Bryn Mawr Classical Review

Bryn Mawr Classical Review 2008.09.22

Hansen on Rzepka on M.H. Hansen, The Return of the Polis.   Response to 2008.02.46

Response by Mogens Herman Hansen, University of Copenhagen (

Towards the end of his review of M.H. Hansen (ed.), The Return of the Polis in BMCR 2008.02.46 Jazek Rzepka wrote: "We should note, however, that a vast majority of entries on individual cities in IACP [Inventory of Archaic and Classical Poleis] would not exist without information from the post-classical age". What he claims is, in fact, that out of the 1,035 entries a minimum of 600 and probably more would simply disappear from the inventory if the members of the Copenhagen Polis Centre team had not used Hellenistic and Roman sources extensively. But the inventory is intentionally restricted to a study of the poleis of the Archaic and Classical periods.1 If Rzepka's statement were correct, the Polis Centre project would have failed and the Inventory should not have been published.

Rzepka is not the only scholar who has censured the Polis Centre for extensive use of Hellenistic and Roman sources,2 but his criticism of the project is the most explicit I have seen so far, and I intend in this note to demonstrate that it is unfounded.

The Polis Centre's Inventory comprises a total of 1,035 entries, each describing a community which we believe to have been a polis in the Archaic and/or Classical periods. Updating the evidence we have now reached 1,040 communities which either indisputably or probably or possibly were Archaic and/or Classical poleis.3

One major purpose of our project was to investigate what the Greeks thought a polis was. Consequently, being called a polis in sources of the Archaic and Classical periods was our primary criterion for inclusion. As the evidence stands there are altogether 499 such communities (registered in the Inventory as poleis type A). Thus, included in the inventory are 541 communities not directly attested as being poleis in Archaic and/or Classical sources but known for properties and/or for having performed activities characteristic of a polis, i.e. characteristic of the 499 communities attested as poleis in sources down to the end of the Classical period. Such characteristics are, e.g., being a member of the Delian League or of one of the regional federations, having a mint, bestowing proxenia on a foreigner or having proxenia bestowed on one of the community's citizens, having a citizen as victor in one of the pan-Hellenic games, being perceived as patris, etc. Other characteristics -- such as to have a defence circuit, or a sizeable habitation centre, or a temple -- are included as important aspects of the community but they are not indisputable signs of polis status. If a community is known for one or more of the essential characteristics, it is listed in the Inventory as a polis type B, if it is known for one or more of the other characteristics, it is listed as a polis type C. There are 272 poleis type B and 269 type C.4

I have -- once again -- worked my way through the Inventory and checked the 541 communities not directly attested as poleis. To what extent is inclusion in the Inventory of these communities founded on sources of the Archaic and Classical periods and to what extent on sources of the Hellenistic and Roman periods? My investigation reveals that out of the 541 communities 489 are included in the inventory on the authority of a wide range of sources which are all of the Archaic and/or Classical period: 150 communities were members of the Delian League, 163 struck coins, for 370 a city-ethnic is attested; there are 45 instances of proxenia bestowed on a foreigner or having proxenia bestowed on one of the community's citizens, 42 instances of a theorodokos being appointed to host a theoros, and 16 instances of having a citizen as victor in one of the pan-Hellenic games. In 6 cases the community is perceived as patris. Citizens from 12 communities filled an office in the Delphic Amphiktyony. 12 communities are known as parties to treaties, and 4 had their city-ethnic inscribed on a kerykeion. In 3 cases a cult of Zeus Polieus, Athena Polias and Athena Poliouchos points to polis status. 67 of the communities are dealt with in inscriptions in a way that indicates polis status, and 58 in literary sources (Hdt., Thuc., Xen., Arist., Hell. Oxy., Ps.-Skylax etc). In 15 instances remains of a large fortified settlement combined with the attestation of a city-ethnic is the principal reason for inclusion. An index which summarises the results of my investigation can be found on the Polis Centre's home page.

Thus, for the great majority of communities classified in the Inventory of Archaic and Classical Poleis as poleis type B or C the inclusion is based on contemporary evidence. If we had restricted our investigation to Archaic and/or Classical sources, we would have to delete 52 entries out of the 541 for which polis status is not explicitly attested in Archaic and/or Classical sources. Thus out of a total of now 1,040 communities recorded in the inventory no more than 5 per cent are included on the authority of sources of the Hellenistic or Roman period.

Furthermore, in most of these fifty-two cases the post-classical sources used are "retrospective", i.e. the source in question is Hellenistic or Roman but treats events and institutions of the Archaic and Classical periods, and the information stems from earlier sources which, however, we cannot any longer identify. Let me adduce one example. The Third Sacred War was ended with the destruction of almost all Phokian poleis. According to Demosthenes (19.123) there were twenty-two poleis in Phokis altogether. Pausanias 10.3.1-3 lists the names of twenty poleis that were destroyed and one, Abai, that was spared. In Classical sources twelve of these are recorded as poleis and four more are known for activities characteristic of a polis. But for five of them the best evidence we have for polis status is Pausanias' account, which probably, as in many other cases, is reliable.5 On the authority of Pausanias, a Roman source, Jacques Oulhen classified these five communities as poleis type B, and, in my opinion, with good reason. In twenty-six other cases the authority for including a community in our inventory has been a retrospective source, either Polybios or Strabo or Diodoros or Pausanias or Arrian or occasionally some other late source.

In the remaining twenty instances a community has been included in the inventory on the basis of a Hellenistic source which concerns contemporary, i.e. Hellenistic events and institutions, but has been interpreted retrospectively by the member of the Polis Centre team and adduced as evidence for polis status in the Classical period.

Five Cretan communities illustrate this practice. In a treaty of the mid-third century B.C. between Miletos and some Cretan poleis, twenty-five Cretan communities are listed as subscribers to the agreements (Milet 1.3. 140). In the treaty they are classified as poleis and many of them are known as poleis of the Classical period as well. The author of this chapter of the Inventory, Paula Perlman, decided in five cases to interpret the treaty retrospectively and to include some of these early Hellenistic poleis among the Archaic and Classical poleis, viz., if their classification as poleis in the Hellenistic treaty could be combined with substantial physical remains of a settlement of the Archaic and/or Classical period (Inventory p. 1149). The poleis in question are Apellonia, Herakleion, Istron, Matala and Petra, and in the inventory they are cautiously classified as poleis type C, i.e. communities which were possible poleis.

So much for our criteria of inclusion and exclusion and the retrospective use of post-Classical sources. Another matter is that for all the communities included in the Inventory we sometimes add information that concerns the Hellenistic period. Thus, for each poleis we list both toponym and city-ethnic. If the city-ethnic is attested in post-classical sources only, we do give a reference but point out that the evidence is late; and the index of city-ethnics (pp. 1310-18) records only city-ethnics attested in Archaic and Classical sources. Similarly, for a number of poleis type B or C we report if they are explicitly attested as poleis in the Hellenistic period. But the post-classical evidence is excluded from the index of poleis (pp. 1298-1306) and from the analysis of the Archaic and Classical city-state culture in the introduction and in all the Polis Centre publications. Again, it is often mentioned that a polis had a defence circuit even if it is currently dated to the Hellenistic period. One reason is that the dates of city-walls are often in dispute, and in recent years many defence circuits have been updated from the Hellenistic period to the fourth century B.C. Also, we record evidence of Hellenistic cults, especially if they can be combined with remains of a sanctuary of the Archaic and/or Classical period. In every single case we do our best to date the sources we use so that users of the Inventory know where they stand.

To conclude: of all the 1,035 communities recorded in the Inventory of Archaic and Classical Poleis 95 per cent have been included on the authority of contemporary evidence, i.e. sources of the Archaic and Classical periods. Post-classical sources have been decisive for the inclusion of 52 communities only, and for 32 of these the authority is a retrospective source which explicitly refers to the Archaic and Classical periods. There are no more that 20 instances of post-Classical sources which have been interpreted retrospectively by the author of the chapter and adduced as the principal reason for inclusion in the inventory.


1.   Inventory 10-11.
2.   cf., e.g., Philippe Gauthier in REG 118 (2005) 240-1.
3.   See The Return of the Polis p. 44 with n. 83 and p. 52 with notes 3 and 4.
4.   Inventory pp. 7, 1298-1306; Return p. 52.
5.   See C. Habicht, Pausanias' Guide to Ancient Greece (2nd edn. Berkeley and Los Angeles 1998).

Comment on this review in the BMCR blog
Read Latest
Index for 2008
Change Greek Display
Books Available for Review

HTML generated at 13:33:10, Friday, 03 April 2009