majority ofpoleis hadlost their independence either bybeing dominated byoneof the hegemonic cities or the King of Persia or by becoming a member state of a confederation. I return to this problem in section III infra. (2) To define the city as an autonomous political unit conceals an essential difference between contemporary English and ancient Greek terminology: the English words »autonomy« and»autonomous« areambiguous in meaning andcan be used to denote anything from the sovereignty of states to the self-government exercised byconstituent states orprovinces oreven local communities. The Oxford Advanced Learner’s Dictionary of Current English illustrates the meaning of the adjective bythephrase: »analliance of autonomous states«.5 Here »autonomous« is

2 3

4

5

W.G. Runciman, “Doomed toextinction: ThePolis asanEvolutionary Dead-End,”inMurray & Price (supra n. 1) 348. 7; E. Lévy, “ La cité grecque: E.g. F. Hampl, “Poleis ohne Territorium,”Klio 32 (1939) 16– Die invention moderne ouréalité antique?”Cahiers duCentre Glotz I (1990) 55; D. Lotze, “ 92) 239; K. Raaflaub, “Homer unddieGeschichte sogenannte Polis,” Acta Antiqua 33 (1990– des8 Jh.s v. Chr,”inJ. Latacz (ed.), Zweihundert Jahre Homerforschung (Leipzig 1991) 241 n. 122; idem, “Homer to Solon”inM.H. Hansen (ed.), TheAncient Greek City-State. Acts of the 20. Copenhagen Polis Centre 1 (Copenhagen 1993) 44. Cf. also myownaccount ibidem 18– In addition to the literature cited supra n. 1 add: A. Aymard, Les cités grecques a l’époque 4; The Colonial “ J. Graham, “ classique,”Recueils de la Société Jean Bodin 6 (1954) 52– Interaction byDesign: theGreek Expansion of Greece,”inCAHIII.3 (1983) 83; A. Snodgrass, “ City-State,”in C. Renfrew & J.F. Cherry (eds.), Peer Polity Interaction and Socio-Political Change (Cambridge 1986) 47; cf., however, the modification added in idem, Archaic Greece

(London 1980) 28;O.Murray, Early Greece (2ndedn.London 1993) 62;R. Osborne, Classical city”is usedto refer to the Landscape with Figures (London 1987) 11: »Throughout thebook “ independent political unitof townandterritory«; Oxford English Dictionary 3 (2ndedn. 1989) 254: »City-state, a city which is also anindependent sovereign state.« 4th edn. (1989) 68. Both meanings arerecorded in Collins Dictionary of theEnglish Language (2nd edn. 1986) s.v. autonomous: (1) possessing a large degree of self-government; (2)

The »Autonomous City-State«

23

indisputably usedsynonymously withtheadjectives »independent« or»sovereign«; but the negotiations between Israel andthe Palestinians in 1993 ended with a treaty by which theGaza strip andJericho became »autonomous«; andhere the adjective is used in the sense of »self-governing«.6 The ambiguity of the term autonomy is best illustrated by quoting the description of it in The Blackwell Encyclopaedia of Political Thought: »Literally meaning ‘self-rule’, autonomy is ascribed in popular

political parlance to self-governing states, or to institutions or groups within states thatenjoy a substantial degree ofindependence andinitiative.«7 TheGreek adjective ς, on the other hand, means »living under one’s ownlaws«8 but in the μ ο ο ν α τό ὐ sense of being »independent« rather thanjust »self-governing«; moreover, it was never applied to »institutions or groups within states« only to what wetoday call ία μ hadoriginally a muchnarrower meaning than states, see section II infra. Α ο ν το ὐ autonomy hastoday, andto speak indiscriminately of the autonomos polis andthe ςπ ό λ ις? Thus inthis ο μ ο ν τό ὐ autonomous city-state obscures theissue: whatis anα article I prefer to avoid the English words autonomous andautonomy, andto use autonomos andautonomia as simple transliterations of the Greek terms. (3) To define thepolis asanindependent andautonomous political unitclarifies the ambiguity in so far as the phrase »independent andautonomous« is taken to ς«. In that case μ ο ν ο τό ὐ meanwhatwecall »independent« andtheGreeks called »α supra, it does not make (a) the English term tallies with the Greek one but, as in political ) ς unit ο if, inthe6th μ ο ν τό ὐ sense to define thepolis asanindependent (orα and5thcenturies, many, andinthe4thcentury, evenmostpoleis, weredependencies.

Now, historians who define the polis as an independent and autonomous (or autonomos) political unit are, of course, aware of the fact that quite a few poleis were dependencies. What dothey doto avoid thecontradiction interms? Somejust leave it there butothers maintain that the archaic andclassical polis wasideally an

6

7 8

independent. The same ambiguity applies to the German andFrench terms as well. For the German terms Autonomie/autonom see, e.g., Duden. Das große Wörterbuch der deutschen Sprache 1 (Mannheim 1976) s.v. Autonomie: »[Verwaltungsmäßige] Unabhängigkeit, Selbständigkeit: dieAutonomie dieses Landes ist gefährdet; DieAutonomie derFraktionen stellt daspolitisch bedeutsamste Gegengewicht gegen autokratische Tendenzen dar.« FortheFrench terms autonomie/autonome see, e.g., Lepetit Laroussse compact (1993) s.v.: »Indépendence, possibilité dedécider, pourunorganisme, pourunindividu, parrapport à unpouvoir central, à unehierarchie, uneautorité. L’autonomie desuniversités« compared withDictionnaire Hachette encyclopédique 1994 s.v. »Indépendence quijouissent lespays autonomes.«. The agreement itself is called »Declaration of Principles of Interim Self-Government Arrangements«, andinthedocument thetermself-government is theonly termused, butinthe subsequent political debate »autonomy« hasfrequently been used synonymously with »SelfGovernment«. See United Nations General Assembly Security Council A/48/486 S/26560 of TheMiddle 11 October 1993 compared withtheIsraeli Government’s report of 2 June 1994: “ 8, where Gaza and Jericho are referred to as East Peace Process, an Overview”pages 7– »autonomous areas«. D. Miller (ed.), TheBlackwell Encyclopaedia of Political Thought (Oxford 1987) 31. ς μ ο ό ν το ὐ ν(the Samians) α τῶ ὐ ιςα ο ετέρ θ ιτ α ο ιςχρῆσ ὲνόμ ο ῖςσφ 6: τ ο ῖςδ IG I3 127.15– ς. Seeinfra page 26 withnote 24. τ α ν ὄ

24

Mogens Herman Hansen

autonomous political unit. Oneexample is E. Will: »la pleine indépendance d’une polis suppose la pleine possession de son territoire: or l’idéal de souveraine indépendance (l’autonomie dans la liberté) est indéracinable de la pensée politique grecque.«9 Here the definition of thepolis is moved from the empirical world of political realities tothenormative world of political ideology. Asa result there is no longer anyclash between theconcept of autonomia (which excludes dependencies) andactual poleis (many of which were dependencies). Whatthe definition implies isjust that allpoleis wanted tobeautonomoi andthat dependencies didtheir best to obtain autonomia. Butto be true this definition presupposes that wehave a number of archaic andclassical sources inwhich autonomia asanideal is matched withthe concept of thepolis. Such sources, however, cannot be traced futher back than ca.

375 B.C., see infra pages 38– 40. Furthermore, the view that every polis strove for autonomia is universally asserted butnever substantiated andnotsupported byoursources. Quite a fewof the members of the Delian League, for example, seem to have been content with their status as hypekooi poleis anddidnot avail themselves of the possibility of becoming autonomoi, not even during the last years of the Peloponnesian War.10 Similarly several of the perioikic communities in Messenia (explicitly called poleis in our sources) seem to have supported Sparta even after thebattle of Leuktra in 371.11 So a definition of thearchaic andclassical polis which focuses onautonomy as a universal ideal andimplicitly identifies autonomy withautonomia is without any foundation in the sources whenapplied to thearchaic andearly classical periods. It canmake sense only if applied to thepolis of thelate fifth andfourth centuries. Other historians again seek refuge in the meaning of the concept of autonomia andclaim that it covers notjust independence but also self-government in a much more restricted sense. Such a stand takes usback to the problem mentioned under (2) supra: the relation between α ία μ andautonomy. Without anydiscussion ν ὐ ο το of therelation between theancient andthemodern concept thetwoterms aretreated μ ία is asbroad andvague astheEnglish ο ν το ὐ as synonyms andit is assumed that α

9

10 11

Le monde grec et l’Orient I (Paris 1972) 416. Cf. F. Gschnitzer, “Autonomie,”Lexikon der Alten Welt (Zürich & Stuttgart 1965) 419: »Die Autonomie (»Selbständigkeit«) aller griech. Städte wurde etwa seit derMitte des5. Jh. dasvielberufene, aber nicht erreichte Ideal griech. Politik«.; H.Bengtson, Griechische Geschichte (5th. edn.München 1977) 286: »... doch hatdie Autonomie, dasPalladium dergriechischen Gemeindestaaten, dieEntstehung größerer Machtbildungen vonvornherein sehr erschwert, ja vielfach geradezu unmöglich gemacht«, cf. 149; V. Ehrenberg, “Autonomy,”OCD(2ndedn. Oxford 1972) 155: »one of theleading ideas of a Greek polis«; A. Mannzmann, “ ία μ ,”Der Kleine Pauly 3 (1964) 782: »Selbstgesetzο ν το Α ὐ gebungsmacht, die auf der genossenschaftlichen Grundlage der Polis beruht.« R. Brock, polis,”in G. Speake (ed.), A Dictionary of Ancient History (Oxford 1994) 507: »freedom to “ manage its ownaffairs byits ownlaws (autonomia) wastheideal of every polis.« 55; for a See F.A. Lepper, “Some Rubrics in the Athenian Quota-Lists,”JHS 82 (1962) 25– Forschungsbericht see P.J. Rhodes, TheAthenian Empire (G&R. NewSurveys in the Classics 45. 17, Oxford 1985) 36– 9. Cf. N.Demand, C. Roebuck, AHistory of Messenia from369 to 146B.C. (Chicago 1941) 38– Urban Relocation inArchaic andClassical Greece (Bristol 1990) 111 with note 31.

The»Autonomous City-State«

25

concept of autonomy;12 consequently itcancover dependencies aswell asindependent political units. Oneexample is theLakonian perioikic communities in Lakonia and Messenia. All sources agree that they werepoleis.13Again all sources agree that they were dependencies.14 Nosource claims that they were autonomoi. Onthe contrary the notorious exchange of words between Agesilaos andEpaminondas suggests, rather, that they were not autonomoi.15 But historians who hold that poleis were by definition autonomous tend to assert that the perioikic communities were »autonomous« or, more cautiously, »internally largely autonomous« or »in some sense politically autonomous«.16 This is inmyopinion totamper withtheconcept of autonomia in order to uphold anunwarranted connection between thetwoconcepts of autonomia andpolis.

II. The Concept of Autonomia

1.Autonomia Denotes Independence, notjust Self-Government Further analysis, then, of the relation between the two concepts of polis and μ ὐ ία το ν ο autonomia requires an answer to the question: what is α ?17 Since Bickermann’s andOstwald’s careful studies there is consensus that »in classical ευ ία θ ερ . Its proper context was the μ ία wasno mere synonym for ἐλ usage α ὐ το ν ο relationship between states, specifically between a stronger oneanda weaker, andit

/ 12 Theunexplained switch between autonomy/autonomous, autonomia/autonomos andα ὐ το μ ία ν ο ςoccurs inall thearticles inEnglish quoted infra n. 17. μ ο α ὐ τό ν ο 13 Hdt. 7.234.2; Thuc. 5.54.1 (pace the note in Gomme/Andrewes/Dover); Xen. Hell. 6.5.21; Ages. 2.24; Lac. Pol. 15.3; Skylax 46; Isoc. 12.179; Strabo 8.4.11; Paus. 3.6.2; Polemon Π ρ ὶ ε τῶ νἐ νΛ α κ εδ ν . See S. Isager andJ.E. Skydsgaard, Ancient Greek Agriculture α ω ε λ ό ιπ ν ο ίμ (London 1992) 131; P. Rhodes, “ The Greek Polis: Demes, Cities andLeagues,”in Hansen (supra n. 3) 163– 4. 81; Thuc. 4.53. 14 E.g. Isoc. 12.178– η ς, ε σ ίλ α ο ἰκ γ α τ ρ οἈ ὰ ε τ μ π ό ἤ ν λ ύ ιν ὀ να ςἐά ιΒοιω ν α δ το ν ώ μ ιν 15 Paus. 9.13.2: Ἐ σ ὺ ο υ α π σ ιν ”ε ε ρ ἶπ ρ ό νγ ς, “ ό τε νὦ ε η ὑ ο ὐπ ρτῆ π ή ν ςεἰρ ὲ α ὶ το ρ τιᾶ τα π ρ ςπ α ὶνἢκ ὺ Σ ι, π ερ ιο ίκ ς ο υ ς.”This exchange of views took place in υ ο ετέρ ςὑμ ντο ύ μ ε μ ςκ ν ύ ο ν τ α ὀ α τ ὰπ ό ινἴδω λ “ connection with theCommon Peace at Sparta inthespring of 371 andthemainclause of that ςἐᾶ peace wasτο υ μ ο ν(Xen. Hell. 6.3.18), see P. Cartledge, Agesilaos and ό ν το ὐ ςπ ό λ ὺ ε ιςα the crisis of Sparta (London 1987) 380. That not even the Thebans believed that the other 3, 36. Boiotian poleis were autonomoi is apparent from Xen. Hell. 4.8.15; 5.1.32– 192 B.C. (London 1968) 30; A.J. Toynbee, SomeProblems 16 G. Forrest, AHistory of Sparta 950– of Greek History (Oxford 1969) 204; P. Cartledge, Sparta andLakonia (London 1979) 178; P.J. Rhodes, “ TheGreek Polis: Demes Cities andLeagues,”inHansen (supra n. 3) 163. 44; 17 E.J. Bickerman, “Autonomia: surunpassage deThucydide (1.144.2),”RIDA5 (1958) 313– M. Ostwald, Autonomia: its Genesis andEarly History (NewYork 1982); E. Lévy, “Autonomia 70; K. Raaflaub, Die Entdeckung der Freiheit et éleuthéria auVe siècle,”RPhil 57 (1983) 249– (München 1985) 193– 207; B. Bosworth, “Autonomia: the Use and Abuse of Political 52; D.Whitehead, “Samian Autonomy,”inR.M. Rosen & J. Terminology,”StIt 10 (1992) 122– Farrell (eds.), Nomodeiktes: Greek Studies inhonor ofMartin Ostwald (Ann Arbor 1993) 321– 9; M. Ostwald, “Stasis andAutonomia in Samos: A Comment on an Ideological Fallacy,” 66. Scripta Classica Israelica 12 (1993) 51–

26

Mogens Herman Hansen

wasa wayof describing (and hence protecting) theposition of thelatter inrelation to the former«.18 We must not forget, however, that in quite a few passages the context is theinternal political life of a community andthatthenegative is tyranny.19 But what does it mean? Here the views are divided andI prefer to side with those whobelieve that autonomia originally meant »self-government«, notnecessarily in the negative andrestricted sense of »self-government which is willing to accept subordination to a superior power insomematters«20 butrather inthepositive sense of an often unqualified independence in one’s own affairs.21 If the concept of autonomia hadbeenloaded withallthenegative aspects assumed byBickerman and Ostwald, it would never have become a popular political slogan invoked with increasing frequency in the course of the late 5th and4th centuries.22 In modern text-books of international law the independence of a state is described as: (a) the power exclusively to control its owndomestic affairs, (b) the power to admit and expel aliens, (c) the privileges of its envoys in other communities and(d) the sole jurisdiction over crimes committed within its territory.23 These are modern requirements, andin many respects modern states are very different from ancient poleis, but the four points listed above square fairly well with what autonomia implied in the ancient Greek world: to give oneself one’s ownlaws24 including the right to decide about taxes,25 to control one’s own territory,26 to have the sole 18 Whitehead (supra n. 17) 321. 19 See Bosworth (supra note 17) 123andalltheexamples cited anddiscussed infra page 19. 20 Ostwald (supra n. 17 [1982]) 29; cf. Bickerman (supra n. 17) 327: »La subordination est autonomie«, cf. 328, 330, 337. toujours présente surl’arrière-plan mental del’idée d’ 21 Lévy (supra n. 17) 256: »cette conception del’a utonomie [asa kindof independence inferior to eleutheria] est contradite par les nombreux passages oùle mótdesigne sans équivoque une indépendance pleine etentière« 260: l’autonomie fait bienallusion à ladépendence, maispour 200: »Wer eleutheria sagt, blickt nach außen und la nier«. Raaflaub (supra n. 17) 199– bezeichnet das Fehlen oder die Abwehr vonHerrschaftsunterworfenheit ... Werautonomia verwendet, blickt nach innen und hebt die Tatsache hervor, daß das Subjekt sich selber bestimmt. Bei beiden Begriffen gehtes umdenGegensatz vonSelbstbestimmung undFremdbestimmung. Deshalb stehen siesichoftsehrnahe. Aberautonomia betont dieSelbstbestimmung, eleutheria dasFehlen derFremdbestimmung.« 22 Judiciously pointed outby Raaflaub (supra n. 17) 198. 23 J.G. Starke, Introduction to International Law(10th edn. London 1989) 100. 6 quoted supra n. 8; Xen. Hell. 6.3.9: κ ρ ο σ έτ ςπ α τ ιλ ὺ ε σ α νβ ὲ α ὶὅ τ εμ 24 IG I3 127 = (II2 1). 15– α ίν ε σ θ ἱ ιε ν σ ο ιε νο εὅ ώ τ ςἐφ σ τ ὴἐά ν γ ιγ κ ο ε ςπ ἰμ ό λ ά α λ ε ιςεἶνα ι, μ ςτὰ τ ε ναὐτονόμ ο υ α ι, ο ιςχρῆσθ η τα ο ὐ ινόμ νβούλ ἷςἂ β ὶο α α ςκ ῖο η υ τ ρ χ εἑα ῆ Θ ιἑκά ε ιντ η ντῶ νἄ νπόλ ε σ ω τ 22 (378/7): ἐξεῖν [ιἐλ ]ῶ α [τ ]ω ια ὐ ευ ι θ έρ μ α τ α μ . IGII243.20– ςγρά ιλ έ σ ω α β σ ο υ σ ικ α τ ὰ τ ὰ π ο ιή 4 (334/3 ora η τ α ι...; I. Priene 3.1– λ ό β ν ἂ ιτε ν ν ]ω ιπ ο λ ία ἣ μ έν ο υ λ ιτ[ε ο ω ιπ ὶαὐτονόμ α ικ τ ν ὄ ς η ν ὸ μ ίω ν ς, μ ρ ο φ ο ρ ό υΦ ο η ν α τεφ ]ὶσ [ω μ π ι], [ἐ ή ῆ ικ α ὶ τῶ ιδ λ ο υ ντῆ ε ιβ ξ ο ]δ little later): [ἔ ν μ ω ό ν ν έ ν το ω α ὐ ιη ρ γ ω ι, Π ό λ λ ισ]υ ίω ρ υ κ ν ις[ἐ ία εσ ιρ α υ ο , τιμ ν έ μ τα ιἱσ δ ά ςτετρ ο ]ν γ ειτ]ν ι[ῶ Μ ετ[α . Same formula in 2.3; 4.4; 6.4; 7.4. Cf. 1.3; 11.14. ν τω ν ἐό ν ο τι, cf. J. Cargill in The έρ ρ ο νφ ό τ ή εφ μ ω ι .. μ ό ν το λ ὐ ]ω ὶα ν τ α ιὄ ικ ευ θ έρ 25 IG II2 43.23: [ἐ 8. Second Athenian League (Berkeley andLos Angeles 1981) 124– ς ο υ ]μ ό [ν τ]ο ρ ιη , α[ὐ ι[Π ἰσ σ νὅ ν ε ]τ ω έ ῖ]ς ο ε ιμ ν ν ο ικ ύ το α λ σ χ α υ ν Ν ω ι[κ ἐ ν 5 (334/3): τῶ 26 I. Priene 1.2– ]ς α [σ ςἐ ά ιπ]ό ῆ ςτὰ ν[τ ςοἰκ ιπ ία ε λ ]α ὶτὰ γ ν ῆ κ ν ή τ[ε ςτ τ]α ς, ἔχ[ο ν υ ]ο θ [ὶἐλ ευ έρ α ικ α ἶν ε 2 (378/7); Thuc. 2.71.2: Π ς... ία ν α σ υ α 2 (427/6); IG II243.10– ; cf. IG I366.11– ν α ρ ν χ ώ κ α ὶτὴ υ μ ο ό ἰ ὐ κ ε ς ῖ α ν ;5.18.2. τ ο χ ἔ ρ ν α φ ετέ σ ν τ ὴ ιν λ ν κ α ὶπό ῆ ἀ π εδ ιγ ιε ίδ τα α σ λ ῦ ο Π υ

The»Autonomous City-State«

27

jurisdiction within the territory,27 to admit28 or expel29 whom you want, and freedom of action in foreign affairs.30 All these aspects of autonomia cover muchof the same ground asthevarious aspects of independence. So I amnotsurprised that μ ίαsuggested in LSJ s.v. It is, in my ὐ ν το ο »independence« is the rendering of α opinion, fully justified andfully borne outbymanymoreexamples thanthose listed in the Lexicon. In the sources we are told that, for example, the following peoples were autonomoi, orhadtheir autonomia guaranteed by a treaty.

(1) Asian peoples before they became subjected to theMedes (Hdt.1.96.1). (2) Plataiai in 479 according to the oath taken after the battle of Plataiai (Thuc. 2.71.2, 4; 72.1, cf. infra page 38).31 (3) Boiotia after theAthenian defeat at Koroneia in 447/6 (Thuc. 1.113.4). (4) Thracian tribes which were independent of the Odrysian kings (Thuc. 2.29.2; 4;2.98.4; 2.101.3) 2.96.2– (5) Delphi as stipulated in the Peace of Nikias (Thuc. 5.18.2, cf. infra pages 30– 2).

(6) The Sikels living inland (Thuc. 6.88.4). (7) Leontinoi, Messene andthe Sikels according to the treaty concluded in 405 between Dionysios I andthe Carthaginians (Diod.13.114.1). (8) The peoples along the south coast of the Black sea who, apparently, were never brought under Persian rule (Xen. Anab. 7.8.25). (9) TheMysians (Hell. Oxy. 24.1, Chambers). (10) The Greek cities in the mainland andonthe islands according to the King’s Peace whenfirst imposed in 386 andlater renewed in 375 (Xen. Hell. 5.1.31, 40. cf. infra pages 38–

Some of these sources are descriptive, some areprescriptive. Often, of course, the autonomia granted by the treaties above wasnot respected by the contracting parties, cf. e.g theSpartan violation of Theban autonomia bytheoccupation in 382/ 1 of the Kadmeia,32 butthat is a different problem anddoes notaffect anattempt to establish the meaning of autonomia. Conversely, there can be no denying that a ]ικα [ο μ ν ο ὐ τ]ό ιςα ίο ν υ [τ ρ ιΓ]ο 350): ἐ ,[ο π ῖδ ὶτο ε[Ῥ τέν]ι 27 I.Cret. 4 80.1 (ca. 400– ]ι[τ ὐ τό ι; ο δ ικ ’ 2 (2). ς. See infra υ ο ίκ δ το ὐ ὶα α ῖςκ ε τελ το ςεἶν ὐ ὶα μ ο υ α ικ α ό το ν ὐ ςα φ ο ὺ ελ Thuc. 5.18.2: Δ ν ρ ιη έ ω ]τ ω ν Π ν ν ό ω ν[ἐ μ ω ι] ... αὐτονό ιδήμ ῆ ικ ὶτῶ α υ λ 9 (334/3): [ἔ δ ] τῆ ο ι, β[ο ξ ν ε 28 I. Priene 2.1– σ ιν τη κ ]γ ὶ[ἔ α κ ν ία ιτε λ ο εν ξ ὶπ α ο κ ρ ν ιπ ῶ τ ὐ ίη ια α θ ι... δεδόσ ν α ό Μ κ υ εδ ίπ π ο ιλ ιΦ ω ν ό τιγ ν ... Ἀ ςκ ῆ α ὶο γ ἰκ ία ς. ςἐ ιν ῇεἰρή ῇκο ντ ςτὰ α κ ῃ ν ςσυνθή ὶ τὰ υ νκ α ςτοίν υ ο ρ ὰ το α ςὅρκ 8: (4): π ὺ 29 Dem. 17.4– γ ντο ὼ γ ς η α νκα ὺ τα ή ν σ εσ ςε ἰςΜ ο γ μ ρ μ εγ έν ς(the common peace of 338/7) Ἀλέξανδρ α α τ τ ε ιἡ ιτά ὶἐπ α α κ ιτ ε ; (8): ἔπ υ ρἐφ ίο α ικ δ ῦ ετο τισ ό ν ρ ς, ἆ ο υ ν ν ςτυρά τα ς , ὄν Φ ιλ ιά δ ο υ π α ῖδ α ς. α ν η λ ςἝλ ὺ ςτο υ ο ρ ςεἶν χ ςἐ ῇἐλευθ ἀ υ ὶαὐτονόμ ν η ο εὐ ή α κ θ έρ α σ ικ υ ν θ ὺ ᾽ 30 Xen. Hell. 6.3.7. 31 Both passages testify to Thucydides’understanding of theconcept. Whether theoriginal oath included the term autonomia is a moot point, see Ostwald 1982 (supra n. 17) 16 with note 64 andE. Badian, “Plataea Between Athens andSparta: In Search of Lost History,”in From Plataea to Potidaea (Baltimore 1993) 115 with n. 16. 32 Xen. Hell. 6.3.11.

28

Mogens Herman Hansen

rigorous implementation of theKing’s Peace would haveresulted inthebreaking up of many more confederacies than the Boiotian and the Chalkidian, and in the liberation of many more dependencies, including the perioikic communities in Lakonia and Messenia, but again that does not change the fact that the term autonomoi intheKing’s Peace means »independent« rather than»self-governing in some more restricted sense«. Next, in the modern world a state can enter into an alliance with other states without anyinfringement of the independence of the contracting parties. Similarly an ancient polis could enter into a symmachia with other poleis without any infringement of its autonomia.33 But alliances between independent poleis were frequently turned into hegemonic leagues orevenempires andthatmeant, of course, a violation of the original autonomia of themembers of thealliance. They became hypekooi instead of autonomoi.34 The line between one andthe other was often blurred, but the essential difference was connected with the concept of consent. Members of a symmachia might have to paya tribute, in the Delian League called phoros, in the Second Athenian Naval Confederacy called syntaxis. If the tribute wasenforced bythehegemonic city, it wasaninfringement of theautonomia of the member states; if the members of their ownfree will hadagreed to paying it, there wasnoviolation of theautonomia of themembers.35

2. The Concept of Autonomia in Thucydides

Allthis is fairly straightforward. Butthere areproblems. Among ourliterary sources twotower over all theothers: Thucydides’Historiai with forty-eight ία μ ο ν το ὐ forα occurrences36 andXenophon’s Hellenika with thirty-nine.37 Xenophon is ourmain ὴ ή εἰρ ο ιν η source fortheconcept of autonomia inconnection withtheconcept of κ ν ε ν ο ιδ ὲα μ ω ντ ὸ ὐ το ν ό 33 Cf. e.g. IG II2 34, 35, 43. Thuc. 4.86.1; 4.88.1; Thuc. 1.97.1: ἡγούμ νβουλ ω νξυνό δ ῶ ευ ό ν ὸκοιν τω π ντοσ ὶἀ α νκ ά δ χ ω εἐπῆλ ν... Thuc. θ ο ντῶ νξυμμά ρ ῶ το π

5.27.2; 5.79.1; 7.57.3.

4 (Elis); 7.57.3 (the Delian League). 34 Thuc. 5.31.2– ςἐ[ν ςἙ η ν λ ίδ λ ςτὰ Χ α ρ ερ ο ή ν σ ω ιὑπ ὰ ρ ο ιτελ ε ι]ςτ λ ό ςδ ὲπ ο ςΒ]η ύ 8: [τ σ α ὰ 35 IG II2 126.13– ό ]ρ ο μφ ντὸ μπ ι τὸ ά τρ λ ιο τη έπ νκ η α [θ ν σ ὶἈ α εβ ίο ιςτὴ ὶΚ ερ ν α ικ [ω ό κ μ δ α ,κ α ὶἈ σ ά δ ε ι ο υ ς... Note that both terms (φ ) ςandσ ρ ο ό τα ύ ξ ὶαὐτονό[μ ν ις α ςεἶν α ικ α ρ θ έ τα ξ ιν , ἐλε]υ ν σ ύ have beenrestored and(with goodreason) questioned byCargill (supra n. 25) 127n.36. –On thewillingness tosubmit to international arbitration cf. Thuc. 5.27.2. ςandtheverbα ο μ μ ε ῖσ θ ν α ὐ ο το ι. Theattestations , theadjective αὐτόνο μ ία ο ν το ὐ 36 Ofthenounα are listed in Lévy (supra n. 17) 255 note 51. ςoccur inconnection withthepoleis along the μ ο ο ν τό ὐ ortheadjective α ία μ ο ν το ὐ 37 Thenounα coast of Asia Minor: Xen. Hell. 3.1.20; 3.2.12 (bis); 3.2.20; 3.4.5; 3.4.25; 4.8.1; 4.8.14 (bis); λ ςπ ό ε ις: 3.2.23; 6.5.3. ικ ε ιο ίδ ερ 5.1.31. TheAthenian klerouchies: 4.8.15; 5.1.31. TheElean π 6 (quater); 6.3.9. Thebes: 5.4.1; 6.3.9; 6.3.11; 6.4.3 (bis). The Boiotian poleis: 3.5.18; 5.1.32– 8 (quinMembers of thePeloponnesian League andother poleis dominated bySparta: 6.3.7– 5 (bis). Messene: 7.1.36. Korinth: 4.8.15. μικ ὶ α ρ ὶκ α quies); 6.3.18; 6.4.2. Mantineia: 6.5.3– μ 2 (bis); 6.5.3; cf. 4.8.14; 6.3.9; 6.3.12 (the distribution of the attestaεγ ά λ α ιπ ό λ ε ις: 5.1.31– tions onindividual poleis entails that somepassages arerecorded twice).

The »Autonomous City-State«

29

and his use of the term will be discussed infra pages 38– 40; Thucydides provides most of the evidence for howautonomia wasunderstood in relation to the Delian League. As usual Thucydides is difficult to interpret and it is particularly some passages in his work that lie behind the prevailing view that α ὐ το μ ν ίαis an ο extremely vague term, that it is usedindifferent meanings indifferent contexts, that the adjective α ςcan be used almost synonymously with the adjective μ ο ν ὐ ο τό ςandthat, accordingly, α ο ο ή κ μ το ία ὐ ν ο canbe predicated even of dependencies. π ὑ Inthefollowing I will discuss three oftheproblematical passages inThucydides: (1) thecase of Aigina, (2) theclause about Delphi inthePeace of Nikias and(3) the list of allies of theAthenians andSyracusans in413 after thedefeat of theAthenians

inthenaval battle intheGreat Harbour. Inmydiscussion of thetwofirst passages I

will take issue with the interpretation offered by Ostwald, in the third with the analysis of thepassage offered byLévy andBosworth. (1) In section V of his study (pp. 26– 30) Ostwald treats the case of Aigina, beginning withtheAiginetans’allegation in432 thatthey hadbeendeprived of their 9): »What conclusions can autonomia andending with the following statement (28– μ ία we draw from all this for the α of Aegina? By itself neither the razing of ο ν το ὐ her walls, nor the loss of her fleet, nor the payment of tribute constitute a loss of μ ία . But since there is no evidence for any state being called α α ν ο ς το ὐ μ ο τό ν ὐ ο which wascompelled (andnotmerely requested) todemolish its walls andsurrender μ ία only if it is ν ο το ὐ its fleet, andsince the payment of tribute is compatible with α ς μ ), we mayconclude that a state is α not exacted under compulsion (β ο ν ο τό ὐ ίᾳ whenit is left to exercise onits ownthemost rudimentary powers necessary for its survival.«

This is indeed a minimalist view of autonomia. But is it true? What had happened wasthis: in 457 Aigina wasdefeated by Athens andforced to demolish the walls, surrender the fleet andbecome a tribute-paying member of the Delian League.38 In the peace of 446/5 the Athenians had granted autonomia to the Aiginetans;39 but in 432 the Aiginetans claimed that the autonomia they had obtained hadbeen infringed bytheAthenians. Wehave noinformation about when andhowtheAthenians weresupposed tohaveinfringed theAiginetans’autonomia,40 but Ostwald seems to argue as follows: since Aigina had to demolish its walls, to surrender its fleet andto pay tribute andsince Aigina was nevertheless declared autonomos in the peace of 446/5, then by itself neither the razing of her walls, nor ία μ . But in ν ο το ὐ the loss of her fleet, northe payment of tribute constitute a loss of α a discussion of the autonomia of Aigina as recognized in 446 it is irrelevant what happened in457. Whatmatters is whether Athens hadinfringed Aigina’s autonomia between 446 and 432. The Aiginetans claimed that their autonomia had been infringed. The Spartans took the opportunity to make peace conditioned upon 38 Thuc. 1.108.4 ς μ ο ικ τὰ α τ ὰ ο ν τό ια ὐ α γ ο ςο ὐ κεἶν ντὸ ν τ ο ν ο ε λ νπ ό ,λ εμ έγ ι... ἐνῆ 39 Thuc. 1.67.2: Α τα ῆ ιν ἰγ 1 argue convincingς . Both Ostwald (supra n. 17 [1982]) 23 andLévy(supra n. 17) 250– σ π ο ν δ ά ly that the reference must be to the Thirty-Years’Peace of 446/5. 10. 40 S. Hornblower, A Commentary onThucydides I (Oxford 1991) 109–

30

Mogens Herman Hansen

Athens’leaving Aigina andother poleis autonomoi,41 andPerikles’answer to that was that the Athenians would grant autonomia to their own allies if they hadbeen granted autonomia in the peace of 446/5 andif the Spartans would dothe same to their allies.42 Perikles’answer to the Spartan envoys amounts to a concession that the Aiginetans’autonomia hadin fact been infringed.43 In conclusion, the case of Aigina does not warrant the minimum interpretation of autonomia suggested by Ostwald.

(2) In the Peace of Nikias the provision

about Apollo’s sanctuary

in Delphi is

ςκ α ὶ Δελφ ῦἈπόλλω ὸ ο that τ ὸδ ν ς νκ ὶ τὸ ο α ὺ νν ε ὼ ντὸ νἐ ο ῖςτο νΔελφ ᾽ἱερ ςτῆ ῆ ς ςεἶν ςγ μ ο α ικ α υ α ὶα ν ό το ὐ ὐ το ὶαὑ α νκ τελ ςκ τῶ ὶτῆ α ὶα α ε ῖςκ ὐ το υ δίκ ο (Thuc. 5.18.2). Ostwald offers thefollowing interpretation: ια τρ ά π τ ὰ ὰ τ α ν κ ῶ τ υ ἑα »theemphatic andspecific nature oftheguarantee, expressed inthethree adjectives, leaves no doubt that it was dictated by historical circumstances, particularly the frequent encroachment by the Phokians on Delphic territory. But this pedantic specificity enables us to see that, closely related though the three terms are, a form μ ία of α canbe envisaged that does involve payment of tribute andwhich does ν ο το ὐ ςexcluded by definition the μ ο υ ό ν το ὐ affect the administration of justice: for if α payment

of tribute andautomatically

guaranteed

the functioning of the lawcourts

free from theinterference of anexternal power, there would have been noneedfor ς.«44 τελ ο ίκ υ το ὐ δ το ὐ ε ῖςandα the addition of α I amnotpersuaded byOstwald’s attempt tominimize theconcept ofautonomia.45 ςand the terms ο υ μ ό ν το ὐ His interpretation of the relation between the term α α τε το ὐ ςis an illustration of a recurrent and very important υ ο δίκ το ὐ λ ὶα α ε ῖςκ problem of howto read ancient Greek documents: in thetext of a lawor a decree, how are we to interpret a technical term followed by other terms paratactically connected withthefirst one? (a) Aretheadditional terms just explanatory?, i.e. they could have been left outwithout anysubstantial change of meaning or (b) dothey addsome newaspects notcovered bythefirst term? Letmeadduce anexample. Among Athenian citizens subject toatimia themost important group was the debtors to the state. In the law quoted at Dem. 24.45 the ὶ δ η ερ ὲπ relation between the atimoi andthe opheilontes is described as follows: μ ῖς το ν τω ν ό ειλ φ ςεἶν ὀ ὴἐπ ν ςαὐτο ο ρ ιτίμ υ ςχ ω ν ὺ ὶτῶ ,ὅ τ π ω ρ ῶ νἀ τίμ ε ὲπ δ η α ι, μ νis α ν ίω η δ η ὲ... Ἀθ ῷ η ν α Ἀθ ίω ν . Here the second clause μ τ ίῳ σ ο μ η δ ο ῖςἢτῷ ε θ explanatory andcould have been left outwithout anyessential change of meaning. 41 42

43 44

45

θ α σ ι τα ίσ ν α ςτ π εἀ ία τειδα ο ςΠ υ ίο α ρ ςπ α ν ιτῶ ν τε Thuc. 1.139.1: Λ ο ν ιο η ό ι... φ α κ εδ α ιμ ᾽Ἀθ η ν ν νεἰρή ὴ ιτ α ν τ ο ιβούλ ιο ν ό ιμ α εδ κ α φ ι... 139.3: Λ ιέ α ν νἀ μ ο α να ὐ τό ν ο ιν ἴγ ὶΑ α νκ ο υ ε έλ ἐκ ςεἶν υ μ ο ςἝ ςα η ν α ὐ λ το ν ό α λ ι. ι, ε ἴηδ α ἶν ὺ ε ἰτο νε ᾽ἂ ς ο μ υ ό ν το ὐ ια τ ιςὅ λ ε ό ὲπ ςδ ν... τὰ ε μ ω ψ π ο π έμ μ ε ν ο ιἀ ά ιν ρ ο π κ ν ῦ δ ιςἀ ὲτού το Thuc. 1.144.2: ν ι σ ῶ δ ο π ἀ ν ῶ τ ῖςἑαυ ιτα ο ῖν ε κ ἀ κ ν τα ὶὅ α α θ ,κ μ ε μ ή ά ν ςἔχον σ ε ο ςἐσπ εισ ο ,ε υ φ α τ ἰκ ὶαὐτονόμ ε ἀ ι. α τ ν ο λ ύ ο ςβ ιςὡ το σ λ ι, ἀ λ α θ μ ῖσ ε ῖςἑκά φ ίσ ιἐπ ο το ν ιτη δ π ςα ὐ το ε ό λ ίω ε σ ιμ ὐ ὴ σ ᾽α ιςat ε λ ό ςπ μ ή ε σ ο νis a concession thatautonomia hadbeeninfringed, andτὰ ςἀφ υ ο μ α ὐ το ν ό mentioned at α ιν ἴγ ςat 139.3 which, again, mustcomprise Α ν α ςἝλλη ὺ 144.2 refers back toτο 139.1. SeeBickerman (supra n. 17) 322. Ostwald (supra n. 17 [1982]) 7. ῖς ε τελ το ὐ Accordingly, I prefer tosidewithGomme’s interpretation inhisCommentary p.667: »α ι.« ο μ ο ν τό ὐ ς: words necessary todefine thevague termα υ ίκ δ ο το ὐ ὶα α κ

The»Autonomous City-State«

31

But if we had not known from innumerable sources that opheilontes formed a subgroup of atimoi, it would havebeenvery tempting toassume that thetwogroups were distinct and that the clause μ η δ έ... Ἀθ η ν α ίω νadded a new provision. Conversely, in theanti-tyranny lawof 337/6 thepenalty clause is: ἄ ςἔσ τιμ ο τ ω κ α ὶ ςτ ςκ α ὶγέν ο ὸἐ ξἐκείν α τὸ μ ὐ ο ο σ ία υκ η α σ ὶἡοὑ ία δ ἔ σ τ ω α ὐ το ςθ ῦκ ῆ α ε ὶτ ο ῦ τ ὸ έκ α το ν ἐπ ιδ . Since atimia onlyexceptionally washereditary andnever automatically entailed confiscation of property, the provisions after ἄ ςἔσ ο τ ω τιμ addsomething newandarenotjust a specification of whatatimia is. Back tothePeace of Nikias. In hisinterpretation Ostwald prefers (b) above, but other sources point to (a) asthepreferable interpretation. Oneexample is theclause about autonomia in the so-called Charter of theSecond Athenian Confederacy, IG 23: ἐά ο ν ύ τ τιςβ [η λ α ι] ... Ἀθ ςεἶν η ν α ά II243.18– μ ίω α [ο νσύμ χ μ νσυμ ὶτῶ α α ι] κ ιὄ ]ω ν μ τ ω [ιἐλ ικ ι, π θ ια ευ α μ ὶα ]ω α έρ ο ν ν έν , ἐξεῖν ὐ τῶ ό ὐ το λ ιτ[ε ω υ ο ιπ χ ο λ ιτε ία ν ρ ]ὰ ή ρ νεἰσδεχομ ο τ υ ρ χ έν εἄ ω ο ι, μ ν ή τ τ α ε[φ ὑ π [δ ο μ ]ο εχ έν τα ι, μ η ω ι, λ ο ύ νβ νἂ ἣ ή μ τ εφ ρ ό ο φ ν έρ φ ο ν τ ι, ἐ π ὶδ β α ῖο ὲτο ῖςα ὐ το ρ η ικ ῖςἐ α ὶο π Χ ᾽ο ἷσ ε ἱἄλ ῖο ὶΘ α ικ λ ο ι μ α χ ο ι. Applying Ostwald’s interpretation of autonomia in the Peace of Nikias μ σ ύ to this document the inference is that a form of autonomia can be envisaged by which anallied polis didnothave its ownconstitution butwasgarrisoned, under the command of foreign officials, andliable totribute. Buthere Ostwald believes, inmy opinion correctly, thateverything after α ιisa specification ofwhatautonomia ω μ ν το ό ὐ means,46 andthat view is corroborated by thereference to the Athenian treaty with Chios, of which twocopies have survived. Inboth copies it is stated that thealliance ςτῶ νἐ β α ίν ο ν τ ντα α α ρ ὴπ α ιμ ία μ ο ν ις α το ὶα ὐ α ικ π ῖςστήλ be ἐ ία ερ θ ευ ᾽ἐλ γ ρ εγ μ μ α έ ν η δ ω μ ν έ ν , ... i.e. theonly specification is a reference to the King’s Peace (see Xen. Hell. 5.2.31) which, again, seems to have prescribed autonomia without anyspecification of whatthat meant. Intheimportant document passed in 378/7 the Athenians felt the need to spell out whatautonomia implied both positively (to have one’s own constitution) andnegatively (to avoid garrisons, foreign officials and taxes being imposed). Myposition, then, is that autonomia is incompatible with a phoros imposed by an external power. But howdoes that view square with the other sources? In the Peace of Nikias six Chalkidic poleis are singled outas autonomoi butwelearn that ρ ο ν ό ςτὸ νφ ο υ σ α έρ ιςφ ό λ ε ὲπ δ εδ σ they shall have to pay the original phoros: τά α ι... (Thuc. 5.18.5); butin thelist of Athenian ςεἶν ο υ μ ό ν ’Ἀρ το α υ ὐ ο π ν ἐ τ ὸ τείδ ισ allies in 413 Thucydides seems to assume that autonomia is incompatible with ς ν ρ ο τ έχ ε ςδ α επ ῦ α ο ρ υ ,ν ό ςφ ε τ ῖςὄν τελ ε ο π ὑ χ ὐ ιο ῖο being liable topaying phoros: Χ ο(Thuc. 7.57.3).47 Ontheface of it, these twostatements are τ ν ο μ ο ιξυνέσπ α ν ο τό ὐ contradictory, but on reflection they are perfectly compatible. The problem is not payment of phoros itself but howthe phoros is imposed. In the second passage the word ὑ ο τελ π ῖςindicates a tax imposed unilaterally by the hegemon. The first ε passage refers to theassessment of Aristeides, butweknowfrom other sources that

46 Ostwald (supra n. 17 [1982]) 48. ςεἶν α ι] κ α ὶ τῶ ν μ α [ο χ νσύμ ν α ίω η τα ι] ... Ἀθ 23: ἐά [η ο ύ λ ντιςβ 47 See also IG II2 43.18– φ έρ ο ν τι... ν ρ ο ό ή τ εφ ι... μ μ ω ό ν το ὐ ὶα α ικ τ ν ιὄ ]ω έρ [ιἐλευθ μ ά τῶ ὐ μ χ ω ν ια υ , ἐξεῖνα σ

32

Mogens Herman Hansen

the allies consented to Athenian leadership and, by implication, to Aristeides’ phoros, undoubtedly ratified during one of the meetings on Delos.48 So phoros is compatible with autonomia only if the paying polis has consented to the phoros. Similarly α τελ ὐ το ῖςinthePeace of Nikias does notmeanthat theDelphians must ε never pay any tribute to any other polis or alliance of poleis, but that the Delphians have to consent themselves to all taxes imposed. To be α τελ το ὐ ε ῖςis anaspect of μ ία , to beὑπ α μ ὐ το ν ο ία τελ ν . ο το ὐ ο ε ῖςis theopposite of α (3) At 7.57– 8 Thucydides lists all the allies first of the Athenians andthen of the Syracusans. He sets out making a clearcut distinction between those who are μ ςα α χ ία μ ο ὐ ι τό ὸξυμ π ν ο ρ ο υὑ ο τελ ε ῖς) and those who are ἀ π ι (= φ ο ο ή κ ό π ὑ (7.57.3). But in the following sections (4– 5) he seems, according to the orthodox interpretation, to mix the two categories so that some poleis appear to be both μ ο ή ιandὑπ κ ο ο α ι. Most scholars infer that Thucydides is making a mess of ν ο τό ὐ thevarious categories of allies, or, being more indulgent towards Thucydides, they assume that autonomia wasa muddled concept andthat Thucydides simply mirrors the contemporary controversy over what autonomia was or rather the confusion about how to define it.49 But interpreting the passage we must not forget that Thucydides loves skewed oppositions andthat the clue to the problem is rather to understand howThucydides uses thelanguage thanto assume that hisclassification is self-contradictory or that the concept of autonomia is understood differently in different contexts. So let usexamine Thucydides’list of Athenian allies. The Chians andthe Methymnians are the twopoleis whoare supposed to be both autonomoi andhypekooi. But in my opinion a closer look at Thucydides’ text leads to a different conclusion. ι, ο ἱδ ο ο ή κ νὑπ ὲ νδ ὸ ῶ π ἱμ First the Chians. Thuc. 7.57.3: τ ᾽ἀ νο ω λ ᾽ἄλ ,κ α ά ρ ὶτῶ τευ ο ιξ ν ὲ ν νμ ν υ ε σ τρ ο φ ό ο μ α ςα μ ο ι, εἰσ χ μ ία ὐ ἳ μισθ τό ν ο ὶδ ὲκ α ὶο ξ υ ςκ α ὶΚ ρ ῆ α ύ σ τιο ι ρ ςκ ὶΣ τυ α ῆ ςκ α ὶΧα λ κ ιδ ιῆ ετρ ρ ο υ τε λ ο ῶ π νἘρ ὑ ό α ὶφ ν κ ω ό κ η π ὑ ή ν ιο ι, ἐ κδ ὶΤ α ιο ικ ς α ν σ ω νΚ ,ἀ εῖο ικ α ὶἌνδρ π ία ὸδ ςἦσ ν ὲνή ο ία ’Εὐβ ᾽Ἰω ἀ π ρ ο υ ςδ ό α ε ,ν ῦ ςφ τ ε ιο ὐ ῖςὄν ῖο χὑ τελ νΧ ο ιο ικ π ε μ τω α ή ὶΧ σ ι. τού ιο ῖο ικ α ὶΣ ά Μ ιλ μ ο τ ιξυνέσπ ο ν . Theorthodox interpretation of this passage is ςα ο ρ ὐ ν τ τό ν έχ ο ο ε π α that all the poleis listed, from the Eretrians to the Chians, depend on the initial partitive genetive: τ ρ ο υ ό ν . I have twoobjections. ὑ π ο τελ ῶ η κ ό ω νκ α ὶφ νὑπ ὲ νμ ῶ (a) on this interpretation Thucydides involves himself in an outright selfρ ο υ ὐ χ ιο ό ὑ π ν... Χ ο τε λ ῶ ν Χ ῖο ῖο τω η ν ό ω ὶφ κ κ α ι. τού ν ὲ π ὑ νμ ὶτῶ α contradiction: κ ρ ο υ... The traditional remedy –applied especially by 19th ό ςφ τ ε ῖςὄν ε τελ ο π ὑ century German scholars –was, of course, to emend the text anddelete the words ρ ο υὑπ .50No 20th century editor of Thucydides has accepted such a ο τελ ῶ ν ό ὶφ α κ tampering with the text. (b) The orthodox interpretation of the passage takes nonotice of Thucydides’ έwhich introduces an ambiguity: from a formal έ... δ ν... δ έ use of the particles μ 65. 7; R. Meiggs, TheAthenian Empire (Oxford 1972) 58– 48 Thuc. 1.96– 5. 5; Bosworth (supra n. 17) 124– 49 Gomme/Andrewes/Dover IV 434; Lévy (supra n. 17) 264– Fora succinct synopsis of theorthodoxy seeW. Schuller, DieHerrschaft derAthener imErsten Attischen Seebund (Berlin 1974) 110 with note 184. 50 Stahl’s edition (Leipzig 1880) andClassen’s edition (revised bySteup, Berlin 1908).

The»Autonomous City-State«

33

of view the passage is organized into three parallel parts: τῶ νμ ὲ νὑπ η κ ό ω ν ρ ο υ ό ὑ π ο τε λ σ ω ῶ ν... ἀ ν... ἐ α ὶφ κ π ὸδ ὲνή κ δ ς... In theprepositional group ν ία ᾽Ἰω ἀ π σ ὸδ ω νthe preposition itself echoes ἀ ὲνή π ςbut at the same time the ία ᾽Εὐβο particle δ έechoes to some extent the μ έ νat the beginning of the period.51 In his usual crabwise fashion Thucydides begins with a category defined by status (τ ῶ ν μ ὲ νὑπ η κ ό ω νκ ρ ο ό α υὑπ ὶφ ο τελ ) butwith a slight twist he adds a geographical ῶ ν subdivision which brings himawayfromtheopening partitive genitive (ἀ πΕὐβο ς ία ω ν... ἐ σ κδ ς). Yet, Thucydides is well aware that the geoὲνή ὸδ π ... ἀ ία ν ᾽ Ἰω ᾽ graphical listing of the allies is to some extent in conflict with the status criterion expressed in the partitive genitive, and, accordingly, he adds a note about the exception, viz., the Chians. He stresses that they are not ὑ ρ ό ο π τελ ο υbut ε ῖςφ μ ο ι. Hesays nothing about whether they were ὑπ ή κ ο ο ο ιbutwhyshould he? ν α ὐ τό ςwastheopposite of being α ή κ ο ο Every reader would knowthat tobe ὑ π ὐ τό μ ς. ν ο ο ρ ό ο υ(stated π ο τελ So whatThucydides means is that theChians wereneither ὑ ῖςφ ε ή ο ι(implied by the opposition to the following α κ ο π μ ὐ τό ν ο ο ι) but explicitly) nor ὑ μ ο ι. I conclude that Thucydides does ν ο ὐ τό they alone of all the cities listed were α not mix up the concept of autonomia with the concept of being hypekoos. His stylistic idiosyncracy hasjust induced himto make use of a skewed opposition, but duly he clears upthe ambiguity by adding a note about the exception: Chios. Second the Methymnians, mentioned at 7.57.5: π ςδ ρ ὸ ἰο λ ῆ το ᾽α ῖςΑ ὐ ς, ῳ ή ὑ κ π ο ο ι, Τενέδιο ρ κ ο λ ο φ ύ ό θ ) να ὶκ υ σ μ ν ο ν α ὲ ν(ἠ υ α ῖο ὶο ιδ ιμ ὐ θ η υ ὲκ α ὶΑ Μ ἴν ιο ι ῳ ή ὑ π κ ο ο ιare traditionally ρ ό α ὶο υ σ ὶκ ὐφ ῖς. In this period the words να ε τελ ο π ὑ taken together as one phrase and rendered: »subject with ships and not with tribute.«52 But on this interpretation Thucydides contradicts his statement at 6.85.2: ς. Again the inference hasbeen to allow μ ο υ ρ ο κ ω χ ν ό ςν εῶ ῇα το ν π α ὐ ν μ α ίο υ θ η υ Μ the contradition in terms, to blame the Greeks for having confused views about ςandὑ ή ς. But κ ο ο μ ο π ο ν τό ὐ autonomia andconclude that a polis could be both α ῳ only, in which case the ρ ό ή ιwith ο ο ο κ ὐφ π nothing prevents us from taking ὑ ῳ ή ὑ κ π ο ο ι), Τεν ρ ό ννα α ὲ υ σ ιμ έ ὶ(κ ῖο ὶο α δ ν μ ιο ὐφ ι η θ υ contradition disappears: Μ δ ὶΑ α ὲκ ῖς»The Methymnians (followed the Athenians) with ships, ε ἴν τελ ιο ιὑ ο π not with tribute as hypekooi.«53 Apart from the Chians andthe Mytilenians, all other members of the Delian League hadbeenenslaved andsubjected. TheChians andtheMytilenaians werethe only autonomoi symmachoi left, andbecause of Athens’imperialistic policy they are autonomoi in namebutnotin reality. point

solitarium andinnowayconnected withthetwo ν έ isaμ ν έ 51 Thealternative viewwould bethatμ νsolitarium is not common in prose andrare in Thucydides, see J.D. έ έ . But μ following δ 4. Denniston, Greek Particles (2nd edn. Oxford 1954) 380– Thucydides Book VII with an introduction and commentary (Oxford 1965) 49; Schuller (supra n. 49) 54 n. 304. This is theinterpretation offered bye.g. Steup & Classen adlocum: »So werden dieBewohner ῇ α μ ο ιgenannt: dieselbe Sache wirdhier ὐ τό ν ο ω ο κ χ ρ α π ν εῶ 6,85,2 mitdenChiern zusammen ν ή ῳ κ ιausgedrückt: »mit Schiffen, nicht mitTribut untertänig.« ο ο ὑ π ρ φ ό ὐ ὶο α υ σ ὶκ α durch ν

52 K.J. Dover,

53

34

Mogens Herman Hansen

3. Autonomia Defined e Contrario

If myinterpretation of the three passages is on the right line, the problem about autonomia is not so much that it is a vague concept, butrather that it is a negative one. Asaptly noted byBosworth: »The wordis context hungry, acquiring its precise connotation from the circumstances in which it is used. It is also what J.L. Austin once termed a “ trouser word” , that is a wordthat is largely defined by its negative. Autonomy maybe elusive, butlack of autonomy is often easy to define.«54 Furthermore, analysing autonomia we must distinguish between (a) what the concept means and(b) bywhomit is usedandfor whatpurpose. Such a distinction

shows that most of the ambiguities and problems about autonomia concern (b) rather than (a)55 andthat if we allow for the vested interest the user has in invoking autonomia it is notall that difficult to establish or atleast to delimit themeaning of the term. This is a fairly universal and well known phenomenon in political ideology. One example will suffice. We all agree that freedom of speech andexpression is an important part of political liberty butthe British government (though it proclaims liberty as one of the basic democratic ideals andclaims thatBritain is a democracy) sometimes suppresses information andenforces censorship invoking national security as the excuse for restricting freedom of speech. It would be a mistake to conclude that theconcept of free speech in a democracy allows for censorship or that free speech is not an integral party of democratic liberty. The inference to be made is rather that the concept of liberty is tampered with by the British government, allegedly in the national interest. With this in mind let us define or at least delimit autonomia by listing its opposites andlet uspayfull attention towhoinvokes theconcept andagainst whom it is used. Everybody seems to agree that autonomia is incompatible withbeing ruled bya 8; tyrant (Hdt. 1.96.1; Hippoc. Aër 16.35; Xen. Hell. 3.1.20; Isoc. 4.117; Dem. 17.7– 6). Plut. Alex. 34.2). Even tyrants may share this view (Arist. Pol. 1315a4– Everybody seems to agree that autonomia is incompatible withbeing subject to Athenian imperialism asdisplayed through Athens’leadership of theDelian League. 4; 3.10.5; 4.87.5) Even Perikles (Thuc. 1.144.2) (Thuc. 1.67.2; 1.139.1, 3; 1.140.3– andDiodotos (Thuc. 3.46.5) admit it. Apart fromthePersians (Hdt. 8.140.1; Xen. Hell. 3.4.25) everybody agrees that autonomia is incompatible withbeing under Persian rule. (Xen. Hell. 3.2.12; 3.2.20; 6). The text of the King’s Peace shows that even the king of 3.4.5; Isoc. 4.117, 175– Persia is prepared to share this view (Xen. Hell. 5.1.31).

(supra n. 17) 123. The same point is made by R. Sealey in A History of the Greek City-States ca. 700–338 B. C. (Berkeley andLos Angeles 1976) 397: »Study of the usage of this wordinthetexts of Thucydides andother authors shows thatit is systematically ambiguous; at eachoccurrence it derives itsmeaning fromthecontext andespecially fromthecondition with which “autonomy”is contrasted.« , ι ίσ ισφ σ λ ό ε Cf. Perikles’ironic description of being autonomos in a waythat suits Sparta: π α ι(Thuc. 1.144.2). θ ῖσ ε ςαὐτονομ δ ίω ε ιτη ἐπ

54 Bosworth

55

The »Autonomous City-State«

35

Apart fromtheSpartans (Xen. Hell. 3.2.20)56 everybody agrees that autonomia

is incompatible with having a harmost anda Spartan garrison. (Xen. Hell. 4.8.1; 2). Sometimes, at least if harmosts 8, 18; Isoc. 4.117; 14.17, 24; IG II243.10– 6.3.7– andgarrisons are involved, noteven the Spartans try to deny it (Xen. Hell. 6.3.18; 6.4.2).

Especially in the4thcentury everybody agrees that autonomia is incompatible with being a member of the Peloponnesian League or what is now called the 8, 18; 6.5.3; Din. 1.73– 4; cf. Thuc. 1.144.2). »Spartan empire« (Xen. Hell. 6.3.7– Agesilaos respects the autonomia of a member (Mantineia) but only hesitatingly (Xen. Hell. 6.5.5). Everybody seems to agree that autonomia is incompatible withbeing a member state of a federation (Xen. Hell. 3.5.18; 4.8.15; 5.1.32– 3, 36; 6.4.3; Andoc. 3.13, 20; 6; 14.10, 17). Not even the Thebans try to deny it (Xen. Hell. 4.8.15; Isoc. 4.175– 3, 36).57 5.1.32– Everybody agrees that autonomia is incompatible withbeing a klerouchy (Xen. Hell. 4.8.15; 5.1.31). Everybody agrees that autonomia is incompatible with being a perioikic community (Paus. 9.2.13). Noteven theEleans tryto deny it (Xen. Hell. 3.2.23; 6.5.3). After this long list of incompatibilities it should be stressed –once again58 – that being autonomos was perfectly compatible with membership in an alliance system, as long as conditions were not imposed unilaterally upon the ally by the hegemon, andaslong asthehegemon didnotinterfere inthedomestic affairs of the ally.59

Theonly pre-Hellenistic example of anautonomos dependency is Rhittenia on [τ ρ υ ν ίο Krete, see I.Cret. 4.80: ἐ ο ιςα [ο μ ὐ ]ικ τ]ό ν ο ]ι[ο ι,] Γ ῖδ ὶτο τέν π ε[Ῥ ]ι[τ ᾽ 350 (SEG 30 1110) and, though in α ὐ τό δ ικ ο ι. Theinscription is nowdated ca. 400– forma treaty between Gortyn andRhittenia, it is infact a decision bytheGortynians forced uponthe Rhittenians whoaretreated as a dependent community.60 56 Bosworth

57

58 59 60

5.1.29.

(supra

9. Cf. also Andoc. 3.13, 20 compared with Xen. Hell. 4.3.15; n. 17) 128–

According toThuc. 1.113.4 theBoiotians became autonomoi bybeing liberated fromAthenian 4 was domination in 446. It is unknown whether the federation described in Hell. Oxy. 19.2– introduced immediately after theAthenian defeat atKoroneia orgrewupinthefollowing years. RJ. Buck, A History of Boeotia (Edmonton 1979) 159 suggests that thepoleis probably lost their autonomia by being forced to adopt anoligarchic constitution. That maywell be true. Whether being a member of the Boiotian confederation before 446 was compatible with autonomia or not, is a pseudo-problem since, probably, theconcept of autonomia hadnotyet been developed. See supra page 28 with note 33. Thuc. 5.18.5; IG I3 66.11 & Thuc. 3.10.5; 11.1, 3. 1. A similar restricted R.F. Willetts, Aristocratic Society inAncient Crete (London 1955) 110– is attested in the 3rdcentury decree of theGortynians concerning the ία μ ο ν ὐ το meaning of α Ϝ ο ικ ίο ν ς σ ν ι. ἐλευθ ῖςτὰ ο [ῦ ]ν δ ν ν ιο ιτο Κ ν τύ α ο έρ ν ο ἱΓορ α σ η ρ ώ island of Kaudos: τ εχ ά δ εἐπ ςϜοικ ῆ ν ν εα υ τὸ ν σ ιἐ , μένο τ ὶψ ρ ν τ ικ ᾶ ο τα σ π α ὰ ςτ σ ά ν ιài ὐ ςκ ὶα α ν ο ὶαὐτονόμ α κ ω μ π ν μ ιχἰρή ο λ α ἠ έμ ιτο η έν ς τυ ο ςκ ῖςΓορ ν ν μ ν ίο ο ιςχρ έν ο νχἠπ α σ τα ικ α τέσ ν ιο τύ ο ἱ Γορ 8), seeWilletts, ibidem 139. μ ν ο ό ιςτο ῖςἰδίο ις(I.Cret. 4 184a.4–

36

Mogens Herman Hansen

III. Autonomia and the Concept of the Polis

1. Down to the Kings Peace After this long exposition about autonomia I return to the connection between autonomia andpolis. If we take autonomia to be a defining characteristic of the polis andassert, e contrario, that a polis without autonomia is not a true polis, the result is that the following political communities will be deprived of the status of being a polis »inthetrue sense of theterm«. All tyrannies, e.g. Athens under the Peisistratids, Samos under Polykrates, Syracuse under the Deinomenids or Dionysios I & II, as well as Eretria in the mid 4thcentury under Ploutarchos or Chalkis under Kallias. All members of theDelian League from ca. 450 andto 404, apart from Samos (until 439 andafter 412),61 Chios andLesbos. Allthecities along thecoast ofAsiaMinor whenruled byPersia or(intheearly 4thcentury) controlled bya Spartan garrison under a harmost, e.g. Miletos, Ephesos, Kolophon andall the other Ionian cities. 371, e.g. Thebes, Phleious Many of Sparta’s allies, especially in theperiod 386– andmany of theArkadian cities. TheParrhasian cities whenthey were members of thealliance ledbyMantineia

3). (Thuc. 5.33–1– Member states of a federation, i.e., in the 4th century, most of the cities in Boiotia, Phokis, Lokris, Euboia, Thessaly, Epeiros, Aitolia, Akarnania, Achaia, Arkadia andChalkidike. The Athenian klerouchies in the 5th and4th centuries, e.g. Skyros, Lemnos, Imbros, Sestos andSamos. Alltheperioikic communities inLakonia (e.g. Kythera), Messenia (e.g. Asine), Triphylia (e.g. Lepreon) andElis (e.g. Tympaneiai).

The concept of autonomia as applied in federations requires an additional note. Some federal states were hegemonic in structure, e.g. Boiotia which not only politically butevenconstitutionally became dominated byThebes. Though all other Boiotian poleis lost their autonomia bybecoming member states of the federation, Thebes itself must have remained an autonomos polis, as is apparent from its continued capacity to enter into relations with other poleis. By becoming members of the federation other Boiotian poleis had lost their right independently to make treaties with other poleis; butThebes wasbound notonly by treaties it hadconcluded jointly with the other Boiotian cities but also by treaties which the city had concluded in its own right, e.g. the treaty of 370 between the Thebans and the Arkadian confederacy.62

61 Thuc. 8.21, see Ostwald (supra n. 17 [1993]) 55. 62 Thebes concludes a treaty in370 withtheArkadian Confederacy (Staatsverträge 273, cf. Dem. β α η ς); withAthens 378/7 ίο ςτα ςπ ρ υ ςΘ ὸ ςκαθ α λ ῖν ὐ το α ελ ε ὺ ςστή ςδ ε ῖτὰ 16.27: ... λ γ ὡ ε ιν έ (Staatsverträge 255 = IG II240);; withPtolemaios of Macedon 368 (Staatsverträge 277); with

The »Autonomous City-State«

37

Summing up,all these political communities, though explicitly called poleis in all our sources,63 would not be »true« poleis according to any definition which singles outautonomia asa defining characteristic of thepolis. More thanhalf of the more than 1,000 Greek poleis would be deprived of the status of being apolis. Such a conclusion is obviously absurd, andthe inference is that the link between the concept ofpolis andtheconcept of autonomia mustbe rejected, andthat cansafely bedonesince thelink is a modern invention notwarranted bysources of thearchaic

andearly classical periods.64

(a) No archaic or classical author discussing the nature of thepolis mentions autonomia as a defining characteristic. Plato, for example, treats the concept and nature of thepolis intheRepublic, especially inBook 2, andintheLaws, especially in Books 3 and4, buthasnota wordto sayabout autonomia; hedoes noteven use the word. Similarly, in Aristotle’s Politics there is no occurrence of the noun autonomia, andthe adjective autonomos is used only once, in a passage in which autonomoi citizens are opposed to citizens ruled by a tyrant.65 To assert with some modern historians that Aristotle in the Politics is writing about »the autonomous polis« is in my opinion misleading.66 For Aristotle it is the concept of autarkeia, not of autonomia, that is inseparably connected with the concept of the polis.67 If independence hadbeen an essential characteristic of the polis one would expect Aristotle either to emphasize that a polis hadto be autonomos or, alternatively, to subsume independence under autarkeia, i.e. to state that one aspect of autarkeia wasto be independent of other poleis. But Aristotle’s concept of autarkeia hastwo aspects only: economic self-sufficiency (i.e. to be independent of import and export) and demographic self-sufficency (i.e. to have the right number of full citizens required to accomplish man’s purpose inlife: tolive apolitikos bios). There is no hint that political self-sufficiency (i.e. independence) wasan element of the Aristotelian concept of autarkeia. Also, the author of theRhetorica adAlexandrum,

ὴ ν ἦμ ρ ὰ τῶ νἈ ν χ π α ὼ α ίω β ν ὰ λ α τ ισ Achaia 367 (Staatsverträge 283 = Xen Hell. 7.1.42: π ν τα β ι); withAlexander of Pherai α ῖο ιἡγῶ η Θ ν ή σ π ο ιἂ ὅ ε ιν μ ά ςἔσεσ χ ο ὶἀκολουθ μ υ α ικ θ α υ σ β α ν γ ῶ η ῖο τα ἡ ικ ιΘ α ὶ ν φ 363 (Staatsverträge 288 = Plut. Pel. 35.3 ὀμ νἐ ὸ σα ὐ τ ό ιδ ὲα ᾽ο ἷςἂ ε α το α ν χ ο ἶν ιΒ ιω ο ); withAthens 339 (Staats. Diod.15.80.6: σύμμ ιν σ ε ή ῖς θ υ ο λ ἀ κ ο ιν σ κ ελ εύ σ ω η ν β ε α ς α ῖνἈθ ίο υ ίω η ν , βοηθ φ ισ τ π ὸΘ ῆ τ α ιςἀ λ ιπ ό ντιςἀ ά verträge 345 = Aeschin. 3.142: ἐ ). InStaatsverträge someof these treaties areerroneously described ῖςἐ ν ις α β ιω το ο ή ῖςτο Β Θ astreaties withtheBoiotians (nos. 273, 277 and283). Forthedifference cf. IG II2 14compared

with IG II240 and43.79. 63 Formostof thecategories listed above I findit otiose toprovide evidence thatthecommunities werepoleis. FortheSpartan perioikic communities see note 13supra; fortheElean perioikic communities seeXen. Hell. 3.2.23, 30; fortheAthenian klerouchies seeIGI3285 col.i 107 &

110, Hyp. 2.18 ande.g. Skylax 58 & 67.

64 Theonly attestation in classical sources of anautonomos dependency is Rhitteneia, ICr 4.80, ς(544.21) we η τ ή ρ ιςΚ λ ό ,π ν ία η ιζ cf. supra page 35, but, apart from Stephanus Byzantius: Ῥ have noevidence thatRittheneia wasapolis. 8. 65 Arist. Pol. 1315a4– 66 E.g. Finley andWelwei (supra n. 1). 8. 8; 1328b16– 10; 1326b2– 4; 1275b20–1; 1291a9– 53a1; 1261b10– 67 Arist. Pol. 1252b27–

38

Mogens Herman Hansen

probably Anaximenes of Lampsakos, distinguishes between autonomoi poleis and poleis ruled by the king.68 (b) The opposite of autonomia is being hypekoos.69 If autonomia hadbeen an essential characteristic of thepolis, theterm hypekoos polis would have been either a nonsense oranoxymoron. Butquite a fewsources speak about hypekooi poleis in a straightforward manner.70 Many dependent city-states would of course have

preferred

to be autonomoi, but obviously a city-state didnot lose its identity as a

polis bybeing subjected to another city-state or, for example, to theking of Persia, or Macedon. (c) Theorthodox view of autonomia as a defining characteristic of theconcept of thepolis is applied to the archaic andclassical polis alike, whereas the concept of autonomia itself did not emerge until the 5th century, and the link between the concepts of autonomia andpolis wasnotdeveloped until some time inthefirst half

of the 4th century. The careful studies of autonomia by Bickerman, Ostwald andLévy have demonstrated that the first unquestionable attestation of the concept of autonomia is the Thirty-Years-Peace of 446/5, though it can not be precluded that that the term mayhave been used already in 479 by Pausanias in his guarantees to the Plataians after the victory over the Persians.71 Next, as will be explained below, the link between the concept of autonomia and the concept of polis was first made in the treaty between Sparta andArgos in418, butwasnotasserted asauniversal principle until ca. 375 in the revisions of the King’s Peace of 386, in paraphrases of the original peace andin paraphrases of the Peace of Kallias, with which the King’s peace is often contrasted. Thus for chronological reasons alone anydiscussion of theautonomos polis asa general principle should bekeptinthe4thcentury andlater andavoided in descriptions of the concept of polis in the archaic period andin the 5th century.

2. From the King’s Peace Onwards (a) The sources in which the concept of autonomia are indeed linked with the concept of polis are the treaties between city-states andalliances of city-states, especially those multilateral treaties of the4thcentury in which the Greeks tried to η , i.e. a »common peace« binding onall Greek poleis. But ν ὴεἰρή ιν ο establish a κ again these treaties show not only that all poleis strove for autonomia but also that city-states deprived of their autonomia nevertheless counted aspoleis. π ὶτ νἐ θ ὸ ο νεἴω ε ῦ ςτῶ νπόλ νδιορθ ε ω μ ο υ ό ν το ςα ὐ ρο ὖ 5: ὥ ντὰ σ π ε 68 Rhet. ad Alex. 1420a22– ιν ε γ ςἄ τ α σ εσ τώ ι νκαθ λ νβα ὴ ε ία νσ ὴ ςὑ π ὸτ ς, ο το ὺ ω τ ὕ ο ςνόμ ὸ νὁκοιν γ ω νἄ ο τ ισ λ κ ά λ ς. γ ο ςλ ό ο νὁσ ρ ὸ φ έ μ υ σ ιτ᾽ἂ ὸ να δ ύ ὶτ π νἐ ]ῶ ν ὶ [τ ]ο ρο ἱ πολ[λ ὰ ἰσ ὶγ 3; 7.57.3; Hell. Oxy. 24.1, Chambers: ε 69 Thuc. 5.33.1 & 3; 6.84.2– ς. ]ε [τ ν ο ῦ ο α κ ὐ χὑπ ςο ω έ μ ο ικ α ὶ] βασιλ ο ν τ[ό ὐ α ν σ ῶ υ Μ 4; 8.64.1; Xen. Hell. 3.1.3; 70 Ps. Xen. Ath. Pol. 2.3; Thuc. 1.8.3; 1.15.2; 2.63.3; 4.108.1; 7.57.3– 5.2.15.

71 Bickerman

(supra n. 17) 339; Ostwald 1982 (supra 7. See also Badian (supra Raaflaub (supra n. 17) 203–

3; n. 17) 23; Lévy (supra n. 17) 250– n. 31) 140.

The»Autonomous City-State«

39

In the Common Peace treaties a recurrent clause is that theGreek poleis, large andsmall alike, be autonomoi. The principal sources for the treaties in question – andso for thelink between polis andautonomia –areXenophon’s Hellenika and some of Isokrates’pamphlets, butharbingers of autonomia inpeace treaties canbe found in Thucydides. The link between polis andautonomia is known from the following peace treaties: The Peace of Nikias (421). Thuc. 5.18.2: τ ὸδ ὸ νκ α ὶ τὸ ᾽ἱερ νν εὼ ντὸ νἐ ν φ ο Δ ῦἈπόλλω ςκ ςα ελ ο α ὶ Δελφ ῖςτο ὺ ο ν ὐ το μ ν ο ό ςε υ ἶν α ικ α ὶα ὐ το τελ ε ῖςκ α ὶ ςτ ςἑαυ ῆ ῆ τ ῶ α ν ὐ το ςκ δίκ κ ο α ςγ υ ὶα α τ ὑ τ ῆ ῶ α ν κ ὶτ ὰ τ ὰ π ά τρ ια . 5.18.5: τά σ δ εδ ὲ 72φ π ό εις λ έρ ο υ σ α ςτὸ ν φ ό ρ ο ν τ ὸ ν ἐ π ᾽Ἀ ρ ισ τείδ ο υ α ὐ το ν μ ό ο ςεἶν υ α ι: ὅ π λ α δ ὲμ ὴ η δ ὲτο μ ςξυμ ά ςμ ὺ χ ν ο α η ίο υ ε ςἐ ινἈ υ έρ θ σ τ ω ἐξ έ ἐπ ιφ ῷ π ὶκα ,ἀ π κ ιδιδό ν τω ν τὸ ν φ ό ρ ο ν , ἐπ ειδ ὴα ἱ σπον έ ν δ α ὶ ἐγ ε τ ο . εἰσ ὶδ ὲ Ἄργιλ ς ς, Ἄκα ,Σ γ ο ο τά ιρ ν ς, θ ο ς. ς, Ὄ ρ τω λ ο ς, Σ λ Σ ν κ ῶ λ ο υ θ π ά ο The peace between Sparta andArgos (418). Thuc. 5.79.1: τ α ὶδ ὲἄλ λα ιπ ό λ ιε ς ῳ κ ιν ν ν ο α εό τω ντᾶ νσπ ο ν δ ᾶ νκ τα ὶἐ νΠ μ ελ ά σ ςξυμ ν α ν α ο ο ὶτὰ π χ ία ςα ὐ τό μ ν ο ο ι ςδιδόν ς, κ τ τ ια α ὰ ά ςἴσ π τρ τ ε ν δίκ ο ς, τὰ ν ἔχ κ α ὶα ῶ ὐ το π ό ν α ιε λ τ ὐ τ α ςκ ςτὰ α ε α ὶ ς. Cf. Thuc. 5.77.5. μ ο ία ὁ ςβα σ η ιλ τα ξ ξ έρ ε ςνομ ὺ ίσ The King’s Peace. Xen. Hell. 5.1.31: Ἀρ ε ιδίκα ιο ν π ό ῇἈ λ ε ιςἑα υ σ το ίᾳ ῦ ε ικ ἶν α α ή ὶτῶ σ ν ω ν ν Κ ν τ λ ὲ ν α ἐ μ ςμ ςκ ζο έ ν α ὰ τ α ὶΚ ρ ύ ο π ν , ςκ ρ α ὶμ ςπ ό α ὶ μικ εγ ά λ ςἙ η ε ιςκ ὰ λ ςα ν α λ ίδ λ α ὐ το ςδ μ ν φ ο ό ςἀ εῖν υ τὰ ὲἄλ λ α α ι ςδ ρ ο ρ ο èὥ υ υ ὶΣ . τα α κ α ύ τ κ σ ύ ρ τ μ π ν ο ὸἀρχα ε υ ή κ α ῖο ὶ Ἰμβ π λ νΛ νεἶν ὴ α η ιἈ θ 3, 36; 5.4.1; 6.3.7–12; Isoc. 4.115– 6; IG II2 7; 175– ν α ίω ν ... Cf. Xen. Hell. 5.1.32–

6. 20; 35.10– 34.13–

2: Ἀ ρ τα ςὁτῶ η ξ ξ έρ νΠ ερ σ ῶ ν The Common Peace of 375. Diod. 15.38.1– β α σ ιλ ε ς... π ὺ ὴ νἙλλ ψ ρ ά β ε δ νε ςπ έσ ρ α ε το α ιςἐξέπ α εμ κ ἰςτ α λ ὺ έσ ο ν ςτὰ τα ς η ν ν σ υ θ θ α έσ ι... συνέθ ο ή ν π ν ςτ τ ά ν ε τ ὴ ν ε ν ὴ εἰρ π ό λ ε ιςκοιν ν σ τ ς η ,ὥ επ σ ά α ε ν ή ἰρ ή ςεἰν το υ α ρ ι. Cf. Isoc. 14.10: ὁμ ρ ο υ ςπ ὶἀφ α ςκ ό τ ὰ λ ρ ο υ ε ὰ ςγ τά ς ιςαὐτονόμ ο ίω ςεἶν ο υ ςαὐτονόμ α ικελ α λ ςτῶ εύ ν ο (α π ό υ σ ιν ὶτὰ λ ἱσυνθ ω ν κ α ε ςμεγά ρ ῆ ὰ κ α ι). τ εμικ

Cf. Isoc. 8.16, 68; 14.17, 24. φ ίσ α ν η τοκα ὶο ἱ The Common Peace of 371 at Sparta. Xen. Hell. 6.3.18: ἐψ ςἐ κτῶ φ ν η ,ἐ ο σ ςτ τ νπ εἁρμ ὰ ό λ το ε ύ ω ν νεἰρήν ιτ ὴ ιδέχεσ α θ ιο ᾽ᾧ ό ν Λ α κ εδ α ιμ ὰνα υ τικ ά ὰκ ςτ α ὶτ α ὶτ , τά ὰπ εζικ ύ ινκ λ ε ε αδια δ π ε τό γ ιν ,τ άτ εστρα ε ά ἐξ 3; ςἐᾶ . Cf. Xen. Hell. 6.4.1– ν μ ο υ π ό λ ε ιςα ν ό ὐ το 5: δό μ γ α ἐπ σ ο α ν ιή τ ο The Common Peace of 371 at Athens. Xen. Hell. 6.5.2– μ ό ὀ σ ν ω α ιτὸ ν δ (ο ἱἈ ετὸ ρ η ν μ ν ε ν ὅ κ ῖνβουλομέν ο ε θ ν α ε . ἐμ ν ῖο τ ὰ νκοινω ι) μ τ ῶ ῶ φ ίσ μ η α σ ιτο ψ εκ η ῖςἈθ ν ὶτο α α ῖςψ ίω ςκ νκ α τέπ εμ σ ςβα ιλ ε ὺ α ὶ π ο ν δ α ῖςἃ τ α ῖςσ ῃἐ π ίτιν α π ό λ ιντῶ ο τε νὀμ α σ α ύ σ ῶ τρ ντό ν δ ετὸ ν , ἐὰ νδ μ έτιςσ ά χ ω ν τῶ νσυμ ῳ .Ἠ λ ρ ῷ ε ὅ κ ῖο ι ντ ςἔχα ο ά ν τ ο ιπ ε ιρ νἄλ λ ὲ νο ὖ ἱμ ι. ο ή σ ν ε θ έ ω π α ν τ ὶσ ρ κ ο ν , βοηθ ὅ ςο γ α ν ο έ α ὔ τ ιε ὔ τ ρ ῖν ςπ ο εΣ εΜ κ ιλ ς λ ο τίο α υ ν υ υ ο ςο ὐ ιαὐτονόμ δ ο έ ν γ ο ὡ δ ν τέλ èἀ ε ἱδ ςπ ρ ό λ ις, ο ε ὰ ε ςτὰ ῖο ν ικ α α η ὶο ἱ ἶν α ςγ τ ύ θ α α ρ ᾽Ἀ ιτα ετέ ς. σφ υ ίο λ υ ο τ εΤριφ ὔ ψ ν ε ,α ςε ςκ μ ο ἶν ο α ίω υ α ι ὁμ ὐ το ν ό ὶ ςἐγρα ιλ σ ε ὺ ρβα π ε σ μ ι, ὥ ο φ ε ισ ν ά η ιψ ο λ λ ἄ ςὁρκ μ ςκ ω τά ικ ψ το ν ρ α α ὺ ὰ ὶμεγά γ ισ ςπ τ λ εμ α ό μ α ε ις, ἐξέπ λ έ ὰ τ ν α σ ὶἐκέλευ α ς ,κ ςπ ά ν τ λ ο σ α ν π μ ε νἨ ὴ λ ε ὶὤ ῖω ν σα ι. κ α . Cf. Xen. Hell. ῶ ηἐ ῃ π ό λ τέ ν ε ιὁρκ ἑκ λ τ ά σ 6.5.3, 5; ςδ èπ ό λ ε ιςMSS: coni. Steup. 72 τὰ

40

Mogens Herman Hansen

The Common Peace proposed at the conference at Thebes in 367. Xen. Hell. ςὑ μ ε σ ν ο π τώ ιλ ὸβα 7.1.36: ἐρω ςὁΠ έ ω ελ ο π ο ίδ ςτ ύ λ α ίβ ο ιτ οἑα ῷ γ ρ ῆ φ α ν υ α τ ι η ντ μ ν ο νεἶν ή εα ὐ τό ν ο α ιἀ σ εσ π ὸΛ ιΜ τ ο α ν ίω κ ε εδ α ιμ ἶπ ε νὅ νκ α ὶ Ἀθ η ν α ίο ς υ ς. Plut. Pelop. 30.7: αὐτονόμ ςἝλ ςνα ῦ ν έλ κ ἀ ε ιντὰ ὲ ςμ νεἶν η ο λ υ α ιτο ν ὺ ς. α The negotiations in 343 to turn thePeace of Philokrates into a Common Peace. 2: π ρ ὶδ ε ν ὲτο τ μ ῦ ρ ο ς α Dem. 7.30– ἑτέ υ ῇ θ ώ το ή ἐπ -,ὃ μ ῃ ν α ε ορ εἰρ ν ῖςἐ ἐπ ὑ α ρ ν ο ςἙ η ς, ὅ λ ν λ α ς, ἐλευθ η ή θ ο σ ν ο ιμ ῦ θ ὴκοινω ςεἰρ σ ε ςἄ , το λ ο λ υ ῆ ὺ ν ο ῦ σ ιτ ςκ ο έρ υ α ὶ ςε μ ο ἶν υ α ι... α ν ό το ὐ η ή ε ιτ α κ α ὶἐπ κ The Common Peace of 338/7: Dem. 17.8; ἔπ ιτά τ τ ε ιἡσυνθ ς. η ν ςἙ ςτο λ α λ μ ο ςεἶν ὺ α ικ υ α ὶα χ ο ό ρ ν υ ὐ το ῇ ἐλ ευ θ ἀ έρ ςἐ ν εὐ θ ὺ Thelist reveals a development of theapplication of theconcept of autonomia. Inthe Peace of Nikias autonomia wasapplied first to Delphi andthento six named poleis; in the peace between Sparta andArgos to all Peloponnesian poleis; in the King’s Peace totheHellenic cities, large andsmall alike, butwiththeexplicit exception of the Asian poleis andtheAthenian klerouchies; theidea behind thetreaty is obviously that some poleis are to be autonomoi, others hypekooi. However, in the revised versions of the King’s Peace, from 375 onwards, theexceptions seem to have been omitted, andfinally, inlater paraphrases of theoriginal peace of 386 theprinciple of ψ ε α ςἔγρ α ιλ σ ε ὺ autonomia is likewise madeuniversal, seee.g. Xen. Hell. 6.3.12: β ςεἶν ςἐ ο μ υ ντ ςτὰ ῇἙλ ό ό ν λ ιπ λ α το ά δ ιςα ὐ ι. Still the autonomia applied ε σ α ά π only to poleis in mainland Greece, but then, by extrapolation of the concept of autonomia, it could be claimed, by the Athenian statesman Lykourgos for example, that the much better peace with the Persians concluded in 449 under Athenian auspices hadsecured autonomia for all Greek poleis, in Europe as well as in Asia ςα ςἝλλη ςε ν ἶν α ὐ μ το α ο ν ι, ό γ ο ν ο ι) ... το ρ ό σ α ὺ ςἐπ ν ἱπ τ ο(ο ή κ α ο ιή θ Minor: σ υ ν μ ὴμό ν ο ντο ςτὴ ὺ η ν ς(Lycurg. νΕὐρώ ,ἀ π το τα ικ λ νκα ο ν ςτὴ ῦ λ ὰ σ ία κ νἈ α ὶτο ὺ 1.73). A fourth-century Greek might nowargue thatif thatclause hadbeenenforced rigorously, all poleis would have become autonomoi, andby further inference one could argue that a polis which lost its autonomia would no longer be a polis in the real sense of the term. It must be a fourth-century argumentation along such lines that lies behind the modern orthodox belief that the concept of polis is inseparably linked with theconcept of autonomia. Ontheother handwemustnotforget that the fourth-century treaties themselves were not universal but allowed for exceptions, that nooneever expected that they would beenforced rigorously andthatthephrase ῇἙ ςεἶν λ λ μ ά ο δ υ ιπ α ό ό ιwaspure propaganda without ν λ ὐ το ε ιςα ςἐ τ ν ςτὰ π ά σ α anybasis in real life. (b) It is in the Hellenistic age that the concept of autonomia spreads and increasingly becomes linked to theconcept of thepolis. During the Peloponnesian war, for example, only three of the several hundred members of the Delian League were autonomoi, viz. Chios, Lesbos and Samos (from 411); all the others were hypekooi. By Antigonos’proclamation in 313, echoed shortly afterwards by that of ς... ο ν ο τίγ ν Ptolemaios, all Greek cities became eleutheroi kai autonomoi: Ἀ ὶ α ι... εἶν ιδ ν α α ὲκ ε ιο ῖν φ ίσ α τ ο π ο λ έμ η ν ἐψ ο ρ μ α κ α θ γ δ ν α σ ψ σ εδό ά Κ ν ὸ ρ α τ ν ᾽ὃ ἔγ ή ς, αὐτονόμου το ς - (Diod. 19.61.3). ρ υ ςἐλ ρ ο υ φ ς, ἀ ευ ο θ έρ υ ςἅ ν π τα α ςἝλλη ν α το ὺ

The »Autonomous City-State«

41

ς... ἀκού α ῖο ςτ το λ εμ μ Π ὰ δ σ α γ έ εδ ν ο α το ε ῖςμ τ᾽Ἀ ό ν ν τιγ ο υ Μ α κ εδ ό σ ιπ ρ ε ὶτῆ ς τῶ ή ν νἙ ω ν λ λ ψ ἐλ ςἔγρ α εκ ευ ία θ ερ α ὶα ὐ τὸ ςτ ὰ π ρ α α ή π σ λ ια , βουλ μ ό ε ν ςεἰδέν ο α ι ς Ἕλλ ςὅ η ν α ρ ο ν τίζ τ ςα το ιφ ὺ ε ὐ το ιτῆ μ ν ςα ία ο ὐ τῶ νο ττο ὐ ’ἧ χ νἈν τιγ ό ν ο υ (Diod. 19.62.1).

Inthewesta similar change becomes apparent bycomparing thepeace treaty of 405 with that of 314 in which autonomia is linked with Syracusan hegemony. Diod. 13.114.1 (405 B.C.): τ νεἰρή ὴ ὶ το ῖσ δ νἐπ η εἔθ ν τ ε ο ν ρ .Κ η χ δ α ο ν ίω ν ςἀ ῆ π ο ίκ ρ ν ςκ χ ω μ Ἐλ ο ε τ νἐ α τ ῶ υ ὰ ξἀ ύ ὶΣ ιμ α ἶν ικ ε α ν ο ύ ς .Σ ελ ιν ο υ ν τίο ςδ υ ὲκ α ὶ ς ςδ γ ς, ἔ α ν ρ τίν ὲτού ο ο ςκ ρ α ὸ υ -,π υ τ ιδ Ἱμεραίου Ἀ κ το α ιςΓελῴ ὶΚ μ α ρ α ιν α ίο ς υ ο ἰκ ὲ νἐ ῖνμ ε ιςτα ο το ρ νδ ίσ τειχ ὲτελ ό νἀ ε ῖντο λ ε σ ι, φ ό η ῖςΚ ρ δ α χ ο ν ῖςπ ίο ις. ᾽σσηνίο ςα ςκ ςδ α ὶ Σικελ ςἅ Λ ο υ εο ν τίν ὲκ α ὶΜ υ ο ὐ ε π ὺ ν το α τ μ α ν ό ο ςε υ ἶν α ι, κ α ὶ ὲ νὑ ρ α ςμ κ ο σ π ὸΔ ιο ν Σ ίο ιο ντετά ύ σ υ υ θ α ι, τ χ ὰδ ὲα ά ἰχμ λ ω τ ακ ςνα α ὶ τὰ ς ῦ β α λ π ο ςτο ο ῦ σ ςἔχον ῖςἀ ι. τ α ἀ π ο δ ο ῦ ν α ιτο ὺ σ α νδ ὲτ ά λ α ια ὰ κ τ εφ ῶ ν σ υ ν τεθ Diod 19.71.7 (314 B.C.): ἦ έν τω ντοιά δ ε , τῶ ν η ν ίδ ρ ω ν π ά ό λ κ Ἑ ε ω ν τῶ ὲ ν α τ ὰ Σ ικ λ νκ λ νμ κ εια α λ ελ νἩ ία ὶΣ ελ ιν ο ῦ ν τ α κ α ὶπ ρ ς ὸ η α δ νὑ ο ρ ν χ ίο έρ ιςτετα χ π ὸΚ θ α ι. κ α τα ις Ἱμ τα α θ ύ ρ ὰκ ο α ὶπ υ π ρ χ ῆ ο ν , τὰ ςδ ᾽ ο γ ν ία ςε νἐχόν εμ ἡ μ ο ἶν ςα α ι, τὴ υ τ ω ν ρ Σ α ὐ το ν ό υ σ α ν ά κ ςπ ο σ ίω ν . ἄ λ λ α What had happened? From the mid 5th to the late 4th century the concept of autonomia had changed its meaning from »independence« to »self-government combined with subordination to a superior power«; in thecourse of the Hellenistic period theconcept wasfurther eroded, andintheendautonomia came toimply little more than self-government in local affairs. In this sense it could easily be asserted

by and predicated of almost any Greek polis.73 In the 5th and early 4th centuries everybody agreed that forapolis toberuled bytheking of Persia orbya hegemonic polis was incompatible with being autonomos, see supra II 3. A few generations later everybody seems to have agreed that being ruled by one of the Hellenistic monarchs or being dominated by a hegemonic polis wasperfectly compatible with being autonomos. The change is reflected by the inscriptions in which individual poleis are granted autonomia or take pride in the autonomia bestowed on them by theruling monarch. Let melist theprincipal sources.

4 (334/3 or a little I. Priene 3.1–

later): decree passed by the Prienian council and μ ω νὄντω ν ν έ ω να people underscoring the autonomia of Priene: Π . Same ό το ν ὐ ρ ιη formula in 2.3; 4.4; 6.4; 7.4. Cf. 1.3; 11.14. Syll3 322 (Miletos 313/2): list of Milesian eponymoi stating that the Milesians recovered their eleutheria, autonomia anddemokratia under Hippomachos i.e. in

313/2. 73 Seee.g. Cic.Att. 6.1.24; 6.2.4. Already Mommsen described theautonomia of theGreek cities under Roman rule as»communale Selbstregiment« (Staatsrecht III.1 [1888] 658). Wemustnot forget, however, that local self-government could be very real and, especially intime of peace, might comprise almost all matters of importance for a city’s inhabitants, cf. E. Bickerman, 61 (die “Autonomie” ). Fora balanced treatment “ Bellum Antiochicum,”Hermes 67 (1932) 56– of autonomia intheHellenistic period seeW.Orth, Königlicher Machtanspruch undstädtische 5 etpassim with further references. 6, 12– Freiheit (München 1977) 3–

42

Mogens Herman Hansen

7: decree passed by thesynedrion at Ilion (ca. 306 B.C.) referring to an Syll3 330.23– embassy to the King (Antigonos) concerning the eleutheria andautonomia of the poleis whoshare the sanctuary.

50 (314/3): treaty between Ptolemaios I andIasos referring tothepolis I. lasos 2.48– as being eleutheros andautonomos andaphrouretos andaphorologetos.

90: Antigonos’ rescript (ca. 303 B.C.) concerning the synoikism of Syll3 344.88– Lebedos withTeos inwhich thesynoikised polis is granted eleutheria andautonomia (SEG 15 717).

9 (ca. 275 B.C.): Athenian decree honouring Phaidros of Sphettos i.a. IG II2682.38– because asa strategos in292/1 hehadsafeguarded thecity’s liberty, democracy and autonomia.

9: letter from Antiochos II (261–46) to the Erythraians reconfirming OGIS 223.26– theautonomia bestowed onthecity by Alexander andAntigonos. 16: decree of Smyrna (ca. 250 B.C.) referring to autonomia having OGIS 229.9– been bestowed onSmyrna bySeleukos II (Staatsverträge 492).

I.Cret. 4.184a (3rd cent. B.C.): In form a treaty between Gortyn andthe Kaudians butin fact terms imposed by theGortynians onthe Kaudians whoare nevertheless described

as eleutheroi, autonomoi andautodikoi.

SEG 19578.1–4 (Chios, second half of 3rd cent. B.C.): list of contributors to the repair of the walls introduced with a statement that the persons listed donated their money to secure that their native city could ophold its autonomia.

IG IX2 1 241 (treaty between Rome andthe Aitolians, 212/1): the last preserved clause of thetreaty is thatpoleis which join theAitolians maybecome members of the Aitolian Confederacy, still being autonomoi (Staatsverträge 536).

5: decree ofLampsakos (ca. 196B.C.) referring Syll3 591.33– to protect thedemokratia andautonomia of the city.

toFlaminius’promise

Greek cities under Persian rule were dependencies buthadenjoyed a considerable amount of self-government. So didthecities in Asia Minor twohundred years later when ruled by Hellenistic monarchs74. It was the ideology that hadchanged andthe concepts used to describe the ideals of a Greek polis. Thus the history of the autonomos polis does notendin 338 with thebattle of Chaironeia. That is where it begins, orrather, it begins withtheKing’s Peace of 387 andcatches oninthecourse of the 4th century so that in the Hellenistic period the concept of thepolis more and

31. 74 See Ph.Gauthier, ‘Lescités hellénistiques,’inHansen (supra n.3) 211–

The»Autonomous City-State«

43

more easily suggests the concepts of autonomia, eleutheria or demokratia –and vice versa.

IV. Conclusion Summing up, in an attempt to understand the nature of the Greek polis it is misleading to focus onindependence asanessential characteristic of theconcept of the polis, since hundreds of Greek poleis were dependencies. Similarly, it is misleading to confine the concept of the polis to autonomoi poleis, first because hundreds of Greek poleis were notautonomoi, andsecond because thelink between theconcepts of autonomia andpolis waspurely ideological anddeveloped inthe4th century only. Finally, it is notadvisable to speak about theautonomous polis, since themodern concept of autonomy is ambiguous andall tooeasily mixed upwith the ancient Greek concept of autonomia. This misconception is probably due to the tendency to think of the polis along modern notions of statehood: a state is by definition autonomous; thepolis is the ancient equivalent to a state; thus thepolis must be autonomous.75 If the term »the autonomous polis« or »city-state« is too deeply rooted to be eradicated, it must be kept in mindor rather explicitly stated that »autonomy« hastobe taken inonly oneof its modern meanings, i.e. intheminimal sense in which it comprises dependencies as well as independent communities and that it must not be mistaken for the ancient Greek concept of autonomia in the classical period. I suggest instead that we speak of self-governing city-states or poleis of which some were independent (autonomoi) but some dependencis (hypekooi). Whattheself-government ofthepolis implied hasbeen described inthis study as well as in previous studies andwill be the topic of future articles from the Copenhagen Polis Centre. Many of these studies will focus on »dependent poleis« or »poleis without autonomia«, expressions which according to the prevailing orthodoxy arecontradictions in terms.76

Mogens Herman Hansen The Copenhagen Polis Centre

75 See e.g. W. Gawantka, Die sogenannte Polis (Stuttgart 1985) 9 n. 1: »Da ein un-autonomer Stadtstaat evident einWiderspruch insich wäre ...« 76 I would like to thank Lene Rubinstein, Pernille Flensted-Jensen, Thomas Heine Nielsen, Antony Keen, KurtRaaflaub andtheanonymous referee formanyhelpful suggestions.

KOME. A STUDY IN HOWTHE GREEKS DESIGNATED AND CLASSIFIED SETTLEMENTS WHICH WERE NOT POLEIS by MOGENS HERMAN HANSEN

I. Introduction One of the main objectives of the Copenhagen Polis Centre is to build up an inventory of all archaic andclassical settlements that areexplicitly called poleis in contemporary sources. Themainpurpose ofthatinvestigation is tofindoutwhatthe Greeks thought apolis was, andtocompare that withwhatmodern historians think apolis is.1Theconcept ofpolis found inthesources andinmodern historiography ought, of course, to be the same. But that is far from always the case. Let me adduce just twoexamples. Theorthodoxy is that thesmall Boiotian townMykalessos was notapolis; it wasrather a komewhich is thetermusedbyStrabo andcited inRE s.v. Mykalessos.2 What is passed over in silence in the RE article andin most other studies of the history of Boiotia is that Mykalessos is called a polis by Thucydides, notjust once, butthree times ina passage where heusespolis both intheurban and in the political sense of the word.3 Similarly, the orthodoxy is that an Athenian klerouchy wasnota polis.4 Nevertheless theklerouchies Hephaistia, Myrrhine and Imbros arelisted aspoleis in theAthenian tribute lists.5 Scores of similar examples could be adduced butit would serve nopurpose to list them here. In such cases the modern historian’s reaction hasnormally been to admit that these settlements were called poleis but then to imply or to state explicitly that they were notpoleis in the true sense.6 The curious result of such a policy is the view that our sources often 1 2 3

4 5 6

323B.C. AComprehensive Research Programme,” M.H.Hansen, “Poleis andCity-States, 600– in D. Whitehead (ed.), From Political Architecture to Stephanus Byzantius. Papers from the

5. Copenhagen Polis Centre 1. Historia Einzelschriften 87 (1994) 14– Strabo 9.2.11, 15. Fiehn, “ Mykalessos,”RE XVI,1 (1935) 1005. 30. M.H. Hansen, “Boiotian Poleis. A Test Case,”in M.H. Hansen (ed.), Sources Thuc. 7.29– for theAncient Greek City-State. Acts of theCopenhagen Polis Centre 2 (Copenhagen 1995). 92. A.J. Graham, Colony andMother City inAncient Greece (Manchester 1964) 166– 110, Hyp. 2.18. See also Skylax 57, 66. IG I3285 col 1. 107– See e.g. P.J. Rhodes, “Demes, Cities andLeagues,”in M.H. Hansen (ed.), TheAncient Greek City-State. Acts of the Copenhagen Polis Centre 1 (Copenhagen 1993) 163 (writing about the 2. perioikic towns inLakonia). Cf also Rhodes’comments in Hansen (supra n. 3) 91–

46

Mogens Herman Hansen

apply

thetermpolis to a settlement that, according to modern orthodoxy, wasnota

polis.

Thecontradiction hasits root inthefact that modern historians whowrite about ancient Greece like to use the term polis synonymously with the term city-state.7 But city-state is a modern historical term which seems to have been coined in studies of the Roman concept of civitas,8 from where it wasrapidly transferred not only to studies of the Greek poleis 9 but also to investigations of Italian city-states fromca. 1100 onwards,10 ofmedieval German Reichsstädte, of Sumerian, Phoenician andEtruscan cities etc.11 Thus modern historical discussions of the concept of the city-state combine characteristics borrowed from many different cultures 12and therefore theconcept of city-state is notnecessarily coextensive withtheconcept of polis. Mykalessos maywell have been apolis in theeyes of theGreeks, although it is not a city-state in the eyes of a modern historian. The twoterms should not be used synonymously, and instead of saying that Mykalessos, though called a polis, was not a polis in the true sense, the historian ought to say that Mykalessos, though apparently a polis in the age of Thucydides, was not a city-state. In this form the statement makes sense. Whether it is historically true is a different matter which I have discussed elsewhere.13 If weacknowledge thedistinction between theancient concept ofpolis andthe modern historical concept of city-state it follows that wecanconduct twodifferent investigations which may lead to different conclusions: if we study the ancient Greek city-state andapply the modern historians’understanding of what a city-state is, weget onepicture of archaic andclassical Hellas. If wegothrough the sources andlist all settlements that are actually called poleis in contemporary sources we investigate the ancient Greeks’understanding of their ownsettlement pattern and get a different picture. It would be wrong to say that one of the two pictures is the right oneandthattheother is misleading. rather thetwopictures arecomplementary. In thefirst type of investigation onemay»start with thesettlement pattern of a landscape, so far as it can be ascertained for macro-periods (archaic, classical, Hellenistic, Roman, late Roman), to trace the actual pattern of political/social/ military articulation andgovernance of a particular landscape andits settlements, to trace theprocess of change ... inthesettlement pattern andpolitical articulation, and only then start looking at the names given to types of settlement ...«14 In such an investigation it does not matter very much how the Greeks classified the different City: From Fustel de Coulanges to MaxWeber andBeyond,”in Economy andSociety inAncient Greece (London 1981) 4. 22. 8 M.H.Hansen, “Polis, Civitas, Stadtstaat andCity-State,”inWhitehead (supra n. 1) 19– 9 W. Gawantka, Die sogenannte Polis (Wiesbaden 1985). 10 A. Molho, K. Raaflaub andJ. Emlen (eds.), City States in Classical Antiquity and Medieval

The Ancient 7 M.I. Finley, “

Italy (Stuttgart 1991).

11 R. Griffeth andG. Thomas (eds.), The City-State in Five Cultures (Santa Barbara 1981). Cf. the 11. list in Hansen (supra n. 1) 10– 53. 12 P. Burke, “City-States,”inStates inHistory (Oxford 1986) 137– 21. 13 Hansen (supra n. 3) 18– 14 Thequote is fromoneof theanonymous readers’report onanearlier version of this article.

Kome

47

types of settlement, and what they themselves thought of the settlement pattern comes second to the study of the settlement pattern itself. A prominent example of such an approach is Robin Osborne’s Classical Landscape WithFigures (London 1987). Inthisstudy thesettlements under discussion are called either »cities« or »towns« or »villages« (11). A discussion of the Greek terminology as applied to each individual settlement is intentionally avoided. And although Osborne states in his preface that he will use the English term »city« synonymously with the Greek term »polis« in its political sense (ibidem), he may, for example, usethe term village about a settlement that is unquestionably called a polis inthepolitical sense ina contemporary source.15 Suchinconsistencies, however, do not necessarily subtract anything from the value of his book, since the Greek terminology and the Greeks’ understanding of their own environment are issues intentionally left outof consideration in this type of study.16 AtthePolis Centre wehave chosen a different approach: wewantto knowhow the Greeks analysed their own settlement pattern and our investigation must be based, first of all, on a careful examination of the terminology used andthe site classification found in our sources. In this type of study it is necessary to describe anddefine the ancient concept of thepolis before webegin to compare it with the modern concept of the city-state. It is extremely important to assess the extent to which the Greeks’idea of a polis matched the settlement pattern as established by modern archaeological and historical research; but in our investigation it comes second. This maynotbe the fashionable wayof doing it. But it is in ouropinion an excellent wayof throwing the modern studies of settlement patterns into relief. Now, the core of the investigation conducted by the Polis Centre is, of course, theconcept ofpolis. But such aninvestigation mustbe supplemented with a number of other investigations starting with the question: if, in the eyes of a Greek, a settlement was not a polis what was it? An investigation of attested poleis is incomplete if notfollowed bystudies ofthealternatives. Andwhatarethealternatives? ν μ ιο λ ίδ ,π α ο ό λ ισ , ις-cognates such as π λ ό Apart from ἄ τ σ υand the various π ηandπ νtheGreeks hada whole range of designations of other forms ν ιο ο λ ίχ π ο λ ίχ ν η , τε μ ς, ἐμ ς, κώ ῖχ ρ ο π ιο ό ο ν , of settlement the most common of which were δῆμ ό ν νandἱερ ή . There is of course a certain overlap. Delphi, for example was a ιμ λ hieron that was also a polis.17 But most of the words are normally found as ηand μ ώ alternatives to polis. The two most important terms are undoubtedly κ η , which outside Athens and Attika is μ ς. The present article is devoted to κώ μ ο ῆ δ by far the more important of the terms. In this article18 I shall have nothing to say

8) Elis itself is called a »city« but all other about Elis, for example, (124– settlements arecalled »villages« although e.g. Xenophon repeatedly refers toseveral of themas poleis, see infra n. 71. Seee.g. thethorough andextremely valuable study of thesettlement pattern of a Greek region byJ.M. Fossey, Topography andPopulation ofAncient Boiotia (Chicago 1988) which focuses onsite location andhasnodiscussion of thesite classifications found inoursources. Polis: Hdt. 8.36.2; Arist. fr. 494, Gigon. Hieron: Dem. 18.157; Aechin. 3.106; Arist. fr. 412, Gigon. The three major studies of the kome in archaic andclassical Greece are outdated but not yet

15 In the section 16 17

18

48

Mogens Herman Hansen

about the Attic demes, but something about δ ςwhen occasionally, in other μ ο ῆ regions, it is used synonymously with κ η .19The other »alternative« terms, such μ ώ as teichos, hieron etc., will be examined in future studies.

II. The Concept of Kome. The Useof the Term in Modern Scholarship –and in the Sources

Theprincipal reason whytheconcept of kome hasbecome so prominent in studies of Greek history is inmyopinion thewaytheterm is usedbytwoauthors whoto an astonishing degree have shaped modern historians’understanding of the settlement pattern of ancient Greece, viz., Strabo andPausanias.20 Letmeadduce their treatment of Boiotia as an example and list the terms they use in their classification of Boiotian settlements. Strabo, Book 9.2, Boiotia:

α μ ρ Ἅ ι σ ία Ὑ ν ώ Ἑλ ε δ η ώ ν Ἀνθ ν μ α ρ υ Λ ά ι α Ἅ λ ς ο Ἴσ Ν ῦ σ α

ζο υ σ ι. 9.2.5. γ τύ νσ ο υ π ο ώ ςἀ ξ ιο λ ό η μ ο ὐ δ ὲκώ ή ν . 9.2.6. ςΛ ιμ ὸ Ἱερ ἱε ρ ὸ ν... π ο λ ν . 9.2.7. ιο ν ίχ ή ν . 9.2.8. λ ιμ ὴ ν . 9.2.8. η ,λ ιμ μ ν , κώ ρ ίο χ ω ρ ίο ν . 9.2.9. χ ω ς, ἱερ ὸ τό π ο . 9.2.10. ν ς. 9.2.15. μ ία ω κ ςτετρα η9.2.11. τῆ μ ώ κ ς. 9.2.15. μ ς. 9.2.11, τῆ ία μ κ ο ω ςτετρα η ηἔρ μ κ ώ ς. 9.2.12. δ α ὲΝ μ υ κ τέ ω τίσ ιέ νκ ρ ω νὙ π ο ικ ο ἄ ς. 9.2.15. μ ία κ ω ςτετρα ῆ η . 9.2.12. τ μ κ ώ έ ν ιμ α . 9.2.13. α σ ἔχ ο υ π ό λ ιςλ η(before ca. 270 B.C. in Lokris). 9.2.13. ν π ο λ ίχ η(before ca. 270 B.C. in Lokris). 9.2.13. ν ίχ λ ο π ςτό ηπόλ ςἔχ ς, ἴχ π ε ν ω ω ν ο ὴ . 9.2.14. ρ π ε ο π ἱερ ῖσ αat Hom. Il. η . 9.2.14. (Probably a just a conjecture for Ν μ ώ κ

ρ α ί Φ α

5. ςτετρα ς. 9.2.14– μ τ ῆ ία κ ω

β α η ῖο ι Θ φ ν ίν ιο Δ ελ ή λ ιο ν Δ ς Β α θ ύ Α ὐ λ ίς ν α εύ γ α λ ς Σ ρ α ῖα Γ ς ό σ σ η Μ υ κ α λ

19

20

2.508).

replaced: E. Kuhn, Ueber die Entstehung der Staedte der Alten. Komenverfassung und 9; H. Swoboda, Komé,”DarSag 3 (1900) 852– Synoikismos (Leipzig 1878); G. Fougères, “ 76. By far thebest recent account is E. Lévy, “Apparition en RE Suppl. 4 (1924) 950– ,” η μ ώ Κ “ 27. Grèce del’idée devillage,”Ktema 11 (1986) 117– See. N.F. Jones, Public Organization in Ancient Greece (Philadelphia 1987). Outside Attica 9), Kalymna demes are found in Amorgos (Jones 214), Elis (Jones 145), Euboia (Jones 73– (Jones 231), Kos(Jones 239ff), Miletos (Jones 323ff), Naxos (Jones 213) Rhodes (Jones 243– 9), Stratonikeia (Jones 335) andThessalonike (Jones 268). See also S.E. Alcock, “Pausanias andthePolis: UseandAbuse,”inHansen (supra n. 3) 326–

44.

Kome

Ὗ λ α ι ς Σ χ ο ῖν ο ς Σ κ ῶ λ ο Π ό τν ια ι Θ έσ π ια ι ρ έο υ σ α Κ η ρ σ κ Ἄ Π ετ ε ώ ν η σ ις Ε τρ ὔ

η . 9.2.20. μ ώ κ ςΘ β η α ικ ῆ ς. 9.2.22. ῆ ρ α τ ώ χ η . 9.2.23. μ κ ώ

ις. 9.2.24. λ ό π ις. 9.2.25. π ό λ ἱΘ νἔχου ινα έ σ σ π ι. 9.2.25. ιο ια ε π ίν ἐ ῇἐ ν ὶκ ρ μ ῃ ώ υ ” . 9.2.25. ο ἰζ η . 9.2.26. μ ώ κ “ ν . 9.2.28. ίο μ ω κ ίο ν . 9.2.28. ρ ω χ μ α ν τε ῖο ν ς. 9.2.34. , ὄρ ο

Π τῶ ν ο ι π α ν μ έ ις. 9.2.27, 36. λ λ κ ο ό Ἀλ α ς. 9.2.39. α Λ εῦ τό π ο τρ κ

Pausanias, Book 9, Boiotia: Π λ ά τα ια

ι ία σ Ὑ ί α ρ Ἐρυθ ς Σ κ ῶ λ ο ι α β ῆ Θ ς η σ σ ό Κ ερ Π ό τν ια ι ς η σ σ ό μ Τ ευ ς Γ λ ίσ α α μ ρ Ἅ σ ησ ς λ Μ υ κ α ό ρ α γ Τά ν α ν ώ δ η Ἀνθ ν ιο ν ίφ α ρ Ἀκ ν ο Π τῷ ν μ α ρ υ Λ ά Κ ῶ π ι α ς μ ω ν Ὄ ε λ ς ο ττ ε Ὕ ρ ς Κ τω ύ ν ε ρ σ ε ία Κ ο ί α Ἁ λ ο β ι εῖρ Κ α ς τό σ η χ γ Ὀ Θ έσ π ια ρ εῦ Κ σ ις

2. ό λ π ις. 9.1.1– ό ις. 9.2.1. λ ια ... π είπ ἐρ ις. 9.2.1. λ ό ... π ια είπ ἐρ είπια 9.4.4. ἐρ ἡμ ὲ νκ ά τ ω π ό λ ις... τὴ νδ ὲἀ ρ ό κ π ο ιν λ . 9.7.6. ν9.14.2. ίο ρ νχω ὸ ρ υ ἐχ ια . 9.8.1. είπ ἐρ ν . 9.19.1. ίο ρ ω χ ρ ε ίπια 9.19.2. ἐ ς. 9.19.4. ω λ ε ό π ια είπ ἐρ ρ ε ς. 9.19.4. ίπ ιαπ ό λ ε ω ἐ π ό λ ις. 9.20.1, 2. π ό λ ις. 9.22.5. π ό μ λ α ισ . 9.23.5. ις– λ ό π ο ν . 9.23.6 ἱερ π ό λ ις. 9.23.7. μ α . 9.24.1. π ό λ ισ η . 9.24.3; 9.34.10. μ ώ κ .

.

η . 9.24.3; 9.36.6. μ κ ώ α . 9.24.4. μ ισ λ ό π μ α . 9.24.5. ισ π ό λ α . 9.24.5. (before μ ισ λ ό π τ ε ο . 9.25.6. π ό λ ις... π ρ ς. 9.26.5. ε ίπ ό ια λ π ε ω ἐ

ca. 270 B.C. in Lokris)

ις. 9.26.6. π ό λ . 9.32.1. ν ω ιέ π εσ èmveiov Θ ις. 9.32.2. π ό λ μ α . 9.32.4. α ισ ό λ Τ ίφ π ς ρ τ ο ις. 9.32.5. Ἁλ ία π ό λ η .σ9.33.5. μ ί κώ α μ ν ε ο Ἀλ α λ κ ί

Θ

β

η

49

50

ύ α ι Φ λ εγ ς ό εν χομ ρ Ὀ η ν δ ώ π λ Ἀσ ά δ ια ε Λ εβ ν ώ ια ε ιρ α Χ

Mogens Herman Hansen

π ό λ ις(in themythical period). 9.36.1– 2. π ό λ ις. 9.34.6; 9.36.6. π ό λ ις. 9.38.9. 2. π ό λ ις. 9.39.1– 6. π ό λ ις. 9.40.5–

The range of terms used for site classification is: ἀ ρ ό π ο λ κ ις, ἐρ είπ ια , ἱερ ό , ν ή ν α η μ ν ,μ , κω ίο τε ν ,λ ῖο μ ιμ ν ,π ώ ό κ λ ις, π μ ό α λ ισ ,π ο η λ ίχ ,π ν ο λ ς, ν ιο , τόπ ν ο ίχ ρ ίο ν ν , χω . But, apart from polisma (which Pausanias uses synonymously ειο π ίν ἐ with polis throughout his work)21 the two recurring terms applied by Strabo and Pausanias to most of the settlements they describe arepolis andkome. Following these two late authorities, it has become almost an orthodoxy, even in studies of archaic andclassical Greece, to hold that a small settlement which wasnot a polis musthavebeena kome, except inAttica andsomeother regions where thepreferred ς. In myopinion the orthodoxy is grossly misleading, andlater in this μ ο ῆ term wasδ return to the Boiotian settlements classified as komai, butbefore I start will I article myowninvestigation of theconcept of komeletmeoffer a short Forschungsbericht. Ever since 1878, when Kuhn published his influential study about the Komenverfassung, theword komehasbeenconsidered almost a technical term for a constitutional unit smaller than thepolis or replacing thepolis in regions where there were nopoleis: inthebackward parts of Greece it wasthefundamental unit of a tribal state (in German: Stammstaat) andin the more developed parts kome, side by side with demos, was the basic territorial civic subdivision of a polis. This constitutionalist approach permeates Swoboda’s major article inRE s.v. andis still apparent in thearticle inDerKleine Pauly.22 A comprehensive analysis of theuses andmeanings of kome, however, shows thattheconstitutionalist viewof thekomeis too narrow and one-sided. Both in inscriptions and in literary sources kome is commonly used to denote a village in the socio-economic sense without any constitutional or political functions. The use of kome, like demos, to denote a municipality is much rarer than usually believed, andwhen the word is used in a constitutional context, the clearcut distinction normally made between polis and kome is muchtoo sharp andignores a certain overlap between thetwoterms. Kome is traditionally rendered by village but, like polis, it has a whole range of meanings anduseswhich mustbedifferentiated fromoneanother toavoid confusion. 21 L. Rubinstein, “Pausanias as a Source for theClassical Greek Polis,”infra 214. 22 Kuhn (supra n. 18); Swoboda (supra n. 18); F. Heichelheim, “Kome,”Der Kleine Pauly 3 (1969) 279: »[kome] kommt zuerst inderPseudo-Hes. Aspis 18 vor. Gemeint ist seitdem eine polit. untergeordnete Dorfsiedlung oderStadtunterteil, inStammstaaten dieunterste Verwaltungseinheit.« See also F. Kolb, DieStadt imAltertum (München 1984) 59: »Eventuell vorhandene weitere Siedlungen auf dem Gebiet einer Polis hießen Demos oder Kome; letzteres Wort bezeichnete aber auch die unbefestigten Siedlungen einer nicht durch Aufteilung in Poleis, sondern alsethnos (‘Stammesgemeinschafft’, ‘Landschaftverband’) organisierten Region«; V. Ehrenberg, TheGreek State (2ndedn.London 1969) 9, 23.N.Jones (supra n. 19) 21 speaks of »the widespread institutionalization of villages asdemoi, komai etc.« Butit isinfact hiscareful collection of theevidence which shows howfewattestations wehave of komeinthesense of 3. »municipality«, seeinfra pages 71–

Kome

51

Since it is often stated in our sources that kome was the form of settlement that preceded the polis we must be aware of the difference between kome used in a historical, sometimes even in a pre-historical context, and kome used about contemporary society. Furthermore, wemustdistinguish between kome in a purely topographical sense, i.e. when it denotes a small settlement, andkome in a more political and constitutional sense, i.e. when, like demos, it denotes one of the subdivisions of the polis and of its citizens. Finally instances in which polis is opposed to kome 23must be treated separately from those in which kome is seen as ἱκ thebackbone of the polis-organization itself, i.e. α α τ ὰ ςοἰκούμ μ κ ώ α ε ν α ιπ ό λ ε ις. Though settlement invillages wasprobably theprevailing form of habitation in archaic andlarge parts of classical Greece,24 the wordkome is surprisingly rare in archaic andclassical sources. Thewordis notfound in theheroic Homeric poems, but occurs occasionally in Hesiod’s more humble hexameters, once for example about his native village Askra.25 In Thucydides the wordpolis is used 849 times, whereas there are only eight occurrences of kome. For Herodotos’ Historiai and Xenophon’s Hellenika thefigures are469 versus 20 and489 versus 10respectively. In Xenophon’s Anabasis, ontheother hand, komeoccurs 91 times, butexclusively about barbarian towns and villages. Most of our attestations of kome come from authors of the Hellenistic andRoman periods. Adding upHerodotos, Thucydides, Xenophon, Plato, Aristotle andall theAttic orators wedonotreach 200 attestations, whereas Strabo alone uses kome 225 times, and in his work the ratio between occurrences of kome andpolis hasrisen from 1:100 (Thucydides) or 2:100 (Xenophon’s Hellenika) or 4:100 (Herodotos) to over 16:100 (225 versus 1424). But it is notthe chronological dimension that is important in itself. It is rather a matter of genre (history versus geography) and of topic (description of Greeks versus barbarians). For both reasons Strabo scores higher than anyother author. Heis a geographer, nota historian, andinmanyof hisbooks hedescribes peoples living on the fringes of or outside Hellas. In Diodoros andPlutarch, on the other hand, the ratio between kome andpolis is as low as in classical historians, whereas another geographer, Pausanias, reaches 6.5:100 andArrian nolessthan8:100, again because hemostly describes barbarian peoples andregions. Nota single oneoftheoccurrences of kome in Arrian concerns a Hellenic community.26 Onemight object that it should not cause anysurprise that historians use the termpolis muchmorefrequently thattheterm komesince inaccounts of politics and warsit is onlynatural tomention »states« rather than»municipalities« or»villages«. But this observation, though it contains a core of truth, is not quite to the point. It is ιξ ]ά ό ςἔπρα ερ εδ θ ε ς. Pl. ν νἱμ ε τ α α π ό μ λ ιν ,κ ε[ρ ςυ ώ ὲκ ἱὸ ο ιμ λ ςἄ κ 2: ... Δ ιὸ 23 Hes. Fr. 43a.61– ὰ τ τ ὔ ν α ετῶ κ ιςο ιθ ίο σ έο υ ερ ν κ ο ο ιπ ή ιο υ ῖςΔ ιλ σ ιτο ἵτ ς...ο εφ ν ε ο μ θ εά ο ιλ ἵτ εφ Resp. 475D: ο μ ε ν ο ι. ό λ ειπ ο π ςἀ α μ ώ κ ὰ τ α κ ν ὔ τ ετῶ π ό λ ε ιςο 16; R. Osborne, Classical Landscape 24 J.N. Coldstream, Geometric Greece (London 1977) 303– with Figures (London 1987) 68, 124ff. ε ι α ακ κ έρ ῇ ῃ ,θ ,χ ῖμ ε ῃ ρ , Ἄσ κ μ ρ ῇἐ ν ὶ κώ ςοἰζυ ο ν ῶ 25 Hes. Op. 639– τοδ χ α σ γ σ ά ικ λ 40: ν ᾽ἄ ᾽Ἑ γ α λ ρ ἀ ; Scut. 18; Fr. 43a 62 (quoted supra n. 23). τ᾽ἐσθ ο ῇ δ έπ λ έῃ ὐ ,ο 26 Diodoros: 42:2541 = 1.5:100; Plutarch 50:2369 = 2:100; Pausanias 50:767 = 6.5:100; Arrian: 37:458 = 8:100.

52

Mogens Herman Hansen

true that many battles were fought outside a polis, e.g., in Boiotia, the battles of Plataiai in479, thebattle of Tanagra in457, thebattles of Koroneia in447 and394 andthe battle of Chaironeia in 338. Butjust as many battles were fought near what wewould call villages, e.g. theBattle of Keressos ca. 520, thebattle of Oinophyta in 457, thebattle of Delion in424, thebattle of Tegyra in375 andthebattle of Leuktra in371. Whereas Plataiai, Tanagra, Koroneia andChaironeia areexplicitly described aspoleis inoursources, noneof theothers is ever referred toin archaic orclassical sources as being a kome. Furthermore the epigraphical sources reveal the same pattern: though manyinscriptions concern local administration andtheinfrastrucure of the polis the term polis occurs much more frequently than the term kome, but in this case it is impossible to present meaningful statistics. For a collection of the 8. attestations see infra pages 63–

III. Kome in a Historical Context

For the view of kome as anearly form of settlement in Greece twosources tower over all theothers: Aristotle’s description of thedevelopment of thepolis inPolitics Book 1, and Thucydides’ substantiation in the Archaeology of his claim that the Peloponnesian Warwasthebiggest of all wars ever fought by manin history. In Politics Book 1 Aristotle states that thepolis is anaggregate of komai, that thekomeis anaggregate of oikiai andthat thepolis emerged by a synoikism.27 The 3Aristotle Book in rest of Book 1 is ananalysis of theconcept of oikia;analyses the polis. Buttheconcept of komeis never given a proper treatment. It isjust mentioned in the beginning of Book 1 as the link between the oikia andthe polis, and in the later books referred to only twice andin passing.28 Theother mainsource forkomeina historical context is Thucydides’Archaeology in which theterm is used twice andboth times notjust in passing butin passages which areessential for hisdescription of thedevelopment of Hellenic societies. (a) ςτ ὸπά η ν λ α ικ ε α ὶ λ λ Ἕ ρ ὰ Writing about the age of Minos Thucydides states: ο ἱγ ῳ π ρ α α ςεἶχ ο θ υ ἠ α λ π ο ά ν... σ ῷ ε ίρ σ ιο ικ α ὶὅ ο ι νήσ ρ σ ω νο ἵτ εἐ ντ τ νβαρβά ῶ α ζ ο ν... α ιςἥρπ έν ςοἰκουμ α μ ώ α τ ὰ κ ὶκ α ςπ λ ιςκ ό ε το σ τειχ ίσ ινἀ τ ε ρ ίπ ν ο σ το π π (1.5.1) (b) In the famous chapter about what Athens and Sparta would look like if left in ruins for posterity to look at Thucydides offers the following explanation of ς η θ ισ είσ οικ εσυν τ ὔ whySparta would fare sobadly incomparison withAthens: ... ο 20: δ ιὸκ α ὶτ ὸ 27 Polis seen as a city, created by a form of synoikismos: Arist. Pol. 1252b 19– δ 8: ἡ ; 27– ν ν ο ω ν λ ρσυνῆλθ ειό κπ ᾽ἐ γ ὰ ω ν ν μ έ ιλ ο σ ευ κβα ό λ ις... ἐ ἱπ ε οα τ ν ο σ ιλ εύ ρ ῶ το νἐβα π ις... SeeN.H.Demand, Urban Relocation inArchaic andClassical λ ςπό ία ν τέλ ιο ε ω ν ιν μ κ ῶ ο ω κ 7: »synoikisms werefrequent inthe5thand4thcenturies ... Aristotle’s Greece (Bristol 1990) 26– acquaintance with this form of polis creation through power-building synoikisms may well have misled himto apply the concept anachronistically to the problem of the origins of the

polis.«

ςὦ σ μ ι, ὴκ ὰ α α τ κ ώ ῳ τ νμ α ς, ὅ κ ν α ὶπ ο υ ό λ ιςἔθ ῷ το ιο τ ύ 9: διοίσ ε ὲτ ιδ 28 Arist. Pol. 1261a27– μ ν ῶ ῶ νκ α ὶκω 81a1: π ὲἡγεν λ ιςδ ό ς. 1280b40– ς ῆ θ ,ἀ ο λ λ δ ε ά ρ κ λ ὸπ ο ιτ ν μ έ νἈ ισ ρ ἷο ω κ εχ ᾽ο ςτελ ῆ ε ςκ ία α ς. ὶα ρ ο κ υ τά ὐ ζ ω ία ν ω ιν ο κ

Kome

53

π ό ςο λ ω ε ὔ τ ε τα σ α ὶκ κ ο α ευ ς, κ α ο μ η λ η υ σ έ τελ α ν ὔ ῖςπ ο ιχρ τ έ σ α τ ῖςκ ε ὰ μ κ ἱερ ώ ςδ α ῷ ὲτ λ λ ῆςἙ ά ῳ δ ςτρ ο ο ό ἰκ π α π λ τ α ισ ιῷ θ είσ ς, φ η ά ν ο ιτ᾽ἂ νὑπ ο δ εεσ ρ τέ α[sc. ἡπ ό λ ις

ο ]... (1.10.2). ν ίω ν α Λ κ α ιμ εδ The account found in Thucydides Book 1 is often juxtaposed with the account inAristotle’s Politics Book 1,29butit isusually overlooked thatthere is a remarkable difference between howThucydides and Aristotle treat the kome as a historical phenomenon. First, in the Politics life in komai is a stage in civilization which precedes thelife inproper poleis, whereas, inThucydides, τ ςοἰκ μ ὰ α ὸκ τ α κ ώ ε ῖνis the way of life characteristic of early poleis. Second, the kome envisaged by Aristotle is different from thekome Thucydides hasin mindwhenhe speaks about ςοἰκούμ μ α α τ ὰ π ό λ κ ώ ιςκ εν ε α ι.

(1a) According to Aristotle the polis emerged by a synoikism of komai; it was formed asa socio-economic community witha viewtothemaintenance of life itself (τ ὸζῆ ) butexists as a political community for the purpose of the good life (τ ὸε ν ὖ α τ ὰ ), andinthis form it is the telos of human life.30 Aristotle’s view of thelife κ ν η ζ ςas more primitive than life in a polis is repeated in the second book of the μ α κ ώ σ ι ςᾦ μ α ὰκώ ὴκ τ α νμ τα ς, ὅ κ α ὶπ ο υ ό ν λ ῷ ιςἔθ το ιο ύ τῳ Politics: δ ε ιδ ὲτ ίσ ιο ς, ἀ ῆ θ ο λ λ ς.31The enigmatic reference to the λ ὸπ ιτ ο μ ν έ δ ά ε κ ρ ισ ρ κ νἈ εχ ω ἷο ᾽ο Arkadians is a notorious crux, butthere canbenodoubt that, according to Aristotle, life in komai without any major urban centre is characteristic of an ethnos andis essentially incompatible with life in apolis. Andthe same viewis found once again in twopassages from Plutarch’s TheGreek Questions. Oneof the main sources for this treatise seems to be Aristotle’s collection of 158 Politeiai of which the 157 are lost butfor a fewfragments.32 Question no. 17concerns Megara andmaybe derived κ ε ῖτ οκ α τ ὰ ὶςᾠ ρ α εγ νἡΜ ιὸ α λ α from the Constitution of the Megarians: τ ὸπ τ οδἩ ν ο , ἐκα λ ῦ α ὶ ν ῶ μ έ ν ιτ η ω ντῶ λ ρ νπ ο α ηνενεμ ε ῖςκ ρ έ ν τεμ έ ς, εἰςπ μ α κ ώ ι (Plut. Mor 295B). ιο ίσ κ δ ο α ὶ Τριπ ρ ρ ε ε ῖςκ ῖςκ α ὶΚ α ο υ ο σ υ ν α ὶΜ εγ α ε ῖςκ Π ιρ polis: it is the νrefers to a period before theformation of the᾽ ιό ὸπαλα Obviously, τ region Megaris thatis split upinto five komai, notMegara; TheMegarians themselves constitute one of the komai, and the presumption is that the polis Megara was created by a synoikism of thefive komai. Similarly, thesource for Question no. 37 is presumably the Aristotelian Constitution of the Tanagraians, of which no other ς ῆ ςΤαναγρικ ῆ ρ ,ἔ τιτ ή τ α π ὁ ν ςτοίν υ υ ο ρ ο π α ν δ ίπ ίμ ο fragment hassurvived: Π Ἐφ ςὑ μ π ὸτῶ ε ν ο ρ ν κ ο ύ ο ιπ ιο τ λ ν ο ῳ Σ τέφ ῷ μ κ α έ ν λ ο υ ς, ἐ η ντ έν ςοἰκουμ α μ α τ ὰκώ κ ρ τω ονύκ ῖν νἐκ ε ίο ρ ὸχω ετ ιπ ιν , ἐξέλ ε τεύ α ισ τρ α σ υ θ ὴβούλεσ ὸμ τ ιὰ νδ ιῶ α Ἀ χ ε(Plut. Mor. 299C). Again the episode concerns the νἐτείχ ισ ία δρ ν α ιμ κ α ὶτ ο νΠ ὴ remote past. Stephon, if correctly identified, is some 6 kmremoved from thecity of ς α μ ὰκώ τ α Tanagra,33 and the passage conveys the impression that habitation κ belongs to thepre-polis-period. E.g. Swoboda (supra n. 18) 951; seealso Fougères (supra n. 18) 853. 7. 30; 1278b24– Arist. Pol. 1259b29– 9. Arist. Pol. 1261a27– FortheAristotelian collection of Constitutions asthemainsource forPlutarch’s treatise see K. Philologus 60 (1901) 464. Giesen, “Plutarchs Questiones Graecae undAristoteles’Politien,” 33 Fossey (supra n. 16) 54.

29 30 31 32

54

Mogens Herman Hansen

To sum up, in Aristotle’s model

η μ . έ ν ο ἰκ ο υ

there

is no room for a π ό λ ιςκ α τ ὰκώ μ ς α

(lb) Thucydides describes the reverse phenomenon: he believes that there were poleis already in the age of Minos,34 andin the Archaeology the life in komai is typical of early poleis and does not precede the formation of the polis itself. So, according to Thucydides, the polis was originally a form of political community only, and the polis as a (walled) city developed between the age of Minos and the Trojan War.35 In agreement with this view of the development of thepolis the synoikism of Attika under Theseus is described by Thucydides in Book 2 as a purely political act which hadnoconsequences forthesettlement pattern: theformation of onepolis out of manypoleis (each probably composed of several komai), andthesynoikism was ὴδ ὲθησ ς ε ὺ ειδ followed only later by the formation of the city of Athens: ἐπ ρ ὴ νχώ α νκ α ὶ καταλύ η σ ετ ςτῶ σ νἄλ α σ α ίλ εἄ ε αδιεκόσμ λ ω σ λ λ ν ε... τ άτ υ ἐβ ςτ ςἐ ὴ νν ρ χ ὰ ςα ῦ ν π ό λ ιν ρ ο ια α ν κ α ὖ σ ή α ,ἓ ὶτ νβουλ ευ ο υ λ τ ρ ευ τ ή ό λ ν τ ά τ εβ π ε ω μ ςτ κ έν ισ ς, κ ο ο αα επ α υ ρ ά ν ὶ νεμ τα ςκ υ τα ν α ὑ ιο νἀ εῖο ὶπ τῶ ν ο δ α π , ξυνῴ είξ ν η π α ι, ἣἁ ά ν δ τ ῃχρῆσθ ω νἤ ια π γ ό κ α λ σ ατα τ εμ ύ ν ἠ ά π ςἅ ῦ ὸτο ρ υ το σ ὶπ ά α ρκ ἑκ ε ηπαρεδό ηὑ έ ηγενομ ν θ π ςτο ὸΘ ά ή σ λ ε ω ὴ νμεγ ῖςἔπ ςα ὐ ε τ ιτ α ξ νἐ υ ν τελ ο τω ύ ν (2.15.2). Onthe other hand, the π ηis, in Thucydides’ ςοἰκουμ έ ν α μ τ α ὰκώ λ ιςκ ό view, a historical phenomenon belonging in the (remote) past, and describing contemporary societies he takes it for granted that a polis is a political community centred on a conurbation. In his work he uses the wordpolis in the sense of town some 250 times but–as far as it canbe tested –only about a town which wasthe political centre of apolis intheconstitutional sense.36 Conversely, apart fromSparta for which see infra (2b), there is no example in his work after theArchaeology of polis in its constitutional sense being used about a community that was still ς. μ α ώ κ ὰ τ α inhabited κ The second major difference between Aristotle’s andThucydides’accounts of thedevelopment of civilization concerns their understanding of what a komeis. (2a) In Aristotle’s model thekomai areproper villages spread outover a fairly large territory. InMegaris, forexample, Tripodiskos is 10kmremoved fromMegara and if theHeraieis andthePeraieis have been correctly located inPerachora andSWof Gerania respectively, the distance to Megara is 40 and20 km.37

ηconveys a different ν έ ςοἰκουμ μ α (2b) Thucydides’ concept of a π ὰ τ κ ώ α ιςκ ό λ understanding of what a kome is. His only example is Sparta. It is universally believed that the komai referred to by Thucydides are the four villages Limnai, 34 Thuc. 1.2.2, 6; 1.3.2, 4; 1.5.1, cf. A. Snodgrass, “Interaction byDesign: theGreek City-State,” in C. Renfrew & J.F. Cherry (eds.), Peer Polity Interaction and Socio-Political Change (Cambridge 1986) 48. 35 Thuc. 1.7.1. 45. 36 See Hansen (supra n. 3) Appendix I 39– 53. 37 R.P. Legon, Megara (Ithaca andLondon 1981) 49–

Kome

55

Kynosoura, Mesoa andPitane.38 Although none of the four is called a kome in any ancient source I accept theorthodoxy, butwanttopoint outthat, though Sparta had nowalls inthearchaic andclassical periods, thefourvillages musthave formed one nucleated settlement. They occupied anarea of some 3 square km., andin theearly 5th century they must have been inhabited by, sometimes, as many as 8,000 spartiatai39 andprobably by their families as well. A population density of several thousand adult male citizens per square km. is quite enough to reveal that Sparta must have been a conurbation, in spite of the absence of walls andmonumental temples. Corroboration of this view is found in Thucydides himself: at 1.134.1 he tells us that the ephors hadPausanias arrested in thepolis, i.e. in Sparta. So here the termpolis is used in the sense of conurbation about Sparta in spite of the fact that, η ςοἰκουμ . Similarly Herodotos 40and έ ν μ α according toThucydides, it wasκ α τ ὰ κ ώ Xenophon 41call Sparta apolis inthesense of urban centre, andinthefamous oracle given to Sparta during the Persian Wars it is even described as anasty.42 Modern historians’belief thatSparta wasapolis without anurban centre 43hasnofoundation in the sources which show that, in the eyes of the Greeks, Sparta was the urban centre of the Lakedaimonian polis and must have accommodated a five digit number of inhabitants. Theonly passage in which habitation in villages recurs inThucydides’work is μ γ α ὲ νεἶν έ ςμ α ιτ ρἔθ ὸτῶ ο ν ὰ ν ὸγ in his description of the Aitolians at 3.94.4: τ ς, κ α ςα ςδ ὶ τα ιὰ τειχ ίσ το υ ύ τ α μ α ὰκώ τ νδ ὲκ α ο ν ο ῦ , οἰκ ά Α ἰτω χ ιμ λ νκ ὶμ ῶ α π ο λ λ ο ῦ ... There aretwosignificant differences between this reference to komai and Thucydides’description of komai in theArchaeology. First, at 3.94.4 it is anethnos ς. TheAitolians seemto have hadno α μ ώ κ ὰ τ α andnotapolis which is organized κ poleis, except, perhaps, Aigition which, at 3.97.2, is explicitly called a polis; and being the political centre of the Apodotai it may well have been a polis in the constitutional sense as well.44 All other Aitolian settlements seem, in Thucydides’ view, tohave beenkomai. Second, Thucydides carefully points outthat theAitolian komai arewidely scattered, undoubtedly anintentional reference back totheSpartan komai at 1.10.2, which were so close together that they formed oneconurbation. So muchforThucydides andAristotle. If, inother sources, welook forparallels to either Thucydides’ or Aristotle’s description of the relation between polis and kome what dowefind? (a) Apart fromTheseus’synoikism ofAttika, which eventheGreeks themselves placed midway between myth and history, there is only one other example of a

38 See e.g. G. Huxley, Early Sparta (London 1962) 24; M. Clauss, Sparta (München 1983) 117. At Paus. 3.19.6, however, Amyklai is called a kome. 39 Hdt.7.234.2. 40 Hdt. 6.58.1. 41 Xen. Hell. 3.3.10–11. 42 Hdt.7.220, cf. J. Fontenrose, TheDelphic Oracle (Berkeley andLosAngeles 1958) 319: Q 152. 43 Cf. .e.g. K.W. Welwei, Die griechische Polis (Köln 1983) 16; A. Wallace-Hadrill, “Introduction,”in J. Rich andA. Wallace-Hadrill (eds.), City and Country in the Ancient World xiii (London 1991); U.Walter, AnderPolis Teilhaben. Historia Einzelschriften 82 (Stuttgart 1993) 116.

40. 44 See Hansen (supra n. 17) 39–

56

Mogens Herman Hansen

political synoikism, namely Herodotos’description of the Ionians’abortive attempt

to unite andbecome onepolis. After the sack of Sardis by Kyros in ca. 547/6 the Ionian cities convened a meeting in which, according to Herodotos, Thales the philosopher proposed that theIonians should set upa common bouleuterion for all theIonian poleis inTeos, whereby all theother poleis, though kept asurban centres

just as before, would change their status andbecome demes instead of poleis.45

Thus, theemergence of thepolis bya purely political synoikism seems tobean abstraction just like the social contract in later political philosophy. As far as we knowthehistory of Greece, there is noexample of a synoikism, whereby apolis was set up through an agreement between a number of settlements by which each surrendered (some of) its self-government to a setof political institutions convened inoneof thesettlements but,inallother respects, wasleft asit was.Allwell attested examples of polis formation by synoikism seem to have involved a physical synoikism as well as a political one,46 andeven Theseus’synoikism of Attika is in other sources described as a physical synoikism.47 What comes closest to the so-called »political synoikism« is an institution which grewupinthefourth century andflourished intheHellenistic period, namely sympoliteia: the agreement to form a common polis out of a number of poleis. Sometimes theresult wasa complete fusion of thepoleis involved, butsometimes the small poleis merged with the larger onewere allowed to persist in the form or municipalities or even as dependent poleis, so that the sympoliteia was a purely constitutional reform. The earliest example is the4th century sympoliteia between Mantineia andHelisson.48 Furthermore, federations, i.e. koina each composed of a number ofpoleis, mustalso havebeenformed bysomekindof »political synoikism«. TheBoiotian confederation of 446, forexample, undoubtedly emerged bya such an act, onlytheparticipating poleis keptsomuchoftheir self-government tothemselves that, after the»political synoikism«, they persisted aspoleis, though hereafter poleis without autonomia.49 Now, themerging ofpoleis by sympoliteia andthe formation of koina took place ina period well covered byoursources, whereas, inmostcases, the merging of komai to form a polis took place centuries before the sources that mention it. So theabsence of reliable accounts ofpolis formation through a purely political synoikism maybe dueto lack of written sources for the history of Greek society in theearly archaic period. (b) Theπ η intheThucydidean sense, though hardly ςοἰκουμ ν έ μ α α ό τ λ ὰ ιςκ κ ώ ever mentioned in other written sources,50 is attested archaeologically, andlet me

45 Hdt. 1.170.3. 10andpassim. Onpage 10shewrites: »Purely political synoikisms 46 SeeDemand (supra n.27)9– donotformpartof this investigation«. I would suggest instead: »purely political synoikisms are unattested, probably because they arepure fiction.« ἰς νε α σ νοἰκοῦ η μ ὰ κ ώ ὶκ α τ α η νκ δ ά ρ ο ιν(Athens) σπ νπ λ ὴ ό ντ ὲ νμ ῶ το ρ ὶπ α 47 Isoc. 10.35: κ υ ο ν ό ρ χ ῦ ο τ υ σ ο (Theseus) ὥ τ᾽ἔ π α σ ν ν ε ῦ ίη ο ν η ἐπ ὶν α λ ικ α ύ τ ικ τ γ ὼ ν τη ᾽ἐκείν α γ α ν υ ν ὐ σ τ ὸ τα μ εγ ίσ η τ ν τῶ η νἙ ν ίδ λ ω νεἶν λ α ι. 48 SEG 37 340 quoted infra IV page 73. 3, Chambers. See“ TheAutonomous City-State”supra page 35 withn.57. 49 Hell. Oxy. 19.2– 50 Theonlyother attestation inaclassical author is apassage intheAreopagitikos where Isokrates ν ι (the μ ο ε ό λ λ αδιελ seems to suggest that Athens was originally a cluster of villages: ἀ

Kome

57

once again adduce a Boiotian example, namely Thespiai. The Boiotian survey project directed by Anthony Snodgrass andJohn Bintliff has come upwith some very interesting results. In addition to the survey of the countryside between Thespiai and Haliartos an investigation of the two ancient cities was conducted as well, andSnodgrass offers thefollowing interpretation of thefindings: »Within the area of the actual city of Thespiai the pattern is interesting: geometric andearlier archaic sherds areconcentrated notin onelocation, butin three or four, suggesting a cluster of villages rather thananurban nucleus, ina manner recalling Thucydides’ description of Sparta (1.10.1). Haliartos, however, presents a different picture. In the city itself, Geometric settlement is apparently confined to the area of the later acropolis; but outside, there is a string of small sites, stretching from 500 to 2500 metres awayto theeast, whose occupation ineachcase begins notfarfrom700 BC. Clearly, contrasting patterns ofpolis growth could coexist, evenindirectly adjacent cities.«51 (c) Apart from Aristotle’s Politics the only passage in archaic andclassical polis is in Plato’s Greek literature in which we find the sequence oikia – kome – Laws Book 1,52andhere it is notina historical context butina discussion of lack of self-control where Plato opposes theindividual to various types of community and, casually, lists the three types wemeet again in Aristotle’s Politics Book 1. Plato’s ownaccount in Laws of the evolution of society is found in Book 3 andhere the formation of the polis is duly discussed, but not that of the kome.53 Similarly, none of the other famous accounts of the evolution of mankind andcivilization has a word to say about komai, whereas several mention thepolis as one of the major achievements in the development, see the first stasimon in Sophokles’Antigone, Plato’s Republic Book 2, Isokrates’Nikokles, a fragment of a tragedy byMoschion and the prooimion of Philochoros’Atthis.54 (d) If we move from political theory to history we are faced with a similar problem: it is extremely difficult to findexamples that will fit Aristotle’s model: the

51

μ ς, ἐθ ο υ νδ ή ρ νκ ὲχώ ά τ ρ α ὰ ς, τὴ ο εώ ίο δ υ ντὸ ντὸ α μ νβ ν ώ ὰ κ α τ ὲ νμ ὴ νπ ινκ ό λ ancestors) τ 6. Foranattestation in a Hellenistic source see υ... (7.46). Cf. Lévy (supra n. 18) 125– ο τ σ ά ἑκ Diod. 5.6.2, quoted infra n. 96. A.Snodgrass, “Survey Archaeology andtheRural Landscape oftheGreek City,”inO.Murray 1;J. Bintliff & andS. Price (eds.), TheGreek CityfromHomer toAlexander (Oxford 1990) 130– 8. Other early A. Snodgrass, “Mediterranean Survey andthe City,”Antiquity 62 (1988) 65– urban settlements resembling Thespiai include Athens, Argos, Sparta andKoressia onKeos, 89) 8 (1987– see A. Snodgrass, “ TheRural Landscape andits Political Significance,”Opus 6–

64. 58– ρ είτ ὲ ν μ κ ν ὁμ ῶ τω ν α ςἡ ὑ το ν σ το ρ α ῦ ἐ σ τ ίν π ό τε ,ὁ δ ὲἥττω ρ ε ῖςἕκα 52 Pl. Lg. 627A: ἐπ ὰ γ ὴ ειδ φ μ ῶ ε ν ῶ ; ὴφ μ ε νοἰκ ν ῖςἢμ το τα ὑ τὸ τἐ ν α ὐ ῦ ινἔχ τα λ ό ιν ε α ὶπ ν κ ία η ν τ μ εκ α ὶκώ 81A. 53 Pl. Lg. 676B, 677C, 678A andespecially 680D– ὴ κ α λ νξύ ὸ ρ ῳ ςχά ιν τ ν μ ὸμ α ; Pl. Resp. 369C: ε σ τ ιτόλ τ ιςὅ λ π ο ις, ἄ λ ψ ίπ ο 54 Soph. Ant. 370–1: ὑ μ ν ε λ η ἀ ὰ λ ςζ γ ρ ιω η λ λ δ ῶ ά η ν ἀ π ῆ ν ο ν το θ ῦ ὐμό ; Isoc. 3.6: ο ῳ ιν λ ἐ ά π ρ μ ν ε ξἀ χ ςπ ο ῆ ιῶ γ ῷ λ ό τ μ ε ν...; Moschion fr. ε ςεὕρο θ α α έμ ςἐθ κ α ν ὶτέχ ο υ ὶνόμ α κ ν μ ε α ίσ κ λ ιςᾠ ό ε ςπ τ ε ν ό κ α ὶσυνελθ η μ ςο ρ ν ςο γ ω έ ο ἶκ ιςὠ χ υ ο ρ η ὔ ρ τ ε ῖα π ύ α ίν ο ιςεὐρ ελ ν ο ὔ τ εστεγή ἦ ρ 8 (Nauck): οὐδέ γ ὰ ω π 6.6– ν ρ ο ρ γ τε ό ό ευ ρ η σ α ν ιὰ π τ δ τ ὸ ιν ο ν σ π ό λ ὴ ρ ὲπ δ υ τ σ π ό λ ις. Philochoros (FGrHist 328) fr. 2b: ἄ ς ςκοιν ὰ ςε η ν ὶσ α ἰςτα τῆ α ςπλά ιἐ τῆ ν κ ῖνκ ε εσυνελθ τ ςτό τα ν νζῶ η δ ά ρ ςκ α ὶσπ μ ά δ ο α ν ο ν . α σ τη έσ ν μ ετα ὐ ο ἰκ ν ο θ ε σ ή ις, ὅ ε

58

Mogens Herman Hansen

emergence of thepolis bya physical synoikism of a number of neighbouring komai. In principle there arethefollowing four forms of synoikism:55 (a) apolis is created by merging a number of komai ordemoi (Aristotle’s model); (b) apolis is created by merging two or more poleis (e.g. Rhodes in 408/7);56 (c) a polis is reinforced by absorbing oneor more neighbouring komai ordemoi (e.g. Thebes in ca. 431);57 (d) apolis is reinforced byabsorbing oneormore neighbouring poleis (e.g. Olynthos in 432 or Arkadian Orchomenos in ca. 350).58 A possible fifth form is a combination

of (c) and (d).

The examples which modern historians traditionally subsume under (a) are: Elis, Mantinea, Tegea, Heraia, Aigion, Patrai andDyme, i.e. all thepoleis listed by Strabo in connection with his account of the synoikism of Elis.59 But the three Achaian poleis are problematic because, apart from Strabo, we do not know anything about thesynoikism, andthefourothers donotfit themodel. Letmefocus first on Elis. The traditional view is that, with the synoikism of 471/0, Elis became a polis bothinthepolitical andthetopographical sense of theterm: a newpolis wascreated

by merging a number of villages.60 Oursources for thesynoikism are (1) Diodorus λ ε ῖο ιμ ὲ νπ ςπ λ ν(472/1 B.C.) Ἠ είο π ὶδ ςκ 11.54.1: ἐ ὲτού τω ρ ό υ α ὶ μικ λ ὰ ε ις κ ὴ η νὀνομ σ ίσ α ντ θ α μ ζο η νἮ έ ία νσυνῳ ν ςε λ ο ἰκ ιν ο ἰςμ τ ν ῦ ε . (2) Strabo 8.3.2: ’Ὅ ν οκ ρ τ α ω θ π ο τισ ν ᾠ ὔ ἔκ ῦ ν η π ,ἀ ό λ ὲἡ κ ιςο Ἦ λ ιςδ λ λ μ ε ῖτο... η δ ὸ ν μ ρ α κ ω ώ χ ᾽ἡ ςτ νν ὴ ό λ ινἮ ῦ νπ λ ιν ε ,μ τ τ ὰ ὰΠ σ νἐ ερ ικ ψ τ ὲδ α έπ ο εσυνῆλ θ ὰ ὀ ,ἐ κπολ ν λ ῶ ςκ μ ιδ ὲ(Oxylos) ὡ ω γ τα α ςανθρώ ν ε ή ὶτο . (3) Pausanias 5.4.3: λ π έ ςἐ δ ο υ κτῶ ὺ ν κ εσ α ν φ εσ τή , κα ὴ τελ ν ςτ ο λ ὺ ἀ π ςο θ ὐ π ό ιν ε λ ῖν ἔπ υ ο εισ ν μ ,ὅ ν ῶ ιτο ε σ ῦ ἐ ο τείχ κ ω ο ν εσ είζ ρ ιμ τέ α ν ο ν α κ α ὶεὐδα ὴ ν η ρ ω νμ ν ετ τό ςτ ὰ η ἄ ή λ θ λ ε α ιτ ἀ π εοἰκ ἐ έφ κ α ὶπλ . Ἦ λ ιν The settlements merged to create the new city of Elis are called poleis by Diodorus, demoi by Strabo and komai by Pausanias. In this context, however, Strabo uses demos synonymously withkomeasis apparent fromhisstatement a few ᾠ κ ε ῖτ ο . If (as usual) Diodoros is brushed aside as an ν η δ ὸ μ ρ ω α κ χ ώ lines earlier: ἡ unreliable source, we have, apparently, a late, but perfect example of Aristotle’s contention that apolis is anaggregation of komai created by a synoikism.61 Epigraphical evidence, however, points to a different interpretation of the synoikism. During their excavations of Olympia in the late 1870s the Germans found a number of bronze placques inscribed with decrees written in the Elean 60; Swoboda, writing about komai, concentrates on (a) and(c) 55 Swoboda (supra n.18) 956– which he further subdivides by distinguishing between whether the entire population in the komai isremoved tothepolis orwhether thekomai areallowed topersist asvillages sidebyside withthepolis. Demand (supra n.27). 56 Diod. 13.75.1; Strabo 14.2.9, 11. 5. 57 Hell. Oxy. 20.3, Chambers, see Demand (supra n. 27) 83– 58 Olynthos: Thuc. 1.58.2. Orchomenos: IG V 2 343 = Staatsverträge no.297. 59 Strab. 8.3.2. 60 Swoboda (supra n. 18) 956; M.Moggi, I sinecismi interstatali greci, I: Dalle origini al 338 a.c. (Pisa 1976) 131ff repeated (with a warning) byJ.K. Davies in CAHV (1992) 28. 61 Demand (supra n. 27) 64 refers to »the synoikized communities« (myitalics) andthus refrains fromtaking a position ontheir status.

Kome

59

alphabet anddialect.62 Their chronology haslong been disputed butthanks to L.H. Jeffery’s careful study most of them can nowbe dated within a period from the late 6th to the early 5thcentury andare thus earlier than the synoikism of 471. Several of these decrees are passed by »the Eleians« andthere are references to a popular assembly, a council of five hundred andvarious boards of magistrates. If we can trust theeditor’s restoration Elis is even referred to as a π ό λ ις.63Nodoubt Elis was apolis in thepolitical sense of theterm before 471, andtheeffect of the synoikism in471 wasapparently tomakeit apolis inthetopographical sense aswell: tocreate a large urban centre forthepolitical institutions which, previously, mayhave metin

Olympia.

This is nowtheprevailing view among ancient historians64 but, in myopinion, of Elis have bought us onestep further. Though Elis wasanunwalled city as late as 400 B.C.65 it wasnevertheless anurban centre of some importance more than a century before: »Inthearchaic period thecity wasextended totheSW. At that time the temple of Athena was probably erected on the acropolis (Paus. 6.26.3). Numerous painted terracotta simas andstone architectural fragments indicate the existence at that time of many monumental structures.«66 According to this description Elis wasapolis in thetopographical sense before 500 B.C. Next, there is doubt about thestatus ofthesettlements involved inthesynoikism of Elis. As stated above, the prevailing view is that they were villages (komai or ρ α ὶπ ό ικ λ ις, and his account should ε demoi), but Diodoros describe them as μ perhaps be taken more seriously than is often the case, because it seems to be supported bysome early documents inscribed onbronze plaques. Some of these are treaties between or laws passed by small communities which in some cases are known from other sources, e.g. Skillous;67 butseveral communities areonly known ρ ιο ι.69 δ ά λ α ο ι, the Μ ιτ α ν to us from the placques, i.e. the Ἄ ετά π ιο ι,68andthe Χ The reason maywell be that they disappeared in connection with the synoikism of ιstipulates ιο π ετά ιandthe Μ ο 471. Furthermore, the treaty between the Ἄ ιτ α ν »friendship« for a period of fifty years, andit is at least notnormal for a village to conclude a fifty years’peace with another village. Thepresumption is that, before the synoikism, they were small poleis exactly as stated by Diodoros. Thus by the synoikism of 471 Elis was enlarged by absorbing a number of neighbouring

excavations

21. 62 IvO 1– 63 ForElis asa polis before thesynoikism of 471 seeIvO7 = LSAG220 no.5 (ca. 500): mention ); IvO3 = LSAG 220 no.9 ν η θ ύ ω ςπλ ο μ ᾶ of five hundred anda popular assembly (δ α λ ώ of a β ]. ς ιο [λ ό ςπ ν , mention ofτᾶ ω ςπλαθύ ο (ca. 475): References toβ μ α λ andtoδα ώ 64 Stated already by V. Ehrenberg in his article: “When Didthe Polis Rise?,”JHS 57 (1937) reprinted in Polis undImperium (Zürich andStuttgart 1965) 89 note 1; nowargued at great length andeven more convincingly by U. Walter, Ander Polis Teilhaben. Historia Einzel25 andfurther developed in M.H. Hansen & T. Fischer-Hansen, schriften 82 (1993) 116– “ Monumental Political Architecture in Archaic andClassical Greek Poleis,”in Whitehead (supra n. 1) 86– 9. 65 Xen. Hell. 3.2.27. 66 N. Yalouris, ‘Elis,’in ThePrinceton Encyclopaedia of Classical Sites (Princeton 1976) 299. 25). 67 IvO 16 = LSAG 220 no. 17 (450– 50). 68 IvO 10 = LSAG 220 no. 12 (ca. 475–

60

Mogens Herman Hansen

communities, some of which were probably small poleis. After all, theterm μικ ρ α ὶ ιςusedbyDiodorus inhisaccount of thesynoikism of Elis maycontain a core λ ε ό π of truth andI will return to this problem in section IV 3 (b), pages 73– 6.

My final

observation concerns the period after 471. The traditional view of the synoikism is that henceforth Elis was the only polis within the region.70 The whole of Elis wastransformed into onelarge city-state just like Athens wastheonly CityState in Attika. But such an interpretation of the politicial structure of Elis, the region, does not square with the sources wehave. Xenophon, for example, states explicitly that, in his age, Skillous was a polis both in the urban andin the political sense of theterm.71 Theplurality ofpoleis persisted after thesynoikism. There is no indication thattheElean poleis wereeverunited ina federation, andastheevidence is, Elis seems to have been a region dominated by onelarge polis, surrounded by a number of small poleis which depended onthe large polis but hadnot been reduced to mere municipalities. The proper parallel is not Attika, but rather Lakonia, composed of onecentral polis, namely Sparta, surrounded bya number of perioikic poleis. This analogy is further corroborated by the fact that in Xenophon’s account λ ό ις(3.2.23). ε ςπ theElean poleis arecalled π ε ίδ ικ ιο ερ Myconclusion is that the comparison between Aristotle’s Politics Book 1 and the synoikism of Elis breaks down completely: both in the political and in the topographical sense ofthetermElis wasapolis before thesynoikism; thesettlements merged with Elis in 471 were probably poleis andnotjust komai; anda number of dependent poleis persisted after the synoikism. The synoikism of Elis in 471 turns out to be of the same kind as e.g. the synoikism of Olynthos in 432: a polis already in existence is reinforced by absorbing a number of other small poleis. A study of theArkadian cities leads tothesameconclusion. Themostprobable 59. The synoikism of Tegea seems to date of the synoikism of Mantineia is 464– belong in the same period, whereas that of Heraia took place either ca. 490 B.C. or ca. 370 B.C.72 Butthere is noreason to suppose thatthese three cities became poleis by the synoikisms. Onthe contrary oursources indicate that the Mantineians, the Tegeatai and the Heraians had existed as political communities, or to be more precise: as poleis, for more than a century prior to the synoikisms.73 But whether they hadanurban centre toobefore thesynoikisms is a mootpoint which I dare not delve into here.74

475). IvO 11 = LSAG 220 no. 8 (500– Moggi (supra n. 60) 161; Osborne (supra n. 24) 124.; Walter (supra n. 64) 121. Xen. Hell. 6.5.2. Forother poleis inElis see 3.2.23, 30; 7.1.26; 7.4.14. Demand (supra n. 27) 66. Tegea: Paus. 10.7.7 (6th cent.); Lazzarini 811, 887 (early 5th cent.); Mantineia: Hdt. 4.161.2 (ca. 600B.C.); Paus. 6.9.9 (= Moretti no. 163) (500 B.C.); Heraia: Syll3 9 (6thcent.); Lazzarini TheArchaic Coinage of Arcadian Heraea,”MN16 975 (early 5thcent.). SeeR.T. Williams, “ 12. (1970) 1– 74 OnArkadian poleis andtheir synoikisms, seetheforthcoming study byTh. Heine Nielsen to whomI owesomeof thereferences inthepreceding note.

69 70 71 72 73

Kome

61

But there is another example –never adduced in this context –which fits Aristotle’s model much better than those discussed above, namely Kassope in the northwestern part of Greece where polis-formation took place as late as in the mid fourth century. In the Periplous erroneously ascribed to Skylax and traditionally 60 three sections aredevoted to theEpeirote tribes: the dated to theperiod ca. 380– Thesprotoi (30), the Kassopoi (31) andthe Molottoi (32). In each chapter the author states that thepeople (ἔ ς) lived κ ο ν α ς, andnopolis is mentioned between θ τ ὰ μ κ α ώ Korkyra described at 29 as a π η λ ν λ ίςandthe Amphilochaian Argos called ό ις Ἑ λ π ό λ ιςat 33. But in the Epidaurian list of theorodokoi of, probably, 356 B.C. Kassope is recorded between Pandosia, a colony of Elis, andthe Thesprotoi, still recorded as a tribe.75 The shift from the Kassopoi in Ps. Skylax to Kassope in the list of theorodokoi strongly suggests that Kassope was by now a polis and that is confirmed by therecent excavations of thesite. They have revealed that thecity of Kassope must have been founded around the midfourth century;76 the excavators consider the circuit to belong to the period immediately after the synoikism and suggest a date around 350 B.C.77 So here at last is an example of the emergence of a polis by a synoikism of komai, themodel advocated by Aristotle in Book Oneof thePolitics. IV. Kome in a Contemporary Context

A kome was often just a village: anything from a cluster of houses to a small conurbation inhabited by families whose principal occupation wasto cultivate the surrounding fields. In this sense a kome may or may not have had a political organization of its own, andit may or maynot have been a unit in the political organization of thepolis or region to which it belonged. 1. Kome in the Topographical Sense

A kome wasa small anda polis wasa larger urban centre, as is apparant e.g. from ί α Isokrates’ rather arrogant comparison between Athens andother Greek cities: κ φ α σ ινο ἱμ η νπ ὲ ς, κ ό ςδ λ ύ τ ντοιοῦ α ὶ ιν α μ νεἶν ι, τα , τα η α ςκώ ν α λ τ ο ι μό ᾽ἄλ ρ γ ο εύ σ α ρ ο εσ α ι.78But the difference θ ςπ ο δ ς Ἑλλά ὴ νἄ υτ ῆ σ τ ςα να δικ ὐ τ α ίω between a kome and a polis was not just a matter of size, it was principally a question of political status, andtheoldview, based exclusively onwritten sources, that a komewasunfortified whereas apolis was(mostly) a walled conurbation 79has 6. 75 IG IV295.24– 76 For a survey seeW. Hoepfner & E.-L. Schwandner, Haus undStadt imklassischen Griechen140. land (München 1986) 75– 77 Hoepfner & Schwandner (supra n.76) 85 withnote 166. 4. 78 Isoc. 15.299. Forthereverse idea cf. Arist. Cat. 5b22– 79 Swoboda (supra n. 18) 951 citing Thuc. 1.5.1; 2.80.8; 3.94.4; 4.43.1. The view is repeated by Kolb (supra n. 20) 59.

62

Mogens Herman Hansen

been refuted by archaeological evidence from e.g. Askra andAigosthena. Askra wasa komeintheterritory ofThespiai.80 It hasbeenlocated andcarefully surveyed,81 andit seems to have been a settlement comparable to e.g. Mykalessos 82which in

the classical period wasa polis.83 In ca. 300 B.C. Aigosthena, today perhaps the most impressive of all classical fortifications, was a kome in Megaris,84 but a century later it was a polis.85 It hadchanged its status but as a town it was probably of thesame size in 300 asin200 B.C. Aninspection of thesources shows that, even in its topographical sense, thewordpolis wasreserved for a townthat wasalso the political centre of a polis in the sense of city-state,86 whereas a kome was a municipality or sometimes just a village, namely in those cases where the komai situated intheterritory of apolis werenotintegrated inthepolitical organization of thepolis in question. Down to the end of the classical period the Greeks seem to have been fairly consistent in distinguishing between polis and kome in the topographical sense. There are in fact only twoexamples of an author using kome andpolis about the same settlement, namely Herodotos whoat7.176.2 describes Anthele asapolis but calls it a kome at 7.200.2, andagain Alpenos is called a kome at 7.176.5 butapolis at 7.216.1. Admittedly, Thucydides tells us about a kome in Lokris by name Polis;87 but this apparent oxymoron is not necessarily an instance of inconsistency in terminology. Here the name Polis is probably used synonymously with the noun akropolis in its original sense of »stronghold« and applied to a small settlement situated on aneminence. Anobvious parallel is the small settlement Ptolis ca. one km.north of Mantineia, situated onaneminence andundoubtedly oneof the small sites merged with Mantineia in the archaic or early classical synoikism.88 Occasionally the wordkomecould be used about a dependent polis, butthat will be discussed inpart 3(b) below. In theHellenistic andRoman periods, ontheother hand, theterminology fluctuated anda settlement called polis byoneauthor is often called komeby another andvice versa, cf. thediscussion of Boiotia infra 2(d). The best example of komai which arejust villages andhave virtually no part to play intheorganization of their polis is undoubtedly thetwelve komai inMagnesia, theutopia described inPlato’s Laws. Magnesia’s territory (chora) is subdivided into twelve parts (dodekatemoria), each centered on a village (kome)89 The komai are villages, notcivic subdivisions in the strict sense. Citizens areregistered in tribes, 80 Hes. Op.640; Ephoros (FGrHist 70) fr. 1, both quoted infra 2(b). 81 A.M. Snodgrass, ‘TheSite of Askra,’inG. Argoud & P. Roesch (eds.), LaBéotie antique (Paris 95. 1986) 87– 5. 82 Fossey (supra n. 16) 80– 21. 30, seeHansen (supra n. 3) 18– 83 Thuc. 7.29– 84 IG VII 1 (after 306 B.C.). 85 IG VII 207.4 (between 223 and201 b.c.). 45. 86 Hansen (supra n. 3) Appendix I 39– 87 Thuc. 3.101.2 quoted infra 2(b). 88 Ptolis: Paus. 8.12.7; synoikism ofMantineia: Strab. 8.3.2. Fortheidentification andlocation of ς (e.g. IG IX2 186.13, 206/5 ύ ιε λ ο Ptolis see RE XIV.2 1305. The attestation of the ethnic Π B.C.) indicates that, like Aigosthena, Polis hadbecome apolis intheHellenistic Period. 22. E; see M. Piérart, Platon et la cité grecque (Bruxelles 1973) 18– 89 Pl. Lg. 848C–

Kome

63

phratries, 90 not in komai; demes and butthe affairs of the kome are not regulated by an assembly of the kometai; there are no komarchoi andthe kome is ruled by the

centrally elected agronomoi andtheir assistants;91 the only public power vested in the komai is judicial: agronomoi accused of misuse of their powers are brought before theneighbours andvillagers,92 andkometai andphyletai shall sitinjudgement together and form a kind of middle tier between the arbitrators (below) and the

central courts (above).93 Apart from Thucydides’ Archaeology and Aristotle’s Politics Book 1, Plato’s description inLawsof thefoundation of Magnesia is theonly account inarchaic and classical Greek literature that explicitly discusses the term kome opposed to the other words used about settlements: chora, asty, polis and akropolis. The other evidence is so scattered anddefective that thebest I can dois to offer a list of our sources.

2.

References toKomai inArchaic andClassical Sources and(Retrospectively) some Later Sources

in

(a) Passages in which Komai are Referred to in General as Subdivisions of a Region

or of a Polis

ρἐ ῖς σ π ντο ε Ainis. Theophr. Nomoi fr. 650 (= Stob. Antol. 4.2.20 p. 129): ... ὥ ελ εύ ο υ ς σ ρ ιγ ὰ ν ο , ἐά ῦἈπόλλω ὶτο ὲ ντιςμ ν π ο ἐ ἰκ ιν ε ύ ρ ι, θ π ίω ν ν ἰν α .94κ Α τ ίη ίη ςοἰκ ὐ τὸ ε ῖ *** κ ςᾗα η ιν μ ε ύ ςκώ ν ῆ ρ ίο νἐ μ α ίο π νδ τ ο υ ἰτ ὲχω , ἐὰ ῦἐπ ικ ω ὶ ὀμ α ς> ἐγγραφ ὴ νὠ ν θ εῖσ α ι ν ῆ , ἦμ ντριῶ ςιο [ι]κ τ εΠ ἰς