Even the Athenians were not prepared to sacrifice efficiency to democratic principle in this most crucial of areas. The number 10 remained sacrosanct and so (probably) did the “one tribe, one general” principle, though later in the 5th century, and in the 4th, it was possible for one tribe to supply two generals, one of whom was elected at the expense of the tribe whose candidate had polled the fewest votes. Again, the object was to ensure maximum efficiency: there might be two outstanding men in one tribe. Another peculiarity of the stratēgia, to be explained in the same way, was that reelection, or “iteration,” was possible. (Actually it is not quite certain that the stratēgia was unique in this respect; it is possible that iteration was possible for the archonship as well.)
The Cleisthenic system was based on the trittys, or tribal “third.” There were three kinds of trittys to each of the 10 tribes, the kinds being called “inland,” “coastal,” and “city.” There were therefore 30 trittyes in all, and each of the 139 demes belonged to a trittys and a tribe. The numbers of demes in a tribe could and did vary greatly, but the tribes were kept roughly equal in population as far as one can see. (The last words represent an important qualification: it is just possible that the whole system was overhauled in 403 to take account of changes in settlement patterns effected by the great Peloponnesian War. In that case the evidence for deme quotas—evidence which is mostly derived from 4th-century or Hellenistic inscriptions—would not be strictly usable for the 6th or 5th centuries. But in fact there is just enough evidence from the 5th century to make the assumption of continuity plausible.)
Each of the 10 tribes supplied 50 councillors to the new Council. In that way, even the remotest deme was involved in what happened in the city; Cleisthenes’ solution can thus be seen in its political aspect as an attempt to deal with a characteristic problem of ancient states, which were mostly agriculturally based. That problem was to avoid the domination of city assemblies by the urban population. Cleisthenes’ system gave an identity to the deme that it had not had before, even though the word dēmos just means “the people,” hence “where the people live,” hence “village” (the word and concept certainly predate Cleisthenes). Now it had a more precise sense: it was an entity with an identifiable body of demesmen and a right to representation in the Council.
The Cleisthenic deme was the primary unit for virtually all purposes. It was a social unit: to have been introduced to one’s demesmen in an appropriate context was good evidence that one was a citizen. It was the primary agricultural unit—though it is disputed whether all settlement in Attica was “nucleated” (that is, whether all farms were clustered together around demes), as one view holds. In fact, there is much evidence for nonnucleated (i.e., isolated) settlement. It was, as stated, a legal unit—although deme judges were suspended from 510 to the 450s. It was a financial unit: temple accounts from the distant deme of Rhamnus date from well back in the 5th century. It was a political unit: as shown, it supplied councillors to the new Council and enjoyed a vigorous deme life of its own (though it seems that there was little overlap between deme careers and city careers). It was a military unit: not only did tribes train together, but a dedication by the demesmen of Rhamnus may show that they participated as a group in the conquest of Lemnos by Miltiades the Younger about 500 bce. (Another view puts that inscription in the years 475–450 and sees it as a dedication by cleruchs or a garrison.) Above all it was a religious unit: deme religious calendars, some of the most informative of them published in the 1960s and ’70s, show a rich festival life integrated with that of the polis in a careful way so as to avoid overlap of dates. It has been suggested that the worship of Artemis of Brauron, a predominantly female affair, was somehow organized according to the 10-tribe system. Finally, and related to the last, it was a cultural unit: at the deme festival for Dionysus (the “Rural Dionysia”) there were dramatic festivals, subsidized, as inscriptions show, by wealthy demesmen and sometimes even by foreigners (one wealthy Theban is attested).
Cleisthenes seems also to have addressed himself to the definition of the Assembly, or Ecclesia. As seen, Solon admitted thētes to the Assembly, but Cleisthenes fixed the notional number of eligible Athenians (adult free male Athenians, that is) at 30,000. One-fifth of this total, 6,000, was a quorum for certain important purposes, such as grants of citizenship.
Cleisthenes’ ulterior motive in all this must remain obscure in the absence of any corpus of poetry by the man himself, of any biographical tradition, and even of good documentary or historiographic evidence from anywhere near Cleisthenes’ own time (the Constitution of Athens is reasonably full, but it was written nearly 200 years later).
That the tribal aspect of Cleisthenes’ changes was central was recognized even in antiquity, but Herodotus’s explanation, that he was imitating his maternal grandfather, Cleisthenes of Sicyon, does not suffice as an explanation on its own. The question is why he should have been anxious that each Athenian tribe should be a kind of microcosm of all Attica. Politically, the tribe does feature in Athenian public life (for instance, tribal support in lawsuits was valuable, and each of the 10 tribes presided by rotation over the Council for one-tenth of the year. This is the so-called prytany system). But the tribe was not a voting unit like the Roman tribe—Athenian votes were recorded as expressions of individual opinion, not submerged in some larger electoral or legislative bloc—and the later political functions of tribes were not quite numerous enough to explain why Cleisthenes felt it necessary to subdivide them into “thirds” in the way he did.
Cleisthenes’ changes should be seen in their context. First, the Attica he inherited had a relatively small number of militarily experienced fighters, many of them former Peisistratid mercenaries. It was essential that these be distributed among the tribes if the latter were to be militarily effective. (It is a corollary of this that one accepts that at some preliminary stage in Cleisthenes’ reforms there was widespread granting of citizenship to residents of Attica whose status was precarious. There was surely plenty of immigration into prosperous Peisistratid Attica, not all of it military in character.)
Second, in the late Archaic period tribal reform took place in other communities, some far removed from Attica in both character and geography. Cleisthenes’ system looks subtle, theoretical, and innovatory in its decimal approach to political reform and its reorganization of “civic space,” but there were precedents and parallels. For example, at Cyrene, three-quarters of a century after its colonization by Thera, there was stasis (political strife), which Demonax, a reformer who was called in from Mantinea on the mainland, settled by reorganizing Cyrene into three tribes. Again, at tyrannical or possibly posttyrannical Corinth, it seems (the evidence is some boundary markers published in 1968) that there was a tribal reorganization along trittys lines not dissimilar to, but earlier than, Cleisthenes’ system.
Finally, there is the Roman analogy: the new system of tribes and centuries, a system based partly on residence, replaced a purely gentilitial system—i.e., one based only on heredity. The word century is a clue: although the term signifies a voting unit, it is military in character. It is evident that tribal reform was a fairly general Archaic solution to the difficulties experienced by states with large numbers of immigrants. Such states needed the human resources these immigrants represented, but they could not admit them under the old rules. The rules had to be changed.