K.H. Kinzl home page     TLG code for Greek text

On the Consequences of Following AP 21,3 (on the Phylai of Attika)*

Konrad H. Kinzl


Chiron: Mitteilungen der Kommission für Alte Geschichte und Epigraphik des Deutschen Archäologischen Instituts. München: C.H. Beck’sche Verlagsbuchhandlung. Band 19, 1989.

Thanks are due to Verlag C.H. Beck for permission to make this paper available on the World Wide Web.
© C.H. Beck Verlag, München





— 347 —




text of AP

1AP says explicitly no more than that "Cleisthenes" did not "create" twelve phylai which were identical with the twelve trittyes.2 There is no proof that AP offers a statement of fact. Scholarly prudence dictates the assumption that he does not,





— 348 —


and there are other reasons as well for not accepting AP's words as a statement of fact; whilst AP may indeed (albeit unconsciously) provide a valuable hint at the actual method applied to the reorganising process. AP must have known that reform of anything, and however modest (which the author of AP must have had an opportunity of observing sometime during his lifetime somewhere in Greece), is, in the best of circumstances, a messy job and an administrator's nightmare. If "Cleisthenes" "created" "afresh his ten tribes", and if he was in command of faculties of logic of at least a rudimentary kind, he ought to have been capable of coming up with a scheme in which logic, at least of a rudimentary kind, reigns. That scheme which we happen to be lucky enough to have documented to the extent that we can actually analyse it in fair detail, however, appears to defy both logic and any attempt by the application of logic to explain it. Chaos reigns supreme.




I shall now proceed to demonstrate the measure of that chaos by merely presenting what we know. In our knowledge we have come a long way, thanks to the work of Traill and Siewert (to name but two scholars, who have written on all this in book form).3 Traill has now abandoned the concept of random-size trittyes; and, it is hoped, the equalised-trittyes principle can no longer be reasonably doubted (one of the principal complaints in my earlier experimental paper need occupy us no longer).4

The "data" on which the present study is based are those of Traill, whose new map of the Organization of Attica. sums it all up. Whilst there remains a multitude of unsolved, or not definitively settled, problems of detail, these have no effective consequences for the purpose of the present paper. (No account has been taken by Traill of my contention that the conventional classification of the trittyes into "City", "Inland", and "Coast" is a figment of AP's imagination with no foundation in reality.5 I have no reason to change my view which I incorporate in the present study; my arguments, however, are not touched by its acceptance.) All this information is tabulated in summary form in "Table 1" below.





— 349 —


Table 1

Table 16

There are two types of trittyes, which I have named "Unitary Trittys" or UT, and "Composite Trittys" or CT.

A UT is made up of demoi which together form a single territorial entity. There are eighteen UTs. These may in turn adjoin to another UT, or to a demos or group of demoi from a non-UT within the same phyle, thus forming a territori-





— 350 —


al unit of a larger size than that of a single UT. A total of fifteen UTs do not stand completely alone within their respective phylai, and there are only three "free-standing" UTs (two in VII and one in X).

A CT consists of at least two demoi or groups of adjoining demoi (which in themselves are too small to be a trittys at the same time: the smallest has one, the largest possibly as many as twelve councillors). There are twelve CTs. The five City demoi located within the walls of Athens7 are all attached to CTs, and they belong each to a different phyle (II, Ill, IV, VII, VIII); three of these are in a phyle in which there is another demos or group of demoi just outside the walls in an adjoining location (II Kollytos, VII Melite, VIII Koile; with (3), (7) and (3) councillors, respectively).

Siewert accorded prominence to the concept of the "enclave",8 and he refined it to distinguish between three different kinds of enclaves. Traill, partly but with great economy, adopted the concept (if not the word).9 I shall here use the term, if only for the sake of convenience. For this concept to make sense, however, all demoi or groups of demoi which only together with other demoi or groups of demoi form a trittys must be so designated.10 Siewert and Traill, because they cling to the AP classification,11 designate only those demoi or groups of demoi as enclaves which are located outside what they think is their appropriate classification. I reject AP 21,4. The number of enclaves in my map is therefore significantly greater (c. [200], against Siewert's c. [80] and Traill's c. [40]).

The following "Table 2" tabulates the distribution and linkages of UTs and CTs by phyle.

Table 2

Table 2





— 351 —








— 352 —


Enclaves rarely stand alone; in most instances they border on other enclaves from other phylai.13 Conversely, CTs are rarely of the most straightforward kind (combining two enclaves); and phylai containing CTs (all except VI) are accordingly equally rarely of that kind.

What emerges from these tables is one simple fact. Logic of the kind which compelled those who use it to create a scheme in which virtually anything is possible (and actually occurs) does not in standard human activity occur: standard human convention does not term it "logic". Nor may we fault AP's logic, who merely reminds us that the ten new phylai were not the old twelve trittyes renamed phylai and, by extension (and because the author surely did not intentionally insult his readers by informing them that ten is not twelve), that chaos ensues if one turns twelve into ten (chaos that, incidentally, becomes no less chaotic by insisting that AP 21,4 is a statement of fact — which, I insist, it is not).14





— 353 —




It is not of course my aim, by demonstrating the "chaotic" nature of the new scheme, to indict the reformers15 who created this "chaos" as, e.g., ignorant of the topography or the demography of Attika.16 Rather, it is my contention that these reformers acted in precisely the way in which one might expect them to act, in the face of a task as daunting as that of the phylai reform, and with inadequate resources, under great pressures of time, etc. They erected their new edifice, not by first tearing down and razing to the ground its predecessor, but by retaining, recycling and reusing components of the old structure,17 to the utmost degree and as far as at all possible in the circumstances (the complexities of which we can barely fathom, and of which we catch only a few and rare glimpses), and given their objectives (which we can fathom to an even lesser degree, and of which we get hardly a glimpse at all): if all we see is chaos, this is only because we fall to see all that. The indictment of their having succeeded only in creating chaos vanishes once we succeed in comprehending their work as part of a historico-political process: a process of an organic growth of the new system from the roots of the old system which, as a consequence of the process of growth, has been nearly completely obliterated.

If, exempli gratia, one were to suggest that the present shape of the Canadian Confederation is owed to "logic" of the kind attributed to "Cleisthenes", one would rightly be accused of unhistorical thinking. Provinces and Territories, ranging from less than six thousand to almost three and-a-half million square kilometres: with boundaries that cut in a straight line across mountain peaks and valleys and the wide open prairie alike, and across linguistic or ethnic divides; which combine within one and the same province the sub-arctic tundra of Northwestern Ontario and the sub-tropical fruitlands of the Niagara Peninsula (to cite but a few examples). Confederation was "created", in 1867, "afresh", as one might say with more justification than in the case of Kleisthenic Attika. (Most of the present provincial and territorial boundaries date from later, with the present map of Canada drawn, with a few exceptions, most notably the addition of Newfoundland as the tenth province in 1948, by 1912.) No student of Canadian Confederation, however, would attempt its study without reference to its antece-





— 354 —


dents and the circumstances of its birth.18 There is likewise no justification for an approach according to which post-c. 506/05 BCE19 Attika was "created" "afresh", with no reference to what preceded it and to the conditions at the time — or indeed how it could possibly have been so created.

The Kleisthenic system as such is an undeniable fact. The evidence presented above is factual. The apparent "chaos" which I outlined, on the other hand, is not a fact but merely an attempt at unbiased interpretation. It is an interpretation which I deliberately put on it, in order to show the limitations imposed by the evidence. The evidence is not suited to provide answers to questions concerning the "why?" and the "how?". Why was this particular scheme devised? How was the end result, which is there for us to see, arrived at? Until evidence surfaces which may be regarded as unimpeachable, either literary (an improbable eventuality because literary evidence tends a priori to be of dubious value at best) or epigraphical (also unlikely because of the nature of the problem: official documents do not fill in the background but preserve only a statement of a reality which is intended), any answer to these questions must remain in the realm of speculation and hypothesis, rather than fact.

The complete absence of any evidence for the background and context of the reforms has not only not deterred scholars from putting forward their own hypotheses, in their attempts at answering the questions of the "whys" and "hows" (which is after all part of the scholar's "job"); it has also been intentionally ignored — and fantasy, dressed up as scholarly conjecture, has been presented in order to expose the most secret and devious schemes and manipulations or gerrymander20 by "Kleisthenes the Alkmeonid".21 It does not seem to bother some stu-





— 355 —


dents of Athenian politics that the opponents of Kleisthenes did not at the time see what they so clearly perceive, about two and a-half millennia later; or that these opponents were unable effectively to thwart him, because Athenian politics c. 506/05 BCE "functioned" the way Aristotle, thinking and writing in the changed world of Athens of one and a-half centuries later (and, incidentally, as a resident alien), wanted it to function, in order to provide him with proof of his political theory.22




A different path of enquiry, I suggest, ought to be attempted or, failing that, it must be conceded and frankly stated that the sources do not permit reconstructing the course of events23 about 506/05 BCE, let alone the forces driving them. If I be permitted my own fantasy, it would run something like this, with some premises stated first.

(1) The period from 511/10, from the passage of the first piece of legislation after the departure of Hippias, and of Kleomenes, to the implementation of the strategies reform in 501/500 BCE, must be viewed as one historical epoch. The bouleutic oath and the new strategoi are firmly anchored to demoi reform, phylai reform, and whatever other reform was or may have been carried out in that period. Kleisthenes, who vanished without a trace shortly after his recall in 507, and who apparently was no longer around in 501/500, was but one player (albeit a very prominent one) amongst a team of reformers who were perfectly capable of keeping up the pace of reform, as well as its direction, after his departure.24

(2) There are two separate thrusts of reform, aimed in two distinctly different directions, behind what is commonly referred to broadly as "phylai reform". The,v meet up in our perception only because they cannot be kept cleanly separated, involving one and the same body of people as they do: from the vantage point of posterity, they are merged into one. These two separate elements address two equally separate and distinctly different problems25 which became acute after the termination of the rule of violence of the last years of Hippias' presence in Athens, and because of the prospects for the immediate future. These are:

(a)  the citizenship question;

(b)  the reorganisation of the Athenian military.





— 356 —


(3) There are two historical peripeteiai,26 separated from each other by an incisive break, the intervention of Kleomenes in the archonship of Isagoras in 508/07 BCE, in these two areas of reform.

In the area of citizenship reform, the two phases are the "diapsephismos" of before and the demoi reform of after 508/07. The salient point of viewing events under this perspective is that a natural historical progression emerges which is otherwise lacking. It is by no means obvious why Athens brought in a procedure of administering its citizenship which, by any standard, represents a most radical departure from previous practice, in order to achieve an end that could have been accomplished by a more traditional procedure.27 A radical new method needs devising only once the means provided by a firm application of conventional practice have been exhausted. Had there been no "diapsephismos" in the years between 511/10 and 508/07 BCE, a "diapsephismos" might have sufficed after 508/07, and there would have been no need to go the radical route of the demoi reform. That route became obvious only because the most obvious one, that of a "diapsephismos", had already been tried (and, presumably, been found wanting).28

It appears logical therefore to surmise that in the case of the Kleisthenic phylai reform (one that also constitutes a radical departure from the past) too, a less radical measure of reform (which may not ever have progressed to the point of implementation) had been tried, and that the ground had been thereby prepared for reform which broke with tradition more completely, in the form of the new ten phylai. The threshold of tolerance for reform, as it were, had to be raised, and was raised, in order for the eventual radical measures to become thinkable and indeed acceptable.

It may indeed be an echo of all this if Herodotos has Kleisthenes bring in the new ten phylai and the demoi before Kleomenes entered to raise the sagging fortunes of his lover's husband Isagoras.29 At the time Herodotos gathered his information, there may still have been people around who recalled that, in the first phase after the expulsion of Hippias, events occurred that were more significant than the game of one-upmanship between Isagoras and Kleisthenes.30 Because he





— 357 —


could not put his finger on anything else, he promptly (although not very astutely for which he need not be blamed; many a modern historian would have done worse than Herodotos) redated the introduction of the new ten phylai to between 511/10 and 508/07 (which leaves him with nothing to report on Kleisthenes after the humiliation of Kleomenes, except that he was "recalled").31

Let us return to the phylai. In the crisis of 511/10, Hippias and only his closest family had been barely got rid of (by sheer luck, thanks to a hostage situation, not because of superior strategy).32 Kleomenes, presumably already into his cups, and into the books of oracles left behind (because of his hasty departure, or perhaps by design?) by Hippias,33 may already have developed an evil mood, barely mollified by the services of Isagoras' wife34 (perhaps Isagoras, anticipating things to come, had offered her to Kleomenes?). Notwithstanding the optimism of the hour, many a thinking Athenian must have grown fearful at the prospect of a reconciliation between Hippias and Kleomenes (which eventually came about),35 or what Hippias might try from the soil of, and with aid from, Persia36 (which, despite the brilliant diplomatic victory of the Athenian envoys in having their little containers of earth and water actually accepted by the Persian satrap,37 thereby cutting Hippias off at the pass, would eventually materialise in the Marathonian plain). Organising a military response must have loomed large in their minds, and with it the need for military reform. If, even after the phylai reform, further military reform (in the form of the strategia reform of 501/500) was deemed necessary, the situation before the phylai reform must have left a great deal to be desired in order to mount an effective defence.38 Siewert (although his thesis, in Traill's judgement, "sets the impossible task of reducing three, indeed four, dimensions to two"39) must be given credit for having drawn due attention to the military dimension of the Kleisthenic reforms.40

The effectiveness of military units depends (If an anti-militarist with no military training be permitted a guess) on their approximately similar sizes, their manageable sizes, and on their command structure.41 These requirements were not





— 358 —


met by at least the old trittyes which (on a none too improbable model, both based on kinship and scattered across their respective phylai) were of excessively unequal sizes. The old phylai, although perhaps not intolerably unequal in size, were of indubitably unmanageably large size. These "imbalances" were of little consequence in the administrative arena: it seems highly unlikely that the boule was under threat of ceasing to function unless their precise numerical equality was forthwith established (their prytanies of one hundred may have been recruited by phyle not by trittys times three: we simply do not know). If reform was a pressing priority, it was in the first instance because the old trittyes were rather unsuited as military sub-units.42 It was for this reason (although it is probable that there were others as well: we do not know) that it was impossible to go the route which AP suggests was rejected (i.e., a straightforward recycling of the old trittyes under the name of phylai; or a modified recycling): they were clearly too unequal in size. That disparity would, under the recycling approach, also have caused severe problems in the boule: in some trittyes there may not have been enough bodies qualifying for service on the council, especially if one thought (as one surely did at the early stages of planning) in terms of one hundred councillors from each phyle.43

The principles which must have been foremost on the minds of the reformers for their consideration — principles which are so obvious as to be regarded as highly probable, if not fact, rather than speculative or even fantasy — are:

(a)  a fixed number (preferably, but not of necessity, a larger number than the preexisting one) of phylai of (ideally — although, as we know, actual practice never matched the ideal of the original goal) an equal number of bodies;44 (b)  a





— 359 —


sub-division of the phyle, also of a fixed number per phyle (not of necessity three, even though that was the one eventually chosen — the trittys, three to the phyle), each approximating the fraction selected; (c) and, although of a lesser degree of probability, regional cohesiveness of both phylai and trittyes, where possible (neighbour fighting alongside neighbour in the phalanx might render it a more effective war machine).45 The consequences of the reorganisation for the boule will have taken a back seat, although they will not have been ignored: whatever the military reform meant for the boule, it may be assumed as also highly probable that no thought was given to the possibility of creating two different sets of subdivisions (or indeed divisions) — if there was to be a fixed number of phylai and trittyes with a fixed number (or fraction) for the total in each category, the same number of phylai and trittyes, with the same number (or fraction) for the total in each category must have been contemplated for of the composition of the boule, at least as the ultimate goal.46 This much may be stated with probability bordering on certainty. all else (unless irrefutable evidence, in the form which alone qualifies as such, i.e. inscriptions or, though unlikely, a reference in a contemporary poem, can be produced, to prove the contrary to be true) must be regarded as speculative. This is true in particular of the regional classifications in AP 21.3.47





— 360 —




I shall now put forward my own hypothesis. It is no more than a hypothesis, nor do I claim it to be anything other than a hypothesis. It is a hypothesis, however, which takes as its point of departure a historical model which is based on factual information only, its inherent principles as outlined above.

Just as, in the case of the administration of citizenship, the "diapsephismos" served, in retrospect, merely to "soften up" traditional attitudes, to the point at which the radical48 demoi based system was introduced, in the military sphere too such "softening-up" took place. Anxiety over what might happen next, mainly, we may assume, on the part of Hippias, caused a rethinking of the state of military preparedness and a realisation that "something had to be done."49 The phylai were (if they had not to a large extent already been)50 equalised,51 and the trittyes52 most certainly were equalised and, almost certainly altered at their base to





— 361 —


assume a more regional character. For this purpose some areas were broken out of their traditional context (thus creating a base for the forming of the later enclaves).53 It is conceivable that serious thought was given to increasing the number of phylai, in order to bring down the rather large totals for both phylai and trittyes. The number contemplated was that of five for the phylai and consequently fifteen for the trittyes. Each phyle had one demos in the city (it will be recalled that there are exactly five city demoi).54 The boule was obviously affected. Its total, without a change in the number of phyletai serving on its phyle contingents which continued at one hundred per phyle, was raised from four hundred to five hundred. If five phylai, were indeed considered, a larger block of territory had to be broken out of the traditional context than with four phylai. It is these areas which form the pool of demoi from which the enclaves under the ten-phylai system would be drawn (as we have seen above, c. 200 councillors fall into this category, whilst the remaining c. 300 continue to be combined in the territorially intact trittyes). In a five-phylai system with one hundred bouleutai from each phyle, a trittys numbers (33/34) councillors. The new trittyes of the Kleisthenic system which we call UTs can be easily created by merely dividing up into two exact halves (16/17) a trittys of (33/34).55 Eighteen UTs of the final scheme de-





— 362 —


rive from nine in a five-phylai system. Six of the new ten phylai have UTs bordering on each other and which were created by halving six trittyes of the five-phylai arrangement. If there was such an intermediate stage with five phylai and fifteen trittyes, the ten phylai were formed by reusing ten trittyes under the name of phylai (or two-thirds of the total of fifteen trittyes), and by redistributing the demoi of the remaining five (one-third) trittyes amongst these. This then accounts for the two: three ratio of 200 councillors from CTs (the latter group) and 300 from UTs (the former group). This process also leads in a very natural way to the fifty bouleutai contingent from each of the new ten phylai: two Uts, with (33/34) councillors of one five-phylai trittys of (33/34) councillors, plus one CT of (16/17) councillors formed with demoi in the pool of demoi in the remaining five trittyes — (5 x 33/34): 10 = (16/17). In itself, the change from 100 to fifty as the new size of the prytany does not appear an unavoidable one (although it might have been more difficult to recruit one thousand bouleutai annually than five hundred).

None of this, as we very well know, ever left the drawing board to be implemented. In its planning stages, Isagoras and his obliging or obliged friend Kleomenes staged their coup. It is surprising that Herodotos reports no military response to Kleomenes' intervention: it was left to the citizens in the boule of four hundred to organise the opposition and rally the support to defeat Kleomenes.56 Was the military in disarray, lacking in organisation because of the





— 363 —


vacuum left between old and new? Once the Kleomenian spook had vanished, military reform became an even more pressing issue. Kleomenes could be expected to be back. Hippias was of course still hovering in the background, and with him the Persian. The short-term success in outmanoeuvring Hippias by their own alliance with Persia could not be foreseen, nor was it welcomed by all quarters, and it was repudiated by the Athenians themselves in short order. 57

I The reform which was eventually implemented was very radical, if we view it against the background of the old four-phylai structure. If, however, one had already begun drawing up plans for reform that went beyond the old four phylai, one would now begin thinking of reform which would go beyond what had hitherto been regarded possible and indeed acceptable. The ten-phylai system may not have seemed at all radical after what had already been advanced at the intermediate stage. As well, Isagoras' own attempt at reform, however unwelcome it was, may also have contributed to rendering the thought of further reform of the boule less unpalatable (and, perhaps, to discrediting, as it were, the figure of one hundred for each contingent to be sent to the council).

At the same time, the question of how best to administer citizenship was tackled again. It found its own radical solution (although somewhat less radical if viewed against the background of the already far-reaching measure, of conducting a "diapsephismos", of before 508/07 BCE).




The rest is history, as they say. Ultimately, the two separate sets of reform, eventuated at about the same time but for completely different reasons, by mere historical coincidence were married, to live happily ever thereafter. The citizen is the soldier; the councillor serving on a trittyic contingent in the prytany did so because the military unit called the trittys also served a civilian rôle. It is posterity, however, which made the most lasting contribution to the fusion of the two separate sets of reform. Once the detailed vision, which allows one to see distinctions in clear contour, became blurred because of the passage of time, all that one could perceive from the distance of a later age was one single and uniform apparition on the horizon of the past. Thus the phylai reform, as presented by all ancient authors from Herodotos onwards, comprising everything from the new 'for the new ten phylai to the new names for the citizens (the demotics), the council of five hundred, etc., was born — entirely in the pages of the history books, rather than in the very real world of 511/10 to 501/500 BCE Athenian politics.

As far as the five phylai are concerned, however, all this is and must remain hypothetical — a historical model (of perhaps, as I would hope, a certain degree of





— 364 —


likelihood). A model, however, which is, it must be stressed, based entirely on what we know (i.e., on the epigraphical evidence and the few facts which can be saved from the literary works), whilst ignoring those statements in the literary sources which cannot pass the test of factuality and must be disregarded, as ancient hypothesis, speculation, or plain fantasy. It is a model therefore which attempts not to build fantasy upon a foundation of fantasy but one which, if it be termed fantasy, deserves at least to be recognised as fantasy built upon a foundation of facts.




Let me sum up: if my positive contribution is dismissed as fantasy, and therefore as in fact representing no contribution at all, there is also a contribution which, although it may be termed negative, none the less ought to further the cause of scientific history because it is not fantasy. AP 21.3 cannot be treated as if it were a statement of fact. The notion, based on a narrow interpretation of AP 21.3 and pressing its "evidence", that the reformers of c. 506/505 BCE "organised the citizen body afresh" cannot be maintained. They did not work from a tabula rasa. Had they worked from a tabula rasa, they would have performed one of the most extraordinary feats in the history of logic, by producing through the application of logic results that defy logic, which are beyond comprehension, and whose logic can be explained only by creating fantasy posing as scholarly conjecture. A tabula rasa approach surely would have produced a "clean" system, not the one illustrated in the opening pages of this study (with "Tables 1 and 2" and "Map").58

The real onus of proof is on those who persist in, and continue, ignoring the principles recently restated by the late Sir Moses:59 that "without a theoretically grounded conceptual scheme, the thin and unreliable evidence lends it-self to manipulation in all directions, without any controls;" that "the modern historian of antiquity cannot simply repeat ancient practice;" that "anything written in Greek or Latin is somehow privileged, exempt from the normal canons of evaluation;" that "with the passage of time, it becomes absolutely impossible to control anything that has been transmitted when there is nothing in writing against which to match statements about the past."60





— 365 —


Whoever attempts to recreate Athenian history on the basis of AP "creates afresh" whatever he elects to. The statistical probability of this creation's resembling anything that actually occurred about 506/05 BCE is infinitesimally small. The fact that these creations still fill the pages of our books and journals is due to the need to fill — just as "Livy and Dionysius of Halicarnassus" "narrated in detail" "the early centuries of the Republic and the still earlier centuries that preceded it," because "like historians ever since, [they] could not tolerate a void, and they filled it one way or another, ultimately by pure invention."61







*     This paper has benefited from several scholars. John Traill provided, by his Demos (n. 3 below), stimulus but also read an earlier draft; by not dismissing it out of hand, he unwittingly encouraged me to consider publication. P.J. Rhodes saved me from attributing to him a view he does not hold and, by his comments, generally aided me in improving the presentation of my ideas. M.H. Hansen generously gave of his time and caused me to add several cautionary flags (in addition to the virtually unchanged present n. 44) on matters of demography. Last but not least, E. Badian provided encouragement. Not one of these scholars may be presumed to agree with any one of my points (and one reader must [be mercifully] accorded anonymity).

Return to Text

1     Aristoteles, *AQHNAI/WN POLITEI/A, ed. M. Chambers, Leipzig 1986, p. 19, 3-7.

Return to Text

2     P.J. Rhodes, A commentary on the Aristotelian Athenaion politeia, Oxford 1981, p. 251, notes that "A.P.'s logic is defective, for if he had wished to do so Cleisthenes could have created twelve new tribes, not by calling each of the old trittyes a tribe but by organising the citizen body afresh as he did to create his ten tribes". Siewert (n. 3 below) offers (at 126-8) a far more measured assessment (under the sub-heading "Die regionalen Trittyen im irreführenden Bericht [my italics] der Athenaion Politeia"), concluding that "durch seine Unvollständigkeit führte das aristotelische Zeugnis über die Trittyenbildung in die Irre [my italics]" (128). — The emphasis in O(/PWS ... MH\ SUMBAI/NH| MERI/ZEIN KATA\ TA\S PROU+A/RXONTAS TRITTU=S surely is on PROU+A/RXONTAS and not on TRITTU=S, for which FRATRI/AS (or even PO/LEIS) might be substituted which are also PROU+A/RXONTA ME/RH. See also H. Hommel, Trittyes, RE 7 A (1939), 342-6, on units, totalling twelve, from early times; also ibid. 354, on the "Zwölfstädte" of the Etruscans, the four tribus and twelve curiae of Mantua in Verg. Aen. 10,201-3 with Servius, and the twelve BASILH=ES of the Phaiakes in Hom. Q 390-1 and LexPatm s.v. GENNHTAI/, p. 152 Sakellion = LexGrMin p. 162 Latte, from the lost beginning of AP. It shows that AP was familiar with the thought of PROU+PA/RXONTA ME/RH of whatever name, totalling twelve (the historical truth in these passages is not what concerns us here; for a thorough examination see Bourriot [n. 21 below], 460ff.).

Return to Text

3     J.S. Traill, The political organization of Attica, Princeton 1975. P. Siewert, Die Trittyen Attikas und die Heeresreform des Kleisthenes, München 1982; cf. my review Gymnasium 93,1986,556-8. J.S. Traill, Demos and trittys, Toronto 1986; cf. my reviews CR [39,1989,67-69], and Gymnasium [96,1989,560f.].

Return to Text

4     Traill, Demos (n. 3). K.H. Kinzl, "On the consequences of following AP 21.4 (on the trittyes of Attika)", AHB 1,1987,25-33, esp. 28. Cf. also W. Schuller, review of Siewert (n. 3), GGA 236,1984,11-21, at 18: "das wichtigste Prinzip, das als Sachverhalt gewiß richtig ist, nämlich das der numerischen Einheitlichkeit der Trittyen". D. Whitehead, The demes of Attica 508/7 — ca. 250 B.C.: a social study, Princeton 1985, xxii appears to concede "that the traditional, geographical trittys" is out.

Return to Text

5     See n. 11 below.

Return to Text

6     Trittys names are given following Traill, Demos (n. 3) "Color Map" (known names are underlined, conjectured names are underlined with dashes; bouleutic quotas are in parentheses). || separates a "free-standing" trittys from the others, or "enclaves" from one another; {and}, and — > — > — >, indicate that a trittys has a common boundary with either another trittys, or with an "enclave" from another trittys; || + + + means that an "enclave" is joined together with, but not in an adjoining location to, a demos in Athens City (on this see n. 7 below). Demoi that cannot be placed on the map appear in the right-hand boxes (in several lines only for reasons of space, not in order to suggest correlation to any of the trittyes in the boxes to the left). — I have allowed myself one arbitrary assumption. Even though one bouleutes from the demos Diomeia (just outside Athens City) is a long distance from the rest of the trittys I have nonetheless classified Halai as a UT. I also arbitrarily ignored demoi which cannot be placed on the map in my considerations. Most of them rate (1), the largest (2/3). Their total quotas add up to no more than between 3.8 and 4.6 percent of the total of five hundred. (It is also conceivable that one or the other such "demelet" is similar to Diomeia; I do not think that this would affect my thesis in a significant way.)

Return to Text

7     These walls were of course constructed after the Persian Wars. It might be prudent to assume that the area enclosed by the walls of 479/8 represented an area which, for many decades, had been thought of as "The City".

Return to Text

8     Siewert (n. 3) "Karte 4"; an elaborate glossary at 131 beneath the text, (3)-(5). Cf. AHB 1,1987,29.

Return to Text

9     Traill, Demos (n. 3) "Color Map".

Return to Text

10     See my "Map" with n. 12 and n. 13 below.

Return to Text

11     The main contention of AHB 1,1987,25-33 is that the regional classification in AP 21,4 has no foundation whatever in fact, and must be discarded. See also below n. 23.

Return to text

12     This Map is drawn after the map of the *NOMO/S *ATTIKH/S published by the *EQNIKH/ UPHRESI/A THS *ELLA/DOS, 1971, and reduced from 1:200 000 by photocopying machine. Solid black squares, and squares in outline, are drawn in the locations of demoi of known name and location, and of known location but uncertain demos name, respectively, on the basis of Traill, Demos (n. 3) "Color Map". Roman numerals indicate UTs, small Roman numerals indicate the phyle association of enclaves. Bold solid lines enclose regions (shaded) made up of "enclaves" (in all but three instances — I Iphistiadai, IV Sounion, and IV Deiradiotai? and ?Potamos Deiradiotes — from two or more phylai); solid lines separate enclaves in such regions from one another. Hatched lines separate UTs from one another. Readers are asked to pay attention to the broad principle, not to inaccuracies of detail, of which there are no doubt many, owing to my lack of topographical experience.

Return to Text

13     I would note that even the most experienced topographer may in some instances find it difficult to determine whether, e. g., VI Thria and VI Lakiadai do indeed border on each other, or whether perhaps they are separated by a joining together of the enclaves of V Kholargos and IV Diakris. I also wonder if IV Phrearrhoi did not indeed represent a UT, rendering V Thorikos a CT (which would mean that the smelting centre of the demos Thorikos was not connected to the ore-carrying hinterland, and that it was also cut off from the other demoi in the same trittys located farther inland). That would of course change the picture presented in "Table 2". If these two hypotheses hold any truth, and if one were to add to it a third one along similar lines, by connecting the neighbouring enclaves of IX Rhamnous and X Eitea with X Semachidai? (and thereby with the band of enclaves adjoining it), a most startling picture emerges. There is one vast crescent of "enclavic" territory, from the far north-east to the far south-east. It stretches from Rhamnous through the heights separating the Marathonian plain from the interior (briefly touching the sea again where the mountains leave only a narrow passage between the plain of Marathon and the mesogeios), the peak of Mt Brilessos, cutting across to the foothills of Mt Parnes (but not including, it seems, its summit; nor the gorge leading upwards to Phyle), along the entire ridge of the Aigaleos range, taking in the harbours of Athens, Athens City with the flats in between, the Hymettos range from the col between Mt Brilessos and Mt Hymettos with X Pallene to the north to its southernmost peak, and much of the rocky terrain in the south-eastern triangle, with X Phrearrhoi, IV Sounion, and IV Deiradiotes? on the south-east coast. Athens City, the main harbours, the mountains with their mineral resources — a glance at the map of modern Attika shows that these regions also largely coincide with those marked as wooded — and quarries, the sanctuaries of Poseidon at Sounion and of Nemesis at Rhamnous ... : all this ought to mean something.

Return to Text

14     See n. 11 above.

Return to Text

15     I refuse to attribute all and everything to the towering giant Kleisthenes but prefer to speak of "reformers" and "Kleisthenic reforms", cf. Kinzl, "Athens: between tyranny and democracy", in: Greece and the eastern Mediterranean ..., Berlin etc. 1977, 199-223, esp. 202-10, = "Athen: zwischen Tyrannis und Demokratie", in: Demokratia (n. 22 below), 212-247.

Return to Text

16     C.G. Starr, Individual and community..., New York etc. 1986, 91, notes the "clear geographical and political knowledge of Attica" behind the reforms (citing P. Lévêque and P. Vidal-Naquet, Clisthène l'Athénien, Paris 1964, in support, at 125 n. 14).

Return to Text

17     Some incipient recognition indicated by J. Bleicken, Die athenische Demokratie, Paderborn etc. 1985, 30: "Da dies an sich schon komplizierte System auf gewachsene Siedlungen Rücksicht zu nehmen hatte [my italics]".

Return to Text

18     See c. g., G.J. Matthews, R. Morrow, Canada and the world: an atlas resource, Scarborough, Ontario, 1985, maps "Territorial evolution", 5f.; W.L. Morton, The critical years: the union of British North America 1857-1873, Toronto 1964; K. McNaught, The Pelican history of Canada, Harmondsworth etc. 1969, esp. ch. 8-10, 106-53; also, G.F.G. Stanley, The birth of western Canada, Toronto 1960; W.L. Morton, Manitoba, Toronto 1957. (I should like to thank Professor John Wadland of the Canadian Studies Programme at Trent University for his advice.)

Return to Text

19     I.e., sometime between the resumption of reform (after the good riddance of Isagoras and his Lakedaimonian bedfellow, with the general climax of crisis occurring in the War on Three Fronts) and the first administering of the bouleutic oath in probably 501/500 (P.J. Rhodes, The Athenian boule, Oxford 21985, 191 with n.3 etc.; R.A. Delaix, Probouleusis at Athens, Berkeley etc. 1972, 22 f. with n. 58).

Return to Text

20     Starr (n.16) very soberly judges it "difficult in looking at the new map of Attica to detect deliberate gerrymander on a wide scale" (92).

Return to Text

21     E.g., esp., G.R. Stanton, "The tribal reform of Kleisthenes the Alkmeonid", Chiron 14,1984,1-41; P.J. Bicknell, Studies in Athenian politics and genealogy, Wiesbaden 1972, esp. ch. 1 "Kleisthenes the politician" 1-53. Also G. Daverio Rocchi, "Politica di famiglia e politica di tribù nella polis ateniese (V secolo)", Acme 24,1971,13-44, at 20-44 (lit.!); cf., contra, F. Bourriot, Recherches sur la nature du génos, 2 vols, thèse Paris I 1974 (1976) 380 ff.

Return to Text

22     Cf., e.g., E. Ruschenbusch, Athenische Innenpolitik im 5.Jh. v. Chr., Bamberg 1979, 10, etc. (cf. my review AAHG 39 (1986) 188-91); id.. "*P*A*T*R*I*O*S *P*O*L*I*T*E*I*A", Historia 7 (1958) 398-424 = In: Kinzl (ed.), Demokratia: der Weg zur Demokratie bei den Griechen, Darmstadt 1985 (Wege der Forschung 657), 87-124.

Return to Text

23     If I may be permitted to restate this view, cf. AHB 1 (1987) 33.

Return to Text

24     This line of arguing is pursued op. cit. (n. 15).

Return to Text

25     I failed to see this while working on ibid. As far as I can ascertain, no one else has.

Return to Text

26     A line of argument developed ibid.; on the rôle which the "diapsephismos" (a terminus technicus which I do not believe was used at the time for a procedure which it may not accurately describe altogether) served, on this model, ibid. 200-1 with nn. 5ff. (lit.!).

Return to Text

27     A comprehensive investigation of the administration of citizenship in ancient Greece is a much needed desideratum, partly filled by C.B. Patterson, Pericles' citizenship law of 451/0, Diss. Univ. of Pennsylvania 1976 [published as a monograph: Pericles' citizenship law of 451/0 B.C., New York 1981; also: P.B. Manville, The origins of citizenship in ancient Athens, Princeton 1990]. Cf. R. Sealey, "How citizenship and the city began in Athens", AJAH 8 (1983) 97-129. There are some very good observations by Bourriot (n. 21) 212ff.

Return to Text

28     We have been forcefully reminded of this by D.M. Lewis, "Cleisthenes and Attica", Historia 12 (1963) 38.

Return to Text

29     Hdt. 5,66,2; 69,2 on Kleisthenes and the phylai and demoi; 5,70-2 on Kleomenes and Isagoras.

Return to Text

30     Hdt. 5,66,1f.

Return to Text

31     Hdt. 5,73,1.

Return to Text

32     Hdt. 5,65,1-2.

Return to Text

33     Cf. Hdt. 5,90,2.

Return to Text

34     Hdt. 5,70,1.

Return to Text

35     Hdt. 5,90-4.

Return to Text

36     Hdt. 5,96.

Return to Text

37     Hdt. 7,73.

Return to Text

38     Cf. broadly on all this Siewert (n. 3), esp. ch. IV, 154ff.

Return to Text

39     Traill, Demos (n. 3) p. iv n. 2. Cf. n. 40.

Return to Text

40     Cf. on this aspect Traill, "Diakris, the inland trittys of Leontis", Hesperia 47 (1978) 89-109 with pl. 1, at 109; also H. van Effenterre, "Clisthène et les mésures de mobilisation", REG 89 (1976) 1-17.

Return to Text

41     Reviewers of Siewert (n.3) remain sceptical. E.g. Schuller (n. 4); A. Andrewes, CR 33 (1983) 346-7; D.M. Lewis, Gnomon 55 (1983) 431-6; my review (n. 3). It is also being acknowledged that previously less than due attention may have been paid to the military aspect.

Return to Text

42     F. Frost, "The Athenian military before Cleisthenes", Historia 33 (1984) 283-94, appears to suggest that, if there was any mechanism of mobilising an army, it was never activated in a regular mobilisation ("unless the army that went to Plataea in 519 was such a force", 293).

Return to Text

43     I see no reason to call the existence of the Solonian boule of 400 into question. See, e.g., Rhodes (n. 19) 208-9 etc.; de Laix (n. 19) 13-7; Rhodes (n. 2) at AP 8.iv (pp. 153ff.). I feel strongly tempted to speculate that the Solonian constitutional reforms did not stop short of phylai reform of some kind; my hunch is that it entailed at least some tinkering with the phylai and trittyes — Solon was not after all oblivious of military concerns (FF 1-3 West). If the phylai are kinship-based of old, there may have existed massive numerical disparity amongst the populations of the ancient phylai and scarcely any degree of regionality. Is it not conceivable that the reforms of Solon, whilst retaining both the ancient number of four and the ancient Ionian names, altered the basis of the phylai from one of ancient kinship alone to one that, at least partially, incorporated a regional element? M.H. Hansen (in a personal communication of April, 1988) would not rule out "important changes in the size of the demes in the period 500-350" — a similarly unattested yet probable change as the one I am proposing for Solon.

Return to Text

44     We must emphasise that the principle of numerical equality cannot be pressed, because of demographic factor. Whilst it is possible to suggest that the military trittys was intended to consist of 300 fighting men (cf., more dogmatic, Siewert (n. 3) 142ff.), and although we do know that the bouleutic trittys did indeed number fifty, the actual citizen population of each trittys and phyle cannot have been exactly equal. Many if not most councillors may have been beyond fighting age, and most men of fighting age will have been too young to serve on the council; and some may have been too old or too sick to serve on either, cf., e.g., Rhodes (n. 19) with nn. 7-8. One also wonders if life expectancy varied from region to region, and whether, by generations' experience, these variations had not become common knowledge to be taken into account in all this. We may surmise that the Akharneis, in the charcoal making occupation, had a low life expectancy, dying of asthma, emphysema, lung cancer, etc., at a relatively young age. There may have been vast numbers of twenty year old Akharneis, to serve in the military (perhaps the "three thousand" of Thouk. 2,20,4 is not as grossly inflated as is commonly assumed, e.g. Gomme HCT 2, 73f., ad locum), but only a disproportionately small number of them reached the age at which they would serve on the boule. The bouleutic quota of (22) for Akharnai may be a reflection of this condition, with the added consequence that these quotas must be regarded a very unreliable indicator in extrapolating total population figures amongst the citizenry.

Return to Text

45     Frost (n. 42) 285, however cites kinship as a more important motivating force (referring us to Homer *B 362).

Return to Text

46     Cf., e.g., Siewert (n. 3) 24; 129; 130. His arguments at the same time ought to weaken Traill's own exercise in conceptual refinement (such as "modified geographical trittyes", etc.; see, e.g., Traill (n. 40) n. 79) (on Siewert's finely demarcated types of "enclaves" see above with n. 8).

Return to Text

47     See above n. 11, etc. The maps of both Siewert (n. 3) "Karte 4", and Traill, Demos (n. 3) "Color Map", are entirely based on this classification.

Return to Text

48     Not of course in the sense of the "radical democracy" — a concept which would provide an excellent topic for an investigation in the footsteps of W. Gawantka, Die sogenannte Polis, Stuttgart 1985 (cf. my review article EMC ns 7 (1988) 403-12). I have avoided using this term in a long time, and am pleased to note B.S. Strauss, "Athenian democracy: neither radical, extreme, nor moderate".

Return to Text

49     Phylai reform with a military component and without a change in the number of phylai could be accomplished at Sikyon by Kleisthenes, see Kinzl, AJAH 4 (1979) 27-8 with 41 n. 27 [revision of id., in: Die ältere Tyrannis ..., Darmstadt 1979 (Wege der Forschung 510), 303-5 with n. 27]. K.-H. Welwei, Die griechische Polis, Stuttgart etc. 1983, 86-7 with 309 n. 90 accepts and partly paraphrases my arguments, including my derivation of the Sikyonian phylai names from names of eponymous heroes (on the other hand, A. Griffin, Sikyon, Oxford 1982, 51 with moving credulity serves up again the Herodotean "punning" version; cf. my review Gymnasium 91 (1984) 423-5); cf. also Kinzl, Klio 62 (1980) 181.

Return to Text

50     Starr (n. 16) 56, in his discussion of the Great Rhetra ("issued in the first decades of the seventh century"), points out the "amazing recasting of people ... into territorial wards", and continues by contrasting this to "Sicyon, Corinth, and Athens", where this "was not to take place for at least two or three generations" — does he mean to imply a Solonian "recasting" at Athens, as I would be prepared to suggest (see above n. 43)?

Return to Text

51     In this and the following clause, "equalised" must be taken cum grano salis, see above n. 44 with its cautionary remarks.

Return to Text

52     *TRITTU/S signifies one of three sub-units of the same name, and qualitative, but not necessarily quantitative, characteristics. H. Frisk, Griechisches etymologisches Wörterbuch vol. 2, Heidelberg 1970, s.v. (who gives an interpretation, rather than an etymology: "der dritte Teil einer Phyle"); Hommel (n. 4) 331-2. (A three-victim sacrifice, or a three-contest victory, are the other meanings.) The -tt- formation might point to ancient Attic origin, but cf. TRIPTU/S (Keos) and TRIKTU/S (Delos), Frisk l.c. Tracing other -ys formations is nearly impossible: P. Kretschmer-E. Locker, Rückläufiges Wörterbuch der griechischen Sprache, Göttingen 31977, fail to list even TRITTU/S! Quoting the ancient lexicographers is not always helpful, in particular if one is not fully conversant with the intricacies of lexicographical tradition. In any event, *LE/CEIS R(HTORIKAI/ s.v. (I. Bekker, Anecdota graeca, Berlin 1814 [= Graz 1965], vol. 1: Lexica Segueriana (preserved in cod. Seguerianus siue Coislinianus 345) 304, 24), explains TRITTU/S: TRI/TON ME/ROS TH=S PHULH=S; verbatim the same s.v. FRA/TORES KAI\ FRATORI/A KAI\ FRATORI/ARXOS (AnGrBekk 1,313,28) for FRATORI/A C.W.J. Eliot, "Aristotle Ath. pol. 44.1 and the meaning of trittys", Phoenix 21 (1967) 79-82 misses the point. No one would argue that, because of its etymology or more ancient meaning, FULE/ cannot designate precisely one-tenth of citizen body, bouleutai, hoplite army, and must mean something else. By the same token there is no reason why TRITTU/S must mean, on purely linguistic grounds, something other than precise thirds and cannot therefore mean these precise thirds (or, to be more exact: 17 + 17 + 16). See also AHB 1 (1987) 27-8 with nn. 13-14.

Return to Text

53     See my "Map", areas enclosed by bold lines, shaded.

Return to Text

54     Cf., however, above n. 7.

Return to Text

55     This hypothesis might go a long way towards explaining the anomaly of Akharnai with (22) councillors, cited by everyone who doubts the exact-thirds meaning of trittys, e.g., Eliot (n. 52) 85-6 with n. 18 (and most recently in personal conversation, at the Winnipeg meetings of the Classical Association of Canada, 1986 06 01), as the "spoiler"; cf. AHB 1 (1987) 28-9 with nn. 18-19. Traill attempts a novel approach. In "Diakris" (n. 40) he stated that "there is no evidence that [Akharnai] was a split deme" (105); ibid. 105 n. 53, however, he speculated that Akharnai, "after the fashion of Aphidna, held a number of smaller communities within its local jurisdiction", and that, "although Acharnai was not technically a divided deme, it may have occupied two different sites" (l.c.) identified by E. Vanderpool, *XARISTH/RION EI)S ... *ORLA/NDON vol. 1, Athens 1965, 172; cf. Traill POA (n. 3) 50. In the "Appendix" (in which category, as in the "Addendum" to "Diakris" (a href="#40">n. 40) 109, one tends to find Traill's most fruitful ideas) to "An interpretation of six rock-cut inscriptions in the Attic deme of Lamptrai", in: Studies in Attic epigraphy, history, and topography, presented to E. Vanderpool, Princeton 1982, 162-171 with pl. 21, Traill suggested "that for military and political purposes a portion of Acharnai was transferred" (169). This view is now adopted without the earlier reservations in Demoi (n. 3): "Acharnai is now treated as a split deme" (123); incorporated in his new "Conspectus of deme quotas and locations" at 133: "... composite deme; the smaller of the 2 known sites ... has been chosen for a city section of Acharnai"; and argued in detail in the new discussion of Agora 15,68 (142-4). Traill admits that there is no evidence in support of his thesis but points out that there is also no evidence against it. An anomaly it appears to be none the less, but one that might be less difficult to understand, as far as its origins are concerned, on my hypothesis: Akharnai, on this hypothesis, was part of a (33/34)-sized trittys and had its representation more or less fixed at that large number (22) before trittyes of (33/34) were split in half to contain (16/17) councillors each. Reducing the quota for Akharnai was out of the question: who would have dared facing an angry crowd of Akharneis, their faces painted black with charcoal (they nearly succeeded, in the summer of 431, in forcing Perikles to break with his strategic plan, Thouk. 2,20-2, esp. 21,3)?

Return to Text

56     Cf. van Effenterre (n. 40) 4. Herodotos no doubt speaks of the boule of 400 (on which cf. above n. 43). It is inconceivable that the Spartan basileus (to heed the cautioning against the use of the word king by R. Drews, Basileus, New Haven, London 1983, esp. 78 ff.; cf. my review EMC ns 3 (1984) 84-7, at 85) would have attempted to abolish the ancient and venerable equivalent of the gerousia (no matter how deep he was into his cups, and how dearly he might have loved to abolish some government body at home, which would of course be the ephoroi) the Athenian boule of the Areios pagos. It would also seem incongruous to suggest that, in those hectic days, Kleomenes and his limp Athenian bedfellow would have attempted to reorganise something equally venerably ancient as the phylai, with all the attendant complexities of such an attempt. The proposed boule of 300 (Hdt. 5,72,1) makes sense only as an attempt at setting up an oligarchic system, drawing one hundred bouleutai each from amongst pentakosiomedimnoi, hippeis, and zeugitai for a total of 300 (Hdt. admittedly speaks only of Isagoras' SUSTASIW=TAI), perhaps in combination with an ekklesia of one thousand from each of the three TE/LH, to total 3000 — thus excluding from both probouleusis and ekklesia the majority of Athenians, in the TE/LOS of the thetes.

Return to Text

57     Hdt. 5,73,3.

Return to Text

58     Cf. AHB 1 (1987) 33.

Return to Text

59     M.I. Finley, Ancient history: evidence and models, London 1985, ch. 2 "The ancient historian and his sources", 7-26 ("a substantially shorter version of this chapter appeared in Tria Corda. Scritti in onore di Arnaldo Momigliano, ed. E. Gabba (Como 1983), pp. 201-14", 109 n. 1). The quotations are at, respectively: 18; 8; 10; 16.

Return to Text

60     For an excellent example of Finley's complaint, see Sealey (n. 23) 121f.: "It may be that the oral tradition had become somewhat distorted by the time it reached Herodotos. But his narrative is intrinsically plausible and there is no positive reason to doubt any of it [my italics]." It would appear to me that, once a tradition has become "somewhat distorted", there is "no positive reason" to believe "any of it" — unless of course there is a "conceptual scheme", if not some "controls", which permit us to determine what is distorted and what to believe; nothing of the sort is available, except the evidence presented in the initial pages of this paper.

Return to Text

61     Finley ibid. 9.

Return to Text

  Return to beginning