95 THESES ABOUT THE GREEK POLIS IN THE ARCHAIC AND CLASSICAL PERIODS

A Report on the Results Obtained by the Copenhagen Polis Centre in the Period 1993-2003

by  Mogens Herman Hansen

A brief survey of the more important results of the research conducted in the Copenhagen Polis Centre is best presented as a number of theses with references to the publication(s) in which the issues at stake have been discussed and the theses have been advanced. The article is intended as a vade mecum for historians to help them to find their way through the numerous publications of the Polis Centre. The theses are stated rather bluntly, without any discussion or argumentation which, however, can be found in the literature referred to after each thesis. The theses represent either novel views or controversial topics to which we believe we have made a significant contribution. References are to the following publications:

The Acts Series

CPCActs 1 = M.H. Hansen (ed.), The Ancient Greek City-State. Acts of the Copenhagen Polis Centre 1. Det Kongelige Danske Videnskabernes Selskab, Historisk-filosofiske Meddelelser 67 (Copenhagen 1993).

CPCActs 2 = M.H. Hansen (ed.), Sources for The Ancient Greek City-State. Acts of the Copenhagen Polis Centre 2. Det Kongelige Danske Videnskabernes Selskab, Historisk-filosofiske Meddelelser 72 (Copenhagen 1995).

CPCActs 3 = M.H. Hansen (ed.), Introduction to an Inventory of Poleis. Acts of the Copenhagen Polis Centre 3. Det Kongelige Danske Videnskabernes Selskab, Historisk-filosofiske Meddelelser 74 (Copenhagen 1996).

CPCActs 4 = M.H. Hansen (ed.), The Polis as an Urban Centre and as a Political Community. Acts of the Copenhagen Polis Centre 4. Det Kongelige Danske Videnskabernes Selskab, Historisk-filosofiske Meddelelser 75 (Copenhagen 1997).

CPCActs 5 = M.H. Hansen, Polis and City-State. An Ancient Concept and its Modern Equivalent. Acts of the Copenhagen Polis Centre 5. Det Kongelige Danske Videnskabernes Selskab, Historisk-filosofiske Meddelelser 76 (Copenhagen 1998).

CPCActs 6 = T. Heine Nielsen & J. Roy (eds.), Defining Ancient Arkadia. Acts of the Copenhagen Polis Centre 6. Det Kongelige Danske Videnskabernes Selskab, Historisk-filosofiske Meddelelser 78 (Copenhagen 1999).

The Papers Series

CPCPapers 1 = D. Whitehead (ed.), From Political Architecture to Stephanus Byzantius. Papers from the Copenhagen Polis Centre 1. Historia Einzelschriften 87 (Stuttgart 1994).

CPCPapers 2 = M.H. Hansen and K. Raaflaub (eds.), Studies in the Ancient Greek Polis. Papers from the Copenhagen Polis Centre 2. Historia Einzelschriften 95 (Stuttgart 1995).

CPCPapers 3 = M.H. Hansen and K. Raaflaub (eds.), More Studies in the Ancient Greek Polis. Papers from the Copenhagen Polis Centre 3. Historia Einzelschriften 108 (Stuttgart 1996).

CPCPapers 4 = T. Heine Nielsen (ed.), Yet More Studies in the Ancient Greek Polis. Papers from the Copenhagen Polis Centre 4. Historia Einzelschriften 117 (Stuttgart 1997).

CPCPapers 5 = P. Flensted-Jensen (ed.), Further Studies in the Ancient Greek Polis. Papers from the Copenhagen Polis Centre 5. Historia Einzelschriften 138 (Stuttgart 1999).

CPCPapers 6 = T. Heine Nielsen (ed.), Even More Studies in the Ancient Greek Polis. Papers from the Copenhagen Polis Centre 6. Historia Einzelschriften 162 (Stuttgart 2002).

CPCPapers 7 = T. Heine Nielsen (ed.), Once Again: Studies in the Ancient Greek Polis. Papers from the Copenhagen Polis Centre 7. Historia Einzelschriften 180 (Stuttgart 2004).

The City-State Series

30 CSC = M.H. Hansen (ed.), A Comparative Study of Thirty City-State Cultures, Det Kongelige Danske Videnskabernes Selskab, Historisk-filosofiske Skrifter 21 (Copenhagen 2000).

6 CSC = M.H. Hansen (ed.), A Comparative Study of Six City-State Cultures, Det Kongelige Danske Videnskabernes Selskab, Historisk-filosofiske Skrifter 27 (Copenhagen 2002).

The Inventory

Inv. = M.H. Hansen and T. Heine Nielsen (eds.), An Inventory of Archaic and Classical Poleis. (Oxford 2004).

Other Monographs

P&P = P. Flensted-Jensen, T. Heine Nielsen, L. Rubinstein (eds.), Polis and Politics. Studies in Ancient Greek History Presented to Mogen Herman Hansen on his Sixtieth Birthday, August 20, 2000 (Copenhagen 2000).

Ark. = T. Heine Nielsen, Arkadia and its Poleis in the Archaic and Classical Periods. Hypomnemata 140 (Göttingen 2002).

Other Articles

No. 1 = M.H. Hansen, "Aristotle's Two Complementary Views of the Greek Polis", in R. Wallace & E. Harris (eds.), Transitions to Empire (Norman 1996) 195-210.

No. 2 = D. Whitehead, "Polis-Toponyms as Personal Entities (in Thucydides and Elsewhere)", MH 53 (1996) 1-11.

No. 3 = M.H. Hansen, "Emporion. A Study of the Use and Meaning of the Term in the Archaic and Classical Periods", in G. Tsetskhladze (ed.), Greek Colonisation 1-2 (Leiden 2003) 1: forthcoming.

No. 4 = M.H. Hansen, "Belonging in a Political Context", forthcoming in the Leicester-Nottingham Studies in Ancient Society (1998).

The Concepts of City-State and City-State Culture

(1) We have introduced and defined "city-state culture" as a concept to be distinguished from the concept of city-state. By a city-state culture we understand the civilisation of a large region whose inhabitants share language (or a lingua franca), religion, traditions etc.; the region usually constitutes an ethnic, a social and an economic entity, but politically it is broken up into a large number of small communities each centred on a city. To be excluded from the concept of city-state culture are isolated city-states which do not form an integrated part of a city-state culture (e.g. Ragusa 1358-1700, Macao 1557-1967, Andorra today), as well as cities which form a cultural and economic network but are capitals of large "territorial" states, and not small city-centred self-governing political entities (The Harappan cities in the Indus valley ca. 2600-1900 B.C.) (30 CSC: 16-17, 609-616).

(2) We have (so far) traced thirty-five different city-state cultures in world history ranging from the Mixtec in Mesoamerica to the Taklamakan city-states in central Asia and from the Viking city-states in Ireland to the Swahili stone towns in Kenya and Tanzania. City-state cultures are attested as early as the late fourth millennium B.C. in Mesopotamia and as late as the late 19th century A.D. in West Africa (30 CSC: 20-22; 6 CSC: 8-12).

(3) City-state cultures emerge in one of the three following ways: (3a) In a period of demographic and economic upsurge, urbanisation and state formation take place simultaneously or in close sequence. The city-state period is preceded by a pre-state period. The formation of city-states is gradual and often imperceptible (the Greek polis). (3b) Colonisation of a region takes the form of the foundation of a number of city-states (the Aztec city-states ca. 1200 A.D.). (3c) In a period of decline, an urbanised macro-state disintegrates in such a way that each of its major urban centres becomes a city-state (the Chinese city-states in the Spring-and-Autumn period) (30 CSC: 16 with n. 64).

(4) The emergence of a city-state culture by devolution (3c) is much more common than usually believed and that disproves the still common evolutionist model that early cities were all city-states and that the territorial state dotted with cities is a later phase of a universal development (6 CSC: 12-13).

(5) Attempts to create larger political units, either peacefully or by conquest, often lead to small city-states being swallowed up by larger city-states (Florence conquering other city-states in Tuscany). But more often such attempts take the form of hegemonic leagues (The Delian League), or federations (the Lykian, the Swiss and the Dutch federations), or "mini-empires" consisting of one large dominant city-state and a number of smaller dependent city-states (Oyo and later Ibadan in Yorubaland) (30 CSC: 17, 612-613; 6 CSC: 14-15).

(6) When, occasionally, one city-state succeeds in long-term conquest of all the others, the city-state structure usually persists so that the result is a large "capital" in control of an empire built up of dependent city-states (Rome, Tenochtitlan) (30 CSC: 17, 613-614).

(7) Thus, the city-states of a city-state culture are not necessarily "peer polities", but can be hierarchically organised systems of polities, of which some are hegemonic, some independent, and some dependencies (30 CSC: 17, 606).

(8) Dependent city-states are self-governing communities, but as regards foreign policy or defence, they have either limited independence or no independence at all, and usually they have to pay tribute and provide troops to a neighbouring overlord or a hegemonic city-state within the region, or a central government in regions in which the city-states are united in a federation (30 CSC: 17, 608; 6 CSC: 14).

(9) A city-state culture ceases to exist either (9a) by the (temporary) disappearance of the urban centres which, of course, is associated with the disappearance of the political structure of the cities as well (the Maya city-states ca. 900 A.D.) or (9b) by being conquered by a neighbouring Great Power: the citystates are transformed into cities, sometimes abruptly (the Sumerian city-states when conquered by Sargon of Akkad in ca. 2330 B.C.), but sometimes the citystates are allowed to persist for some time, and the transformation from city-states to cities is slow and gradual (Mixtec and Aztec city-states for some generations after the Spanish conquest) (30 CSC: 17 with nn. 71-73, 610-611; 6 CSC: 13).

(10) City-state cultures often appear in neighbouring regions, and in some cases one can almost speak of clusters of city-state cultures (Mesoamerica, West Africa, the Fertile Crescent) (30 CSC: 17 with n. 74; 6 CSC: 13-14).

(11) In some cases a region is split up into city-states only once in history, but there are several examples of regions which have been a city-state culture at least twice and sometimes three times in world history (Mesopotamia, Toscany) (30 CSC: 17 with n. 75; 6 CSC: 13-14).

(12) By contrast with a modern nation state, the population of a city-state has a political identity which is different from its ethnic identity. It shares its ethnic identity (language, culture, religion, history, etc.) with a number of other city-states, whereas its sense of political identity (including patriotism) is primarily centred on the city-state itself rather than on smaller entities (municipalities) or larger entities (ethnically based political organisations, federations, monarchies) (30 CSC: 18).

(13) The name of a city-state is either identical with the name of its major urban centre, or it is an ethnic derived from the name of the urban centre. The name of a territorial macro-state is usually identical with the name of the country (30 CSC: 18 with n. 80; 6 CSC: 14).

(14) In many small city-states the majority of the population lives in the town. In middle-sized and large city-states a substantial part and sometimes even the majority of the population may have been settled in the hinterland, either nucleated in villages or dispersed in homestead farms. But in all citystates the population of the urban centre constitutes a much higher percentage of the total population than in any other type of pre-industrial community (30 CSC: 18, 32, 614; 6 CSC: 15-16).

(15) Small city-states may have what is essentially a subsistence economy; but the urban centres of middle-sized and large city-states are cities in the Weberian (historical) sense of this term. Although Ackerbürger (see infra no. 24) may have constituted part of the population even of large city-states, the cities of middle-sized and large city-states were centres whose inhabitants acquired an essential part of their necessities in the local market. Thus, specialisation of function and division of labour are essential aspects of the economy of a city-state (30 CSC: 18, 602-604).

(16) A city-state is a self-governing polity, but not necessarily an "independent and autonomous state". It suffices that a city-state is a legislative, administrative and judicial unit and, by and large, possesses what in modern terms is called "internal sovereignty", i.e. a government which enforces a legal order within a territory over a population. Many city-states are independent, many others possess some of the powers that are commonly subsumed under the concept of "external sovereignty". But external sovereignty (= independence or autonomy) is not a necessary requirement for being a city-state. Nothing prevents a city-state from being a tributary polity or a dependency of another citystate, or of a federal central government, or of a monarch. Even (some) interference with a city-state's internal sovereignty does not necessarily undermine its identity as a city-state. A central criterion is the citizens' own perception that their city is their fatherland (no. 29 infra) and constitutes a polity (30 CSC: 18, 608-609; 6 CSC: 14).

(17) The city-state is a highly institutionalised form of political community and the percentage of the population directly involved in the government of the community is larger than in any other type of state in world history at least until the 20th century. That applies not only to republican city-states; even in city-states ruled by monarchs a high percentage of the population participates in the running of the political institutions (30 CSC: 17, 611-612).

(18) In his description of the ideal polis Aristotle emphasises economic self-sufficiency (autarkeia) as an unobtainable but desirable aspect of the Hellenic polis. Undoubtedly following Aristotle, it has become customary to include economic self-sufficiency among the defining characteristics of the city-state. City-state cultures, however, are characterised by urbanisation which entails specialisation of function, division of labour and trade – not only local trade but also trade with other city-states in the region as well as with states outside the region. Thus, compared with other types of early state formation, the city-state is characterised by its lack of economic self-sufficiency and by a high degree of economic interaction with its neighbours (30 CSC: 18-19, 616 n. 15).

The Concept of Polis

(19) In our study of the ancient Greek city-state culture we distinguish between (a) the ancient Greeks' understanding of their own settlement pattern and political system, and (b) the modern historians' analysis of the ancient Greek settlement pattern and political system. When investigating the Greeks' perception of their social and political organisation we focus upon the Greek term polis as attested in Archaic and Classical sources, whereas we restrict the concept of the city-state to our modern analysis of ancient Greek society. The analysis is thrown into perspective by comparing the concept of polis as found in the sources with the concept of city-state as found in modern historical accounts. Many modern historians, however, are not sufficiently aware of the distinction between (a) and (b). They use the term polis synonymously with the term city-state and accordingly they erroneously transfer characteristics of the more general concept of city-state to the ancient concept of polis (CPCActs 3: 7-34, especially 7-9).

(20) In a study of the history of the modern terminology we have discovered that the English term city-state was probably coined in 1885 as a rendering of the German term Stadtstaat. The German term Stadtstaat was probably coined in 1842 as a rendering of the Danish term bystat, coined in 1840 by Johan Nikolaj Madvig. The terms bystat and Stadtstaat were not originally connected with the Greek concept of polis, but first developed and used in connection with the Roman concept of civitas, and only later transferred to studies of the Greek polis and the Italian città. The French term cité-état and the Italian term città-stato are both derived from Stadtstaat and/or city-state and neither is attested earlier than the 20th century (CPCPapers 1: 19-22; CPCActs 5: 15-16).

(21) When studying the concept of polis, we have conducted two separate investigations, one of the intension (meaning) and one of the extension (coverage) of the term. (a) All attestations of the term polis and its derivatives are collected and analysed in order to determine the intension of the term (what is a polis?). (b) All attestations of the term polis applied to a named and identifiable community are collected in order to determine the extension of the term (the total number of Archaic and Classical poleis). This second investigation is the foundation of our inventory of all communities that are actually called polis by the Greeks. (c) The results of (a) and (b) are compared in order to describe and define the concept of polis in the Archaic and Classical Periods (CPCActs 3: 7-34, especially 9-14). The inventory will be the first documented survey of the number and identity of ancient Greek poleis in the Archaic and Classical periods, and it will enable historians to compare e.g. Plato's and Aristotle's general view of the polis and their ideas about what a polis ought to be with what a polis actually was (CPCPapers 1: 14-15; CPCActs 3: 55-62, 73-116; Inv.).

(22) Our investigation covers the Archaic and Classical periods, i.e. the period from ca. 650 B.C. (when our sources begin) to ca. 323 B.C. (when Alexander's conquest of the East led to the foundation of several hundred new poleis). Our investigation is based on contemporary sources, and later sources are used only if their information is explicitly retrospective and, accordingly, based on lost Archaic and Classical sources. We thereby avoid the all too common anachronistic use of, e.g., Pausanias' understanding of what a polis is (10.4.1), or Strabo's classification of named urban settlements as either poleis or komai (CPCPapers 1: 14-15; CPCActs 2: 326-344; CPCActs 3: 55-62,73-116).

(23) The inventory of poleis comprises 1,035 entries and covers the Archaic and Classical periods down to 323 B.C. Given the lack of sources, especially for remote regions, it can be presumed that there were altogether ca. 1,500 poleis, but not all at the same time. Throughout the period new poleis emerged while others disappeared. In Euboia around 400 B.C. the number of poleis was reduced from ca. twelve to four, but in the same period numerous new poleis were founded in the Adriatic by Dionysios I, and in the fourth century many indigenous communities became Hellenic poleis, e.g. in Sicily and Karia. A list of poleis in the year 400 B.C. comes to 850, and the presumption is that there were never, in any given year, more than ca. 1,200 poleis at maximum (Inv.).

(24) Pace Moses Finley, Max Weber's antike Stadt as an ideal type is a very valuable model when applied to the Greek polis of the fifth and fourth centuries. Because of his "primitivistic" view of the ancient economy it was Finley who argued that Weber's Idealtypus or "model" of die antike Stadt did not fit the over one thousand middle-sized and small poleis. The investigations conducted in the Polis Centre indicate that Weber's ideal type of city does fit even small poleis: the Classical polis (in the sense of state) was a self-governing (but not necessarily independent) political community invariably centred on a polis (in the sense of town). Many poleis were so big that it was impossible for all inhabitants to know one another, whereas in most cases the number of adult male full citizens was small enough to allow the polis (in the political sense) to be a face-to-face society. A considerable number of townsmen were farmers who had their home in the city but their fields in the hinterland (Ackerbürger). The town was enclosed by a defence circuit and centred on an agora in which the inhabitants supplied themselves with a substantial part of the necessities of life, often produced in the hinterland but sometimes imported from abroad (CPCActs 4: 32-54; 30 CSC: 156-160, 602-604); CPCPapers 7).

(25) The polis was a community of politai. Structurally it was a descent group of citizens of both sexes and all ages; but functionally, the politai were the adult male citizens. As a community, the Archaic and Classical polis was primarily a political and a military organisation, a male society from which women and children were excluded, not to speak of foreigners and slaves. The polis was a highly institutionalised community, and at the core of the polis were the political institutions where the politai met and isolated themselves from women, foreigners and slaves. Political activity was a fundamental aspect of the community, and, as a polity, the polis is best seen as a very deliberately planned and highly rational form of political organisation (CPCActs 4: 57, 493-502; 30 CSC: 165-173; P&P 241-242).

(26) The sweeping statement that the ancient Greek polis was a fusion of state and (civil) society is a false generalisation. It is true for Sparta, but false for Athens. Again Sparta seems, in the Classical period at least, to have been the exception and Athens much closer to what we can expect to have been the case in other poleis, at least in democratically governed poleis. The presumption is, however, that Sparta and poleis organised like Sparta, though exceptional, were poleis to the same degree as Athens and poleis organised like Athens. Thus, whether a given polis was a fusion of state and society or separated state from society is irrelevant for its status as a polis, but is, of course, relevant for the modern historian's discussion of whether or not it was a state or, to be more precise, a state in the modern liberal-democratic sense (CPCActs 5: 84-106).

(27) Like the modern state, the polis is often seen not just as a system of political institutions, but as an abstraction, i.e. a permanent public power above both ruler and ruled (CPCActs 5: 67-73; CPCPapers 6: 22-26).

(28) The modern concept of sovereignty is subdivided into supremacy (internal sovereignty) and independence (external sovereignty). Our sources for the ancient polis reveal not the same, but a similar distinction between internal supremacy expressed through the adjective kyrios and external independence expressed through the adjective autonomos (CPCActs 5: 73-83).

(29) Like the modern state the polis provided its citizens with a feeling of common identity, based on traditions, culture, ceremonies, symbols and sometimes (presumed) common descent. For a Greek citizen the polis was his fatherland (patris) for which he was expected, if necessary, to die, just as the modern state expects "every man to do his duty". Both in the ancient and in the modern world victories in the Olympic Games are won by participants representing, respectively, their state or their polis. The polis had no flag; but city-ethnics (Naukratites, Milesios etc.) were used as a kind of surname which, at the same time, indicated the bearer's status as a citizen of the polis in question. Other symbols were the eternal flame burning in the prytaneion, cult festivals, monumental architecture etc. (CPCActs 5: 72-73, 120; Ark. 203-210; forthcoming article in CPCPapers 7).

(30) The increasingly common view that the polis was not a kind of state but a "stateless society" is based on a skewed comparison. It is claimed that in the polis there was no clear distinction between rulers and subjects, no separate political institutions, no prison and law-enforcing apparatus of any consequence, whereas all these features are characteristics of the state from the age of Thomas Hobbes onwards. In the polis, administration of justice was dominated by self-help, private apprehension and private prosecution of criminals. This apparent contrast between the polis and the early modern European state is based on suppression of (a) all the ancient sources which show that even in democratic poleis there was a clear distinction between archontes and archomenoi, but combined with an annual rotation, that every polis had a prison and separate political institutions empowered to enforce the laws, and that selfhelp was legal only in a few narrowly defined cases. (b) It is also passed over in silence that, in the early modern state down to the 19th century, self-help was allowed against the same types of criminal as in the polis, that private apprehension and private prosecution of criminals took place in the great majority of cases, and that the prisons served precisely the same functions as in a Greek polis (CPCPapers 6: 17-47).

(31) The Greeks were conscious of using polis in two basically different senses, viz. (1) settlement, and (2) community. As a (nucleated) settlement a polis consisted of houses, as a (political) community it was made up of human beings. A study of words used synonymously with polis shows that both the local and the personal sense were used in a number of different ways. (1) In the local and physical sense of settlement, polis was used (a) synonymously with akropolis about a small and often fortified hill-top settlement; (b) synonymously with asty about an urban centre; and (c) synonymously with ge or chora about a territory (composed of a town plus its hinterland). (2) In the personal sense of community, polis was used (a) synonymously with politai about the citizen body; (b) synonymously with ekklesia, vel sim. about the people's assembly or some other body of government; and (c) synonymously with koinonia about a political community in a more abstract sense. (Re 1a-c): apart from some frozen formulas found in inscriptions, (1a) is rare; (1c) is not common and mostly attested as a connotation. (Re 2a-c): the three different uses of polis in the sense of political community are very close and are indeed just different aspects of one concept: in (2a) and (2b) the polis is understood in a concrete, in (2c) in an abstract sense, just as we use the term state sometimes about the body politic, sometimes about the government, and sometimes about a permanent public power above both ruler and ruled. Thus, the two important senses are (1b) "city" and (2a-c) "state" of which both are very common, whereas (1c) "territory" is a less common variant. In numerous passages these three senses are indistinguishable. The various senses do, of course, overlap, and especially so when polis is used as a generic term or a heading (see infra no. 38) or is opposed to other terms as, e.g., chora or ge (CPCActs 5: 17-34).

(32) The common view is that polis in the sense of "town" is often used about urban centres which were not poleis in the sense of being centres of city-states. Our examination of the sources disproves the orthodoxy and shows instead what we have called the Lex Hafniensis de Civitate: in Archaic and Classical sources describing Hellenic communities the term polis used in the sense of "town" is not applied to just any urban centre but only to a town that was also the centre of a polis in the political sense. Thus, the term polis had two different meanings, "town" and "state", but even when it is used in the sense of town its denotation seems almost invariably to be what the Greeks called polis in the sense of a self-governing community and what we today call a city-state. The Lex Hafniensis does not apply to barbarian poleis (CPCActs 3: 25-34; CPCPapers 5: 173-215).

(33) In opposition to the modern trend to minimise the urban character of the Greek polis, we want to emphasise urbanism as a paramount aspect of the polis, and we believe that we can substantiate what can be called the inverse Lex Hafniensis: in the sense of polity, the term polis was invariably applied to a polity which had an urban centre also called polis. The common view that there were quite a few poleis without an urban centre has no foundation in the sources of the late Archaic and Classical periods (see no. 62 infra), and it cannot even be convincingly demonstrated for the early Archaic period, for which the lack of sources forces us to suspend judgement (CPCActs 1: 13-16; CPCActs 4: 34-42; 30 CSC: 154-165).

(34) The close connection between the political, territorial and urban aspects of the polis is also apparent in the use of toponyms. The common view is that the toponym invariably denotes the town and the state is referred to by the city-ethnic. An inspection of both literary sources and inscriptions reveals that this view is an exaggeration: a toponym denoting a polis-town (e.g. Athenai, Korinthos or Tanagra) is also commonly used as the name of the polis-state and as the name of the polis' territory (CPCActs 3: 28, 38; Article no. 2).

(35) In a number of studies on individual authors we have demonstrated that the term polis is used much more consistently in our sources than previously believed, and that the site-classifications found in Archaic and Classical literary and epigraphical sources must be taken seriously and cannot just be brushed aside as unreliable whenever it suits a modern historian to question a settlement's status as polis (CPCActs 3: 14-20; cf. CPCActs 2: 39-45; CPCActs 3: 39-54; CPCActs 5: 17-34; CPCPapers 2: 83-102, 128-132; CPCPapers 3: 127-167; CPCPapers 4: 17-27; CPCPapers 5: 133-215).

(36) The highest number of occurrences of the word polis in any ancient source is in Stephanos of Byzantion's Lexicon. Whenever Stephanos quotes or paraphrases his source, he is remarkably reliable, and in all such cases his use of the term polis as a site-classification can be trusted to stem from the author he quotes (CPCPapers 1: 99-124). On the other hand, we must refrain from using Stephanos when we are unable to establish that he took the site-classification from the source he cites. Thus, if we pick out the actual quotations, Stephanos provides us with valuable information about how the term polis was used by otherwise lost historians, such as Hekataios, Theopompos and Ephoros. All three were non-Athenian authors, and thus important sources for checking to what extent the bulk of our sources give a too Athenocentric view of the concept of polis (CPCPapers 4: 17-27; CPCPapers 5: 141-150).

(37) In the Classical sources, the highest number of occurrences of the word polis applied to a named community is in Pseudo-Skylax' Periplous of the fourth century B.C. Comparison with all other contemporary sources shows that whenever the site-classification found in Pseudo-Skylax can be checked, there is a remarkable agreement between what Pseudo-Skylax classifies as poleis and what is found in all other sources. In the chapters about Greece and the regions densely colonised by the Hellenes, Pseudo-Skylax follows the generally accepted usage and restricts the term polis to urban centres which were also the political centre of a polis. Erroneous classifications and "ghost cities" are found only in the chapters that deal with remote regions (CPCActs 3: 30-32; CPCPapers 3: 137-167).

(38) A word has a broader field of meaning when used as a generic term or a heading than when used individually in its specific sense, cf. e.g. the term nation in United Nations. Thus, one has to distinguish between (a) polis used as a generic term and/or heading of a list of political entities, and (b) polis applied to an individual community. A failure to make this distinction has led some modern historians to assume that polis often signifies and denotes not only city-states but also large "territorial states" (as states covering a whole country are usually called). The evidence for such a view, however, is that such states are occasionally recorded alongside genuine city-states in lists headed by the word polis. A more important observation is that the term polis is not applied to such communities when they are mentioned on their own (CPCPapers 4: 9-15; CPCActs 5: 124-132).

(39) The polis was not believed by the Hellenes to be a specifically Hellenic institution, one that separated Greeks from barbarians. True, Aristotle's basic view was that the polis was peculiar to Hellenic civilisation and that in this respect there was a gulf between the Greeks and the others. But the gulf is rather between Aristotle and our other sources. In Hekataios, Herodotos, Thucydides, Xenophon and Pseudo-Skylax we hear about hundreds of barbarian poleis, often in the sense of city rather than state, but sometimes obviously in the sense of political community and applied to, e.g., the Etruscan or Phoenician city-states (CPCPapers 5: 180-182; 30 CSC: 145; P&P 237-238, 242).

Membership of the Polis. The Concept of Citizenship

(40) Modern historians disagree about which of the inhabitants of a community to include or exclude in the concept of polis. On the basis of Aristotle's Politics, books 1 and 3, we have shown that the Greeks themselves had two different views of this issue according to whether they saw the polis as a political community or as a social and economic community. When the term polis is attested in the sense of state, the focus is upon the political institutions and the polis is seen as a community of adult male citizens from which women, free foreigners and slaves are excluded. In this context the "atom" of the polis is the citizen (polites). When the term polis is used in the sense of town, the focus is upon the economic and social aspects of the community and the polis comprises all inhabitants: citizens, free foreigners and slaves of both sexes and all ages. The "atom" of the polis is the household (oikia) (CPCActs 1: 16-18; Article no. 1).

(41) In the ancient Greek world citizenship was, essentially, what it has become once again in the modern world, i.e. the legally defined hereditary membership of an individual in a state whereby the member (in the modern world called a citizen or a national) acquires political, social and economic privileges that a non-citizen member of the community does not enjoy, or enjoys only partially. As a rule, a person is a citizen of one state only. In ancient Greece the corresponding terms used were politeia to denote citizenship itself, and polites to denote the citizen if the emphasis was on a citizen's exercise of his political rights, whereas astos (masculine) and aste (feminine) were commonly used to denote a person of citizen birth. As a rule a person was a polites of one polis only (CPCActs 5: 114-115; 122-123).

(42) In Aristotle's Politics the key concept, politeia, does not just mean "constitution" or "structure of the polis", it often has the connotation body of citizens, citizenry, and, in the sixfold classification of constitutions, politeia – the positive variant of popular rule – should be understood in the specific sense of citizen-constitution (CPCPapers 1: 95-97).

(43) We believe that of all western peoples the Greeks are unique in having used their equivalent of hereditary surnames – i.e. the ethnic – as a sign of political status, primarily as an indicator of a person's status as citizen of a polis. Furthermore, instead of the prevailing defective terminology: ethnic (indicating affiliation with either a region or a polis) and demotic (indicating membership of a municipality and used in this narrow sense about citizens of Athens, Eretria, Rhodes and a few other places), we propose to distinguish between regional ethnics (indicating affiliation with a whole ethnos), city-ethnics (used externally to indicate membership of the polis as a whole) and sub-ethnics (indicating affiliation with a civic subdivision, i.e. a phyle, a demos, a kome or a phratria etc. and used internally to indicate membership of a polis). Apart from the term polis itself, the attestation of a city-ethnic or a sub-ethnic is one of the best criteria for identifying a community as a polis (CPCPapers 3: 169-196).

(44) The use of a regional ethnic as part of a personal name or as a coin legend does not necessarily show that the region in question was politically united and formed a federal and/or tribal state; thus, Arkas is found both as part of personal names and as a coin legend in the fifth century, a period which is characterised by serious political disunity, and, again, Arkas is used as part of a personal name after the dissolution of the Arkadian Confederacy in 324 (CPCPapers 3: 39-61; Ark. 54-66).

The Concept of the Dependent Polis

(45) In most modern accounts of ancient Greek society independence (often equated with autonomy) is singled out as the most important defining characteristic of the polis, and the ancient concept of autonomia is equated with the modern concept of autonomy without any thought for the fact that the modern concept covers everything from true independence to a very restricted form of self-government (cf., e.g., Gaza and Jericho after 1993). Our sources show, however, that in some periods more than half the poleis were dependencies without autonomia. In reply to the "peer polity interaction" model of the polis we have emphasised the hierarchical structure of the polis culture, and in opposition to the view that all poleis were autonomous we have developed the concept of the dependent polls. We have dissociated the concept of polis both from the ancient concept of autonomia and from the modern concepts of independence and autonomy. We hold that the concept of autonomia becomes linked to the concept of polis only in the course of the fourth century, and that the history of the autonomos polis does not end ca. 338 B.C. On the contrary, that is in fact where it begins, viz. after autonomia had lost its original meaning of independence and could be taken to signify self-government only (CPCActs 1: 18-20; CPCPapers 2: 21-43; CPCPapers 3: 113-136; 30 CSC: 172-173).

(46) Dependent poleis existed in many different shapes and sizes, and certain types of dependent poleis were common in some regions but virtually non-existent in others. So far, we have isolated the following fifteen different types of dependency: (1) A polis situated inside the territory of a larger polis. (2a) A polis in the peraia controlled by an island or, conversely, (2b) an island controlled by a mainland polis. (3) An emporion organised as a polis dependent on a larger polis. (4) A colony being a polis dependent on its mother-city. (5) An Athenian klerouchy. (6) A perioikic polis in Lakedaimon. (7) A polis that is a member of a hegemonic federation. (8) A polis that is a member of a hegemonic league (symmachia) which has developed into an "empire" (arche). (9) A polis that persists as a polis after a sympoliteia with another polis. (10) A polis that persists as a polis after a synoikismos. (11) A polis that, together with other poleis, makes up a "tribal state". (12) A polis which is controlled by a foreign monarch. (13) A polis founded as a fortress. (14) A polis which is a major port of an inland polis. (15) A polis that is at the same time a civic subdivision of another polis. There is, of course, a considerable overlap between the different types (CPCPapers 4: 29-37; Inv.).

(47) Coins issued by a polis should not be seen primarily as a symbol of independence or autonomia or even civic pride of that city. Many poleis were happy to use the coins of other poleis and never cared to set up their own mint (CPCActs 2: 257-291). Conversely, coins were often issued by dependent poleis, e.g. member states of a federation (CPCActs 2: 20-21). By the use of city-ethnics as legends to identify the issuing authority, coins struck by a polis are fairly easy to distinguish from coins issued by a sanctuary, a district, a federation, or a ruler etc., and such coins are accordingly a very good indication of the polis status of the issuing authority (CPCActs 2: 10-11; Inv.).

Civic Subdivisions

(48) The self-government of a polis must be contrasted with the activities and powers exercised by civic subdivisions such as a phyle, a kome or a demos: like a polis (dependent or independent) a civic subdivision could have its own temples, including a theatre, its own cults and its own festivals. It could have its own assembly, in which both laws (nomoi) and decrees (psephismata) could be passed and taxes and liturgies imposed; there could be separate local magistrates and a local court. But, in contradistinction to a polis (dependent or independent), a civic subdivision had no prytaneion, no bouleuterion, no boule, and no desmoterion; its members were citizens of the polis of which the subdivision was a part, and were not citizens of the civic subdivision as such; a local assembly had no right to pass citizenship decrees and proxeny decrees; and a local court could impose fines but was not empowered to pass a sentence of death or exile. A civic subdivision did not have its own coins, and it had no right to send out envoys or to enter into relations with foreign states. The members of a civic subdivision could form a unit of the army of the polis, but would not operate as a separate army (CPCPapers 4: 31; 30 CSC: 603-609).

(49) Many poleis had no civic subdivisions and, in particular, the territorial civic subdivisions are not well attested at all. In the Archaic and Classical periods civic subdivisions are unattested in Boiotia, Thessaly, Lesbos and Aiolis, i.e. in the Aiolic-speaking regions of Hellas. The demos and the kome were the two principal types of territorial subdivisions. But demoi were confined to a few prominent poleis, principally Athens, Euboia, Rhodes and Kalymna, and in the Archaic and Classical periods komai are attested as civic subdivisions in two poleis only: Megara and Mantinea (Inv.).

(50) It is commonly held that the civic subdivisions show a high degree of permanence and that almost all innovations took place during the Age of the Tyrants. Yet, a closer look at the attested reforms indicates that civic subdivisions were subject to constant transformations and with the passage of time became more and more artificial – a typical instance of the Greeks' conscious and continuous remodelling of their society and institutions. Reforms and revisions of civic units are so frequently attested throughout the period that a system attested in Hellenistic sources can only exceptionally be retrojected back into the Classical period (Inv.).

Stasis

(51) Our investigations support those historians who argue that most poleis were split up into what in the sources is called two opposing poleis, i.e. two factions among the citizens. The two opposing factions were often one of the rich (supporting oligarchy) and one of the poor (supporting democracy), but sometimes they were two different ethnic groups living side by side as citizens of the same polis, usually a colony, or two subfactions of wealthy citizens in an oligarchy (Inv.; Article no. 4).

(52) The goal of each faction was to control and (if necessary) to reform the political institutions of the polis. The opposition between the two factions within a polis entailed a constant tension and discord resulting in repeated outbursts of civil war, during which each faction was prepared to collaborate with a congenial faction in a neighbouring polis, or in a distant but hegemonic polis. The members of each faction were, in fact, willing to sacrifice the freedom (eleutheria) and independence (autonomia) of their polis if only they could get the upper hand of the opposing faction (Inv.; Article no. 4).

(53) By giving up autonomy in the sense of independence, a faction could keep what was much more important to its members, namely the self-government of one's polis exercised by one's own faction. If the opposing faction came to rule the polis it would impose its will in all matters, day in and day out. If one's own faction ruled it would be in control of almost all decisions that mattered in everyday life. By sacrificing the autonomia of the polis one would have to pay tribute, but not necessarily a large one; in times of war one might have to assist the hegemonic polis. But essentially the polis was left as a dependent, but still self-governing community. Dependent status became a nuisance only if a polis had to suffer a foreign garrison on its akropolis, or if its self-government was constantly interfered with by outside harmosts or episkopoi. On the other hand, apart from the help from the neighbouring polis to subdue the opposing faction, there might be a bonus, namely that a small polis could have the hegemonic polis as its protector, and so be safe from being attacked by neighbours who might be a more severe threat than the, perhaps, more distant, hegemonic polis (Inv.; Article no. 4).

(54) What endangered the prosperity and well-being of a polis was not so much the loss of autonomia as the lack of homonoia. Accordingly, what the Greeks prayed for was not autonomia but homonoia and freedom from stasis. As far as we know, autonomia was never deified in any polis and made object of a cult, whereas homonoia became a goddess whose cult was venerated all over the Greek world, especially from the fourth century onwards (Inv.; Article no. 4).

Types of Constitution

(55) A basic classification of constitutions as tyrannis, oligarchia or demokratia is attested both in Athenian and in non-Athenian sources, both in inscriptions and in literature, and in all types of prose: history, rhetoric and philosophy. The subdivision of each of the three basic types into a positive and a negative variant is peculiar to the Athenian philosophers. Apart from the rulers of barbarian kingdoms (Persia) and remote colonies (Kyrene), basileia is usually treated as an obsolete historical form of constitution and contemporary monarchs are called tyrannoi or monarchoi. Also, aristokratia as a positive form of the rule of the few seems to be invented by Sokrates and taken over by the Socratic philosophers: Plato, Xenophon and Aristotle. Aristokratia is unattested in inscriptions, in speeches delivered before the assembly or the court and in historians, apart from Thucydides (Inv.).

(56) According to Aristotle, oligarchia and especially demokratia were overwhelmingly the commonest constitutions in Greece in his own time; basileia had virtually disappeared and tyrannis was no longer a common form of constitution. The evidence contradicts his views about tyrannis. It was not only found in remote regions such as Sicily or the Pontos. After a low point in the Greek homeland in the fifth century, tyrannies re-appeared in the fourth century all over the Hellenic polis world, and tyrants ruled poleis in the Peloponnese, in Euboia, in Thessaly and in Lesbos etc. (Inv.).

The Polis as an Urban Centre

(57) Almost all towns (poleis in the urban sense) had a population of over 1,000 inhabitants but very few surpassed a population of 10,000, probably some 24 altogether (23 with walls enclosing an area of over 150 ha plus Sparta). The archaeological surveys conducted since ca. 1980 (Northern Keos, Southern Argolid, Central Boiotia, Methana, Melos, Asea etc.) as well as written sources show that small and middle-sized Classical poleis had a large urban and a much smaller rural population. There was no sharp division between townsmen and countrymen but a continuum in which the gap between town and country was bridged by a large number of people who lived in the city but worked as farmers in the hinterland (Ackerbürger). And the urban class of "landowners" was small as against a majority consisting of farmers, fishermen, artisans and traders. To have a sizable part of the population settled in the hinterland was a characteristic of a few large poleis, and it was here an oppostion between town and country emerged and was felt. Thus, in Classical Greece the degree of urbanisation was inversely proportional to the size of the poleis: the smallest poleis had the highest degree of urbanisation, whereas the few large poleis had a higher percentage of its population permanently settled outside the major urban centre, its polis in the urban sense (CPC Acts 4: 25-31, 44-47, CPCPapers 7).

(58) Asty and polisma, the two other common terms meaning "town" or "city", are used only about urban centres which could also be called polis. In Archaic sources asty is sometimes used in the sense of community, and the derivative astos (man of the asty) is never used in the sense of "townsman" but only in the sense of "citizen" (by birth). Polisma is mostly used about urban centres in border areas where Greeks and non-Greeks lived together (for polisma, see CPCPapers 2: 129-32; for asty, see CPCActs 4: 58-60).

(59) Although the ancient Greeks showed a tendency towards clustering together in towns, it is a curious fact that they never coined a word to denote the urban population. The term polites (citizen) is almost invariably linked to the concept of polis in the political sense. The word designates the adult male citizen and is only very exceptionally used in the sense of townsman (CPCActs 4: 10-12). The feminine form politis is sometimes used of females of citizen birth. There is no attestation of politis signifying a woman exercising political rights (30 CSC: 166 with n. 286).

(60) It is commonly held that the polis as a state can be traced back to ca. 700 B.C. or even earlier, whereas the polis as an urban centre emerged as late as the late sixth century. The archaeological record, however, especially recent excavations of, e.g., Eretria, Miletos, Megara Hyblaia and Sicilian Naxos, indicate that urbanisation often took place as early as the first half of the 7th century and, consequently, was contemporary or almost contemporary with state formation (30 CSC: 160-161).

(61) It is well known that in the Archaic period many poleis were not protected by a circuit of walls, but the number of walled Archaic poleis is much larger than usually believed (forthcoming thesis by Rune Frederiksen), and by the fifth century almost every polis we know of (except Sparta and a few others) was protected by city-walls, and a circuit may now be regarded as an essential characteristic of a polis (CPCActs 2: 245-56). Furthermore, references to city-walls in Homer, Alkaios and other early texts show that walls seem to have been an important element in the concept of polis already ca. 600 B.C. (CPCActs 3: 22; CPCActs 4: 52-53).

(62) A polis kata komas oikoumene is not (as usually believed) a polis in the political sense that is settled in komai and has no urban settlement as its centre. The two key passages are Thucydides 1.5.1 and 1.10.2 where he uses polis in the urban sense, and by a polis kata komas oikoumene he understands an unwalled town consisting of a cluster of komai instead of being one unified settlement around an akropolis. Thus, the polis (in the urban sense) is here a conurbation. Sparta was a polis with such an urban centre and is duly described in our sources as a polis in the urban sense and also as a polisma and as an asty (CPCPapers 2: 55; CPCActs 4: 34-36).

(63) Grid-planned poleis are much earlier than commonly believed. The per strigas system of townplanning is copiously attested in the Sicilian and Italian poleis of the Archaic period, and Archaic grid-planned towns are found in other regions as well. The Grid-planned polis seems to have emerged in the western Greek colonies as early as ca. 700 and it spread to the Greek homeland in the following period. Thus, there is no essential connection between grid-planning and democracy (CPCActs 3: 317-373).

(64) There was virtually no monumental political architecture before the Hellenistic period. The palaces which many modern historians ascribe to the tyrants of the Archaic period are at present without support in written or archaeological evidence. Prytaneia and bouleuteria were mostly plain buildings of modest dimensions and cheap materials. The people's assembly was convened either in the agora or in a theatre, connected with a sanctuary constructed primarily for dramatic performances etc. Genuine ekklesiasteria are exceptional. Dikasteria met in the agora or in buildings erected for other purposes (e.g. stoai). Again, in the Archaic and Early Classical periods both the gymnasion (with palaistra, stadion and hippodromos) and the theatre were simple constructions which in most poleis have left no trace whatsoever. Down to the second half of the fourth century B.C. virtually all monumental architecture was sacred, and consequently it is extremely difficult to determine from remains of buildings alone whether or not a nucleated settlement was a polis (with a prytaneion and a bouleuterion) or some kind of second-order settlement (CPCPapers 1: 23-90; CPCActs 4: 103-116; 30 CSC: 162-165).

(65) Whereas every polis seems to have had at least one urban or extra-urban monumental temple (CPCActs 4: 109), the theatre (sometimes extra-urban) was a type of building to be found in a fairly small number of mostly fairly large poleis and especially in some regions. In other regions, e.g. Crete, there were no theatres before the late Hellenistic or Roman periods (CPCPapers 6: 65-124).

(66) In our analysis of the concept of emporion we distinguish between (1) a community that had an emporion, and (2) a community that was an emporion. In the first case the emporion was a polis' centre of foreign trade and was distinct from the agora, which was the centre of local trade. In the second case the emporion was a community, and the traditional view has been to distinguish an emporion (a trading-station) from an apoikia (a colony), and to hold that an apoikia was organised as a polis whereas an emporion was not a polis. But, apart from the toponym Emporion (attested ca. 530-500 as a polis in Spain) and Herodotos' classification of Naukratis as an emporion of the seventh century, there is no evidence earlier than ca. 450 B.C. that the Greeks had developed the concept of emporion; all emporia mentioned in Classical sources were in fact poleis that had an emporion, perhaps even including the inland emporion Pistyros in Thrace. The distinction between a community that had an emporion and one that was an emporion seems to vanish, but not quite. First, in poleis such as Athens the emporion was not the whole polis but only a part of the polis. In the settlements identified as being emporia the centre for international trade may have been the paramount feature of the settlement, where the majority of the citizens worked and from which the polis obtained almost all its revenue. Furthermore, all the sites classified as being emporia are colonial settlements and centres of trade between Greeks and barbarians. Finally, a site classified as both a polis and an emporion seems to have been a specific type of dependent polis, namely one in which the port was the dominant part of the settlement (CPCPapers 4: 83-105; updated and revised treatment forthcoming in Brill's volumes on Greek colonisation).

(67) A comparison of the concept of emporion with the concept of agora shows that every polis had an agora. In Archaic poetry the agora is described as the place where the people had the sessions of the assembly. In the Classical period almost all traces of the agora as an assembly place have vanished, and the agora was now primarily the market place. Conversely, the earliest evidence we have of the economic functions of the agora is a reference in the Gortynian law of ca. 480-60 B.C. We must seriously consider the possibility that the concentration of local trade in the agora and of long-distance trade in the emporion was a phenomenon to be dated in the late Archaic and early Classical periods and to be connected with the development of the institutions of the polis (CPCActs 4: 60-61; 102-105).

(68) "Civic space", in the sense of space reserved for citizens, is a misleading concept invented by modern historians. There was no civic space to which only citizens were admitted. What is attested is an opposition between private property and publicly owned property. Sacred precincts were usually open to everybody (except atimoi) and so were the agora and other publicly owned areas. But public space could, when required, be used for gatherings of adult male citizens from which women, metics and slaves were excluded (CPCActs 4: 12-17).

Territory and Settlement Pattern

(69) The words for an urban centre and its hinterland form a pair of antonyms in most European languages, e.g. city/country (English), Stadt/Land (German), cité/pays (French), by/land (Danish) etc. In ancient Greek it was the word for city (polis) that came to denote the political community, whereas in modern European languages it is invariably the word for country that is also used synonymously with state. The most likely explanation is that most poleis had one urban centre only which was also the political centre of the community, whereas the emerging mediaeval European states had many towns but no political centre. The king and his court moved from castle to castle and from town to town (CPCActs 1: 15; 30 CSC: 153).

(70) Whereas a territory is an essential element in the modern concept of state, it was a less important aspect of the ancient polis than the people (hoi politai); but the frequent use of exile as a penalty is enough to show that every polis had a territory, sometimes marked with horoi (boundary stones). The concept of Poleis ohne Territorium is basically misleading. For most poleis the territory was no larger than the immediate hinterland of the polis town (CPCActs 5: 53-56).

71) Some 60% of the poleis had a territory of max. 100 km2, and close to 80% had a territory of max. 200 km2. Only 10% had a territory of over 500 km2 and only thirteen poleis one of over 1,000 km2. Both the mode and the median fall between 25 and 100 km2. The mean, however, is ca. 150 km2. This shows that a mean can be a dangerous simplification and that the Normalpolis may be a misleading concept. On the whole colonies had larger territories than poleis in the Greek homeland. Only four poleis had so large a territory that it took more than one day to walk from the periphery to the urban centre, viz. Syracuse, Sparta, Pantikapaion and Kyrene. In the great majority of poleis it was possible to get from the border to the urban centre and back again in one day (Inv.).

(72) Modern archaeological analysis of the settlement pattern of ancient Greece is based on a three-tier hierarchy: (a) first-order settlements: (towns/ cities), (b) second-order settlements (villages/hamlets), and (c) third-order settlements (isolated farmsteads); the fundamental distinction is between nucleated settlement (a + b) and dispersed settlement (c). We have shown that the Greeks had a clear terminology for (a), namely: polis, polisma or asty; a defective terminology for (b), namely kome or, to some extent, demos, but no clear terminology for (c); the fundamental distinction was between polis (a) and chora or ge (b + c). To conclude: the Greeks emphasised the polis and the distinction between polis and chora, but they had a defective perception of the settlement pattern and showed little interest in whether settlement in the chora was nucleated or dispersed (CPCActs 4: 20-25).

(73) Any small urban centre that is not a polis is traditionally classified as a kome (in Athens and a few other poleis as a demos). This model implies that in Classical Greece there must have been thousands of komai in addition to the hundreds of poleis. But whereas some 447 urban centres are explicitly called polis in our sources, there are only some 30 known localities that are explicitly called kome in Archaic and Classical texts. The identification of all the other small nucleated settlements as komai is either without any foundation in our sources or based upon an anachronistic projection back into the Classical period of the terminology found in Strabo and Pausanias. In contemporary sources komai are only attested in some regions and mostly in the socio-economic sense. Megara and Mantineia are the only unquestionable attestations of kome used constitutionally as the designation of a civic subdivision of a polis (CPCPapers 2: 45-81).

(74) In many parts of the Greek world, e.g. Boiotia, Arkadia, Chalkidike, the number of poleis constituted a very large percentage of all nucleated settlements. Together with some other regions Attica is exceptional in having over one hundred nucleated centres, only one of which was a polis. Regions settled in poleis with none or very few villages between the poleis seem to be the rule rather than the exception (CPCActs 3: 73-78; 30 CSC: 154-155).

Polis Religion

(75) Since Fustel de Coulanges it has commonly been claimed that religion was the dominant aspect of community life and that religion was the very centre of the Greek polis. We find that this holistic view of the polis is skewed. The opposition between the sacred and the secular is well attested in the sources. Religion was indeed extremely important, but constituted one aspect of polis life only, and not necessarily the focal one, which was the polis as a community of politai. Both as a political and as a military organisation the polis was a male society from which women were excluded. Religion was different. Women took part in the rites and cults of the polis. Most goddesses were served by priestesses. In religion women were insiders. Again, in many poleis the priests (hiereis) and priestesses (hiereiai) who performed the sacrifices and rituals were not technically polis officials (archai), and religious specialists such as manteis (prophets) and kathartai (purifiers) were marginalised in the polis, and not at all well organised. A person who escaped into a sanctuary or held onto an altar was protected against violence. Even when the suppliant was a criminal, the punitive authority of the polis stopped at the threshold of the temple. polis religion was extremely important but not necessarily the core of the polis. On the other hand, there was no institutionalised and organised religious sphere distinct from and, sometimes, opposed to the polis sphere. In the Greek world there was nothing like the Mediaeval opposition between two competing power organisations: the Crown and the Church (CPCPapers 2: 201-210; 30 CSC: 167-169).

(76) Instead of the monistic and holistic view of polis religion, we prefer a dualistic view (religion and polis as two overlapping but distinguishable aspects of society). Thus, we subscribe to the following tripartite analysis of the relationship between polis and cults: (a) the polis makes use of religion, (b) the polis makes decisions about religion, especially by organising all the cult festivals and (c) the polis makes religion, e.g. by introducing new and sometimes specific political cults (CPCPapers 2: 201-210; 30 CSC: 167-169).

(77) Some poleis had no patron god or goddess at all whereas others had several simultaneously. The cults of Athena Polias and Zeus Polieus were originally connected with polis in the sense of akropolis, and should not invariably be interpreted as cults of the polis in the sense of political community (CPCActs 2: 292-325).

(78) To appoint theorodokoi to host theoroi sent out to announce a Panhellenic festival was a political act performed by a polis or, sometimes, a ruler, and in the preserved lists of theorodokoi names of towns followed by a personal name are important evidence of the polis status of the towns listed (CPCActs 2: 113-170; Inv.).

Emergence and Disappearance of Poleis

(79) A new polis emerged either by natural growth or by foundation. Foundation of a polis took place either by colonisation or by synoecism. Most poleis in Hellas itself seem to have emerged by natural growth, usually a slow and imperceptible process so that it is impossible to trace when and how the people of a given community became conscious of being a polis. Due to lack of sources it is no surprise that polis formation by growth is unattested in the Archaic period. It is more surprising that the formation of poleis by growth is equally unattested during the Classical period. Even in the large corpus of sources relating to the fourth century, there is not one single attested instance of the formal creation or recognition of a pre-existing community as becoming, from now on, a polis (30 CSC: 149).

(80) The disappearance of poleis has been one of the neglected problems of Greek history. According to whether the polis is seen as a city or as a state, we can distinguish between two basic types: (a) a polis disappears as a political community (but may persist as a nucleated settlement); (b) a polis disappears both as a political and as an urban centre as a result of the population being killed, moved or sold into slavery, and sometimes the urban centre is destroyed. The following variants are frequently attested: (1) Destruction by which all men are killed, whereas women and children are sold into slavery (andrapodismos). (2) The whole population of a polis is expelled or moved by force to another polis (anastatos polis). (3) The population of a polis is dispersed over a number of villages (dioikismos). (4) A polis disappears when the whole population emigrates and founds a polis in a different place. (5) A polis disappears when the whole population joins in a synoecism of another polis. (6) A polis has its status changed from polis to kome, vel sim. (7) A polis disappears because of a cataclysm or a similar catastrophe (30 CSC: 150-152).

(81) In most cases an andrapodismos was not a complete but only a partial eradication of a community. It has been demonstrated that plundering and destruction of the territory of a polis in a war was less devastating than traditionally assumed. Similarly, a survey of all sources shows that the andrapodismos of the population of a polis did not usually entail an annihilation of the community. Mostly the community was re-established and the settlement rebuilt after a short period. And many of the new settlers were refugees who had escaped the andrapodismos (Inv.).

Synoecism

(82) The distinction between political and physical synoecisms should be changed into a distinction between the political and the physical aspects of a synoecism. The purely political synoecism is a mythological fiction, known only from Thucydides' account of Theseus' unification of Attica. Physical relocation of communities was the central aspect of all attested synoecisms, sometimes accompanied by the setting up of a new polis in the political sense (CPCPapers 2: 55-56).

(83) A synoecism resulted only occasionally in the disappearance of the contributing communities. In the majority of cases the synoecism was partial and the contributing settlements, mostly small poleis, persisted or re-appeared shortly afterwards, and usually with their status of polis preserved (CPCPapers 2: 60).

(84) Synoecism is found in the following four forms: (a) a polis is created by merging a number of komai or demoi (Kassopa, mid-fourth century); (b) a polis is created by merging two or more poleis (Rhodes 408/7); (c) a polis is reinforced by absorbing one or more neighbouring komai or demoi (Lepreon mid-fifth century); (d) a polis is reinforced by absorbing one or more neighbouring poleis (Thebes 431); (e) possible variants are a combination of (a) and (b), and a combination of (c) and (d). Types (a) and (c) are rarely attested whereas (b) and (d) are the two common variants (CPCPapers 2: 57-58).

Ethnos

(85) Contrary to a commonly accepted subdivision of Greek political communities into poleis and ethne, it can be shown than an ethnos was conceived principally as a people, not as a state. Most poleis in Greece were integrated into the differently based ethnos structure. Thus polis and ethnos must be seen as two different but often overlapping forms of organisation, not as two clearly separable and distinct types of political unit. Before the Hellenistic period the common phrase poleis kai ethne should not be taken to mean "polis-states and ethnos-states", but rather "states and peoples" (CPCActs 5: 68).

(86) It has commonly been held that it was a characteristic of the ancient Greek federations that they were based on larger regions such as Boiotia; however, the study of the Mainalians has shown that so-called tribal states could be formed inside a larger region, and that these tribal states were not just ethne but composed of several poleis united in a political organisation resembling a federation (CPCPapers 3: 100-103; CPCActs 3: 132-143).

(87) The regional structure of mainland Greece was not a static phenomenon inherited from the Dark Ages; a study of the Triphylians shows that Triphylia was a region with its own ethnic identity emerging around 400 B.C. but swallowed up by Arkadia a generation later (CPCPapers 4: 129-162).

(88) Hegemonic leagues were much more common than usually realised; in addition to the Spartan-led Peloponnesian League, there existed in the fifth-century Peloponnese smaller hegemonic leagues headed by Elis, Mantineia and Tegea; Archaic Sybaris, too, was the head of such a league (CPCPapers 3: 79-87).

Colonies and poleis outside the Greek Homeland

(89) The traditional view is that the formation of the polis preceded colonisation. In recent years this view has been challenged by the opposite hypothesis: that the polis emerged in consequence of colonisation and that the colonies influenced polis formation in the homeland. We suggest that the polis emerged in the homeland and in the colonies more or less simultaneously; but in the colonies the new start and the proximity of an indigenous and sometimes hostile population speeded up the formation of the polis both as a walled town and as a political community so that, in some regions, the fully-fledged polis emerged more rapidly in the colonies than in the homeland (CPCPapers 1: 15-16; 30 CSC: 147-148).

(90) Not every Hellenic polis outside the Greek homeland was a colony settled with a contingent of immigrants sent out by a metropolis. Many settlements were indigenous communities which became Hellenised by acculturation. Such settlements were neither founded by Greek settlers nor conquered and taken over by Greeks. They became Hellenised over a long period through immigration of individual Greek settlers and through regular contacts with neighbouring Hellenic communities (Inv.).

(91) Many of the major Hellenic colonies were indeed founded by mother cities situated in "Hellas" but others were founded by the colonies themselves, a phenomenon called secondary colonisation. In some cases even tertiary colonisation is attested. In overviews of Greek colonisation secondary colonisation is often mentioned in passing, but hardly ever discussed as an essential element of colonisation, and the ubiquitous distinction between colony and mother city does not always take into account that the colony becomes a metropolis when secondary colonies are founded. Secondary colonisation was particularly important in Sicily, Italy, Illyria and Libya where colonies founded by colonies outnumber colonies founded by poleis in "Hellas" (Inv.).

(92) It is true that most colonies were founded outside the Greek homeland but it must not be forgotten that quite a few colonies were placed in "Hellas" (Inv.).

(93) The extent of colonisation during the Classical period is mostly underestimated. Yet it appears that, including Athenian klerouchies, no less than 72 colonies were either founded or refounded in the fifth and fourth centuries, and it is also worth noting that the big metropoleis in this period were no longer Chalkis or Corinth or Miletos, but Syracuse and Athens (Inv.).

(94) All the evidence we have supports the view that, with very few possible exceptions, every colony was founded as a polis or became a polis not long after its foundation. Furthermore, there is no basis in the sources for the traditional distinction between apoikiai which were poleis and emporia which were not. Almost all, perhaps even all the communities of the Archaic and Classical periods that are described as emporia in the sources are attested as poleis as well (CPCPapers 4: 83-105; Article no. 3).

(95) The common claim that a Greek colony was independent of its metropolis must be modified. It is well known that the Athenian klerouchies and colonies remained dependencies of Athens, but many of the Corinthian colonies were dependent poleis too, and so were the secondary colonies founded by, e.g., Syracuse, Sinope and Kyrene (CPCPapers 4: 32-34; Inv.; see also nos. 45-46 supra).