Part of a gloss in Harpokration s.v. Hipparkhos has been extensively1 studied:
another one is Hipparkhos son of Kharmos, according to Lykourgos in the Against Leokrates. Regarding this one Androtion says in the second book that he was on the one hand a relative of the tyrant Peisistratos, and he was the first to be exostracised under the law on ostrakismos at that time passed for the first time, because of the suspicion against those around Peisistratos, because he became a tyrant being a demagogue and strategos.2
There is universal agreement, as far as I can see, that the gloss means that Androtion dated the first ostrakismos to immediately after the introduction of the law, that is, to 488/87. There is absolutely nothing in the text of Harpokration to support this; and Harpokration is our only source for Androtions alleged remark on Hipparkhos. There is nothing in Harpokration that indicates a date. It is only by basing everything on hypothetical assumptions which cannot be proved that the remark of Harpokration can be understood as referring to 488/87.
Four passages in our sources refer to the ostrakismos of Hipparkhos: (1) Harpokration s.v. Hipparkhos (see p. 109 and n. 2 above); (2) Souda I 523 (deriving from Harpokration); (3) Plut. Nikias 11.8 (in my opinion deriving from a lexicographical source); (4) Aristot. AP 22.4. Only Aristotle offers an absolute date for the ostrakismos of Hipparkhos that is beyond any doubt, i.e. 488/87.
It is an accepted fact that Aristotles text derives from Androtion, because Aristotle and Harpokration are almost verbatim the same, and because Harpokration refers to Androtion. If this fact is indeed a fact, two questions must be satisfactorily answered: (1) how can we distinguish between material from Androtion and non-Androtionian material; (2) what are (and how
do we account for) the substantive disagreements between Androtion and Aristotle in those points of detail on which they in fact use the same phrases?
The first question is nearly impossible to answer. A structural analysis3 suggests that AP 22. 3-6 has been composed as one piece. Since Androtions Atthis was very lengthy,4 we would have to assume that Aristotle built a structure which he then adorned with details from Androtion, including his remarks on Hipparkhos and ostrakismos.
The question regarding divergence or convergence of Aristotle and Androtion is commonly answered to the effect that Aristotle disagrees with Androtion about the date of the introduction of ostrakismos: at 22.1 he unequivocally ascribes the law to Kleisthenes whose reforms he dates equally unequivocally after the expulsion of Isagoras in his archonship in chapter 21; at 22.4 he dates Hipparkhos ostrakismos to 487/87. Since 22.3f. derives from Androtion it is concluded that Androtion too dated the ostrakismos of Hipparkhos to 488/87; and since Harpokration claims that Androtion connected the introduction of the law with Hipparkhos it follows that Aristotle contradicted Androtion precisely regarding this date. (In support of this hypothesis one might add that Dionysios of Halikarnassos attests Hipparkhos archonship in 496/95; it would seem unlikely that Hipparkhos could have risen to archon and candidate for the Areopagos a mere eleven or twelve years after his exile.)
Are these assumptions weighty enough to exclude the possibility that Androtion and Aristotle disagreed, not concerning the date of the introduction of the law, but regarding the date of the ostrakismos of Hipparkhos? I suggest that it is entirely worth our while to investigate the possibility that Androtion dated the ostrakismos of Hipparkhos to c. 506/05. There is no indication that Androtion knew from documents of Hipparkhos archonship in 496/95; or that he had documentary evidence for the date of Hipparkhos ostrakismos which is the most important point.
Now Jacoby5 thinks that the idea is inconceivable that the learned historian should have despised such a document (i.e. the official archon list): all Atthidographers beginning with Hellanikos made use of the list for it provided the best possible framework for arranging the individual facts of the historical period. There seems no reason, however, to assume that each and every Atthidographer mentioned every archon, or at least took note of every archon. Since nothing noteworthy appears to have occurred in Athens
in 496/956 there was no reason for Androtion deal with Hipparkhos archonship.
Jacoby7 takes it for granted that the Atthidographers inspected official documents. He adds,8 however, that Krateros would hardly have composed his YHFISMA/TWN SUNAGWGH/, had the Atthidographers already collected and written down all YHFI/SMATA. It seems fair to assume that each Atthidographer was guided by his own interests in selecting documents he would exploit and quote. There is nothing, on the other hand, in Harpokrations gloss that Androtion could have gathered from official documents. Neither the family relationship of Hipparkhos with Peisistratos nor the political motives that led to the introduction of the ostrakismos law represent the kind of information that Androtion could have brought to light through archival research. Thoukydides notorious digression on Peisistratid genealogy in book six demonstrates how limited the available evidence really was.9 It seems very doubtful that Androtion could still resort to oral sources of the kind which Thoukydides still seems to have been able to exploit. The growth and expansion of historical knowledge10 regarding both Peisistratid genealogy and especially the political and psychological background of the passage of the law on ostrakismos seems most probably to be due to the creative thinking of historiographers and Atthidographers.
There are accordingly no convincing arguments that would compel us to believe that Androtion inspected documents which caused him to take note of the date of the ostracising of Hipparkhos and incorporate it in his historical account. The disagreement between Aristotle and Androtion which results from assuming that Androtion dated Hipparkhos ostrakismos to c. 506/05 is rather inconsequential in comparison to the disagreement concerning the date and political background of the law by which ostrakismos was introduced.
Aristotle, as we have seen, does not seem to depart from his assumed source Androtion in any significant way. It would border on the absurd to assume that Aristotle copied Androtion by and large, but completely rejected Androtions date for the introduction of the ostrakismos law; and that Aristotle accomplished this remarkable feat by transposing the two words TO/TE
PRW=TON alone, without remarking on his thorough revision of Athenian democratic constitutional and political history. It remains equally inexplicable that such a fundamental discrepancy between two authors, both of whom were extensively exploited by the lexicographers,11 remained completely unnoticed by those lexicographers12 (or why their glosses regarding this discrepancy disappeared from this tradition so completely that the extant late manuscripts or compilations contain no hint at it whatever).
In summing up I should like to put forward for further debate the following conclusions. If one wishes to insist on Androtion as the source for Harpokrations gloss on Hipparkhos son of Kharmos, and that Androtion can also be encountered in Aristotles AP 22.3f., one ought to exercise prudent caution. Aristotles silence on his disagreement with Androtion on dating, which amounts to a significant rewriting of the history of Athenian democracy, remains inexplicable,13 unless we view that date as one of only secondary significance: the year of Hipparkhos ostrakismos. Androtion dated the introduction of the law on ostrakismos no differently from Aristotle, as part of the Reforms of Kleisthenes,14 and he believed that Hipparkhos, the cause of the introduction of the law, was ostracised at the first available opportunity.15
1 We cannot here list the abundant literature on the topic. Cf., at any rate, R. Thomsen, The origin of ostracism (1972); P. J. Rhodes commentary on the AP at 22. I should also like to refer to my study in Klio 73 (1991), 28-45.
2 Text of the edition of I. Bekker:
Jacoby printed the entire section as FGrHist 324 F 6.
4 With the end of Book Two, Androtion had arrived only at 463/62: F. Jacoby, FGrHist 3 b suppl. text, p. 105, 2-6.
5 Jacoby, Atthis (1949), 171.
6 Cf. my contribution ["Athens: between tyranny and democracy", in: K.H. Kinzl (ed.),] Studies Schachermeyr (1977), 212f. (also arguing against monumental internal upheaval in the 490s as has sometimes been proposed.)
7 Jacoby (n. 5), ch. III § 5.
8 op. cit., 208.
9 Thouk. 6.53.3-60.1. We cannot reference here the entire libraries (K. von Fritz, Griech. Geschichtsschr. , 269, n. 51) produced on this excursus. I attempted to emphasise the genealogical element in this excursus in Historia 22 (1973), 504 ff.
10 Cf. Jacoby (n. 5), 152ff.; 188ff.
11 Harpokration explicitly cites Androtion 14 times and AP 51 times. Both are cited in the gloss s.v. *I(/PPARXOS.
12 Cf. Keaney, Historia 19 (1970), 2.
13 To think that no firm opinion had emerged on the date of the introduction of ostrakismos (as Keaney proposes, op. cit. [n. 121, 2) by the fourth century appears unreasonable. The same, however, is entirely thinkable and even probable as regards the date of the ostrakismos of Hipparkhos.
14 Cf. D. Kagan, Hesperia 30 (1961), 395. Kagan unfortunately arrived at this conclusion by way of a reinterpretation of the alleged meaning of Harpokrations TO/TE PRV=TON, and by mixing up the arguments of the historian with those of the philologue on this text; only the latter arguments are valid.
15 Thus, before the rediscovery of AP, G. Busolt, Griechische Geschichte, 11, 1885, 620 with n. 2.
The confused information in Ailian. POIKI/L. I(ST. 13.24 might be the result of sloppy abridgement of Androtion: the introduction of ostrakismos by Kleisthenes and the first ostrakismos, of Hipparkhos, are compressed into the introduction by and first application against Kleisthenes. Similarly silly is the excerpt by Herakleides Lembos (F 371 Rose = p. 14, 23-15, 2 Dilts), were the omission of the name of Kleisthenes elevates Hippias to the creator of ostrakismos.
Read at the meetings of the Classical Association of Canada in Québec in May 1989.