BLANK, DAVID L., Dialectical Method in the Aristotelian "Athenaion Politeia" , Greek, Roman 
and Byzantine Studies, 25:3 (1984) p.275 

Dialectical Method in the Aristotelian 
Athenaion Politeia 

David L. Blank 

B OTH CONTENT AND FORM of the Athenaion Politeia, as scholars 
have long observed, are in agreement with what Aristotle l says 
elsewhere, and are indeed what one might well expect an Aris­

totelian treatment of the Athenian democracy to contain.2 The work 
consists of a history of the mechanisms of state, showing how one 
organ after another became in turn the most important,3 followed by 
a synchronic account of the state machinery in the late fourth cen­
tury.4 The account of Solon as the ~U'oc; '1TOAiT..,.,C; and consequently 
the consummate Athenian politician (Ath.Po/. 5.3, 11.2) squares with 
Aristotle's philosophy in genera15 and with his political theory in 
particular.6 The Ath.Pol. 's insistence on this point, and others as well, 
suits the exemplary and didactic purpose Aristotle assigns to the 
collection and study of constitutions (Eth.Nic. 1181b6-9). 

What is the historical methodology of the Ath.Pol.? I shall argue 
that certain aspects of the gathering, use, or neglect of source ma­
terial in the historical part can be explained with reference to Aris­
totle's theory of scientific and philosophical method. In particular I 
shall address the Ath.Pol. 's 'contamination' and 'rationalizing correc-

I I use this name for the sake of convenience. It is immaterial to my argument 
whether the author of the Ath.Pol. was Aristotle himself or one of his pupils, a sugges­
tion now revived by P. J. RHODES, A Commentary on the Aristotelian Athenaion Poli­
teia (Oxford 1981: hereafter 'Rhodes') 58-63. 

2 See for example K. von Fritz and E. Kapp, "The 'Constitution of Athens' and Aris­
totle's Political Philosophy," in Aristotle's Constitution of Athens and Related Texts (New 
York 1950) 32-66 (= J. Barnes, ed., Articles on Aristotle II [London 1977] 113-34). 
See also, for an extended and controversial treatment, J. Day and M. Chambers, Aris­
totle's History of Athenian Democracy (Berkeley 1962), and responses by K. von Fritz, 
Gnomon 39 (1967) 673-81; D. Kagan, CP 59 (964) 187-90; F. W. Gilliard, "Teleo­
logical Development in the Athenaion Politeia," Historia 20 (1971) 431-35; Rhodes 
10-13; E. Schiitrumpf, Die Analyse der Polis durch Aristoteles (Amsterdam 1980). 

3 Cf Pol. 127SbS-ll, 130Sb25-31. 
4 E.g. F. Jacoby, Atthis (Oxford 1949) 212; cf Pol. 1252a24. See also J. J. Keaney, 

"The Structure of the Athenaion Politeia," HSCP 67 (1963) 115-46, and "Ring Com­
position in Aristotle's Athenaion Politeia," AlP 90 (969) 406-23. 

5 E.g. Eth.Nic. 1096a25, 1105b26ff. 
6 Pol. 1295b4-96a7 (OJ,LOAO'Yt:tTat TO ~TPWV apUrTOV Kai TO ~U'OV). 

275 



BLANK, DAVID L., Dialectical Method in the Aristotelian "Athenaion Politeia" , Greek, Roman 
and Byzantine Studies, 25:3 (1984) p.275 

276 DIALECTICAL METHOD IN THE ATH.POL. 

tion' of sources. Both these techniques can be explained by taking 
into account what Aristotle is likely to have thought about historical 
'facts' and sources, and about history and politics as sciences. One 
cannot find absolute certainty in historical investigation, nor com­
plete agreement among historical sources. Aristotle's epistemology 
will therefore not recognize history as a science stricto sensu. It is not 
the analytical, but the dialectical method that informs the Ath.Pol. 
We may consider first the problematic aspects of its method, then the 
dialectical method itself, and finally an instance of its application. 

Within a given section of his historical narrative, Aristotle often 
follows one main source, but will add at will items drawn from an­
other source. He does this even when the second source has a ten­
dency contradictory to the main source. He also makes up his own 
narrative of some parts of Athenian history, drawing individual data 
from various sources at once. The most extensive treatment of these 
procedures is that of G. Mathieu, who spoke of a tendency to 'mix' 
or also to 'reconcile' various historical sources with one another.7 

Thus, for example, in chapters 14-15 the narration of the rise of 
Peisistratus derives mainly from Herodotus; Aristotle juxtaposes this 
narrative with a (disagreeing) chronology drawn from an Atthis, and 
adds some other matter as well. Thucydides is a main source of the 
narrative in chapters 29-33 on the Four Hundred and the Five Thou­
sand, but an apologetic source, which will have cited documents, is 
also used. On the other hand, the author himself put together the 
account of the Thirty and the Ten from several sources.8 

In a brief comparison of Aristotle's methods in the A th. Pol. and in 
his biological writings, M. Pokrowsky drew attention to evidence that 
even the supposedly strict, empirical naturalist commingled his own 
observations with the reports of others. On occasion Aristotle even 
corrected such second-hand reports not on the basis of any observed 
evidence to the contrary but merely in order to make the reports 
more plausible. Pokrowsky called such changes "rein rationalistische 
Verbesserungen."9 For example, Aristotle (Hist.An. 502a13) repeats 
Herodotus' description of the hippopotamus (2.71, OVpTJ V i7T1TOV Kat 
c/xJJvr,v' ~ya8o~ DUOV TE f30~ <> ~YUTTO~) but makes one change: it 
is not as large as an ox, but only the size of a donkey {,.dYE8o~ 8' 

7 G. Mathieu, Aristote, Constitution d'Athimes: Essai sur la methode suivie par Aristote 
dans la discussion des textes (Brussels 1915) 11, 26f, 5 Of, 72, 115, 124; cf Rhodes 27ff, 
50,55. 

8 See Rhodes 191-99; 29, 365-68; and 420ff. 
9 M. Pokrowsky, "Ueber das Verhaltnis der 'Afhr. zu den naturwissenschaftlichen 

Schriften des Aristoteles," NJbb 151 (1895) 465-76, at 466f. 



BLANK, DAVID L., Dialectical Method in the Aristotelian "Athenaion Politeia" , Greek, Roman 
and Byzantine Studies, 25:3 (1984) p.275 

DAVID L. BLANK 277 

E(TTtV -qALKOV ovoS'). Herodotus' assertion that Ethiopians have black 
semen (3.101.2) is countered by Aristotle on the ground that Ethi­
opians do not have black teeth.l° Again, Aristotle repeats Herodotus' 
note that the crocodile is blind in the water (2.68), but he expresses 
himself less categorically (Hist.An. 503all, /3AE7T'OV(Tt 8' EV /-LEV T~ 
v<>an qxxvAW,). 

Similarly, in the A th. Pol. Aristotle several times corrects or contra­
dicts his sources on grounds of likelihood. He asserts that the second 
of the three original property qualifications, the cavalry, was "more 
likely" to have been determined by a measure of produce en gros, as 
was the case with the pentakosiomedimnoi, than by one's ability to 
keep a horse. The latter was the contention of some sources which 
adduced in evidence Anthemion's dedication of a statue of a horse 
upon becoming a knight.11 Those who say that Solon's laws were pur­
posely vague, so that the courts would have more power, are contra­
dicted by Aristotle on grounds of likelihood (9.2, OV /-LT,v eiKoS'). 
Likewise, Aristotle places the disarming of the population at the be­
ginning of Peisistratus' third tyranny (15.4-5), in accordance with his 
general rules about the establishment of tyranny.I2 Then he is forced 
to contradict Thucydides' account of the disarming of the men in the 
Panathenaic procession by Hippias (Thuc. 6.56.2-3, Ath.Pol. 18.4). 

How could Aristotle make so free with his sources as to mix them 
and play them off one against the other, accepting or rejecting data as 
he saw fit, without pausing to consider what doubt was thereby cast 
both on his sources and on his own account? The answer lies in Aris­
totle's conception of the nature of his enterprise and its limitations. 

In the Poetics Aristotle contends that history, because it tends to 
deal with the particular rather than the universal, is less philosophical 
than poetry:13 

10 Gen.An. 736alO: 'HpOSOTO~ yap OVK aA7jOT/ A~YEL qx{UKWlJ ~AaLlJalJ fillJaL T-rylJ TWlJ 
AiBWTTWlJ YOlJT)lJ, WUTTfiP avaYKawv 01' TWV ~v xpoav I-LEAaVWV filVaL TTaVTa ~)..ava, 
Kat TaVO' opWV Kat TOil<; oBOVTa~ aVTWV OVTa~ AfiVKOV~. See also the immediately 
preceding a priori polemic against Ctesias of Cnidus (736a2ft). 

11 Ath.Pol. 7.4: ov J-LTjv aAA' EVAoywnpov TO'i~ ~TpOL~ (jLTlPT/u8aL Ka8clTTEp TOV~ 
TTfi VTaKOULOI-LESi,.LVOV~. 

12 Pol. 1311a12-14. On the topographical and chronological improbability of Aris­
totle's account see U. von Wilamowitz-Moellendorff, Aristoteles und Athen I (Berlin 
1893) 289-92; Mathieu (supra n.7) 40f. The theory that this disarmament is an Aristo­
telian inference is strengthened by the fact that Aristotle has Peisistratus say almost the 
same thing after the disarmament (Ath.Pol. 15.5, ov xp-ry BavJ-Lcl~ELv OVSE aOvl-LEI.v, a)..)..' 
aTTfiA8ovTa~ ETT!, TW" iSiwv fillJaL, TWV B£ KOLlJWV aVTo~ ETTLI-LEAT)ufiuOaL TTclVTWV) as 
Herodotus does after the battle at Pallene 0.63.2, Oapu~EL" TE KEAEVOVTE~ Kat aTTL~VaL 
EKaUTOV ETT/. TIl: EWVTOV). See also Rhodes 213. 

13 l415b4-11. On this celebrated and puzzling passage see F. W. Walbank, "History 
and Tragedy," Historia 9 (960) 216-34; G. de Ste. Croix, "Aristotle on History and 



BLANK, DAVID L., Dialectical Method in the Aristotelian "Athenaion Politeia" , Greek, Roman 
and Byzantine Studies, 25:3 (1984) p.275 

278 DIALECTICAL METHOD IN THE ATH.POL. 

, , , ~ A.J. ,..." , , , I \~, ... 

aUa TOV~ oW.~PE'" T~ TOV /-LEV Ta 'YEV0/-LEva AE'YEtV, TOV oE ow. 
'" '" ' '"t..,}. ..t...!_ ' "'_ I I av 'YEVOtTO. oW Kat 'fIW~0(T0'fAU' EpOV Kat U7TOVouWTEPOV 7TOt'T/Ut~ 

i(T'Topia~ EUTill' T, ,."Ell 'Yap 7Toi'T/Ut~ JUiUOll Ta Ka8oAOV, T, 8' iu-
I , 8'" \ ' " ~\ 8' ,,..., \ TOpW. Ta Ka EKaUTOll n.E'YEL. EUTtll oE Ka OAOV /-LElI, Tcp 7TO~ Ta 

7TOW. aTTa uvp,{3aillEt AE'YEtV Tj 7TpaTTEtv KaTa TO ElKO~ Tj TO ava'Y-
~ 'P 'y • , " '8 I '",' 8' Kawv, ov UToxa."ETat 'T/ 7TOt'T/Ut~ ovop,aTa E7TtTt E/-LElI'T/' TO oE Ka 

" , 'A ~ Q,..!'" " I: "," 8 EKaUTOll, TL ~KttJCAA.o'T/~ E7TpaO:,ElI 'T/ Tt E7Ta Ell. 

This characterization of history means that, for Aristotle, history 
cannot be an investigation in which demonstration «(bT08E~t~) and 
scientific understanding (E1rUTTT,p,..,,) figure (Metaph. 1 039b27 -40a 7), 
for E1rUTTT,p,.." is only of TO Ka(JoAov. Moreover, in Aristotle's view 
there is no demonstration of what is not eternal and unchanging, nor 
is there any knowledge of it in an unqualified sense (A n. Post. 75b21-
26): 

"t..~ '",! ,. ,~ • , 8/\ 'I: 'P • \ \ 
~VEpOV at: Kat Eav WCTtV at 7TpOTaUEt~ Ka On.OV EO:, CUV 0 UVI\.I\.0'YW-

p,O~, OTt ava'YK'TI Kat TO uvp,7TEpaup,a ai.'"Bwv Eivat rij~ TOW.V'T'T/~ 
~ §:::: l; \,.., to:~ ::..,., 't ,.. 't ~ _: I:. ,,, '" " 
a7TOoE"<:.E~ Kat 'T'T/~ a7T1U.U"j Et7TEtV a7TOot:~E~. OVK EUTtV apa a7TO-

&"g,~ TWV 4>8apTwv ov8' E7TW"r'i,p,'T/ ci7T~, aU' oiYr~ W<T7TEP KaTa 

UVp,fjEf3'T/KO~, OTt ov Ka8' OAOV awov EU'TLV auCx. 7TOTE Kat 7T~. 

If it is not a fit subject for strict scientific investigation, history will 
have to make do with premises that are less than certain and reach 
conclusions that are also only likely, not certain. 

Besides dealing with particular occurrences, political history has 
other characteristics which make it an unfit subject for apodictic 
reasoning. It is about the doings and development of states, which 
are composed of parts which, in turn, are constituted by people. 
Wherever the actions of people are concerned, choice and delibera­
tion figure among the explanatory factors.14 This is not true of an art 
or techne, which does not deliberate about its goals or methods (Ph. 
199b28ft). Politics and the history of-or rather the explanationl5 of 
the development of-politics in any state fall under the science of 
ethics.16 

Aristotle has a particular methodology for dealing with subjects like 
ethics, where scientific accuracy and invariability are not to be ex-

Poetry," in The Ancient Historian and His Materials, Essays in Honour of c. E. Stevens 
(Farnborough 1975) 45-58; R. Weil, "Philosophie et histoire: La vision de l'histoire 
chez Aristote," in La "Politique" d'Aristote (Entretiens Hardt 11 [1965]) 162-63. 

14 Int. 19a7-8; Eth.Nic. 1112a18ff, 1104a5-10. 
15 Cj: e.g. An.Post. 71b9-12: E7rUTTau(Jat 8E oWJ.U(J' EKauTov a7rA~ '" (hal' TI,V 1" 

aiTtav olWJ.U(Ja 'YtvWuKEtV l)t' ill' TI) 1Tpa'Y~ EUTtV, OTt EKEivov aiTta EU'Ti, Kat ILT, 
EV&'XEU(Jat TOW' uUW'; EXEtV; and Ph. 184a12ff. 

16 Eth.Nic. 1180a32ff, 1181b14ff. 



BLANK, DAVID L., Dialectical Method in the Aristotelian "Athenaion Politeia" , Greek, Roman 
and Byzantine Studies, 25:3 (1984) p.275 

DAVID L. BLANK 279 

pectedI7 in the premises or the conclusions of arguments. The meth­
od is called dialectic, and Aristotle describes it in detail in the Topics. 
This work, perhaps an instruction manual for the Academy's course 
on debating, gives a set of rules for solving 'dialectical problems', 
problems that result when the many and the wise disagree with one 
another or among themselves (Top. 1 OOb21-23) . 

One such dialectical problem was that of aKpauUx or weakness of 
the will. In his discussion of the conflicting opinions about akrasia 
(Eth.Nic. 1145f), Aristotle gives his clearest example of the dialectical 
method.18 He begins by explaining his method (b2-7) : 

~ -1""1 ~'tf , , "'" "\. \. (J , 'A...IV' ,1""\ 
Ut:t u , WU1TEp E1Tt TWV aI\J\.WV, Tt EVTafi Ta 'P"tV0/.LEva Kat 1TPWTOV 
~, ,,~" \ " '" ~- i: uta1TOPTJuaVTa!) OVTW uEtKVVVaL J..UX,,-LUTa /.LEV 1TaVTa Ta EVuu~a 

, 1""1 , '6 ' ~, , , \."" \ , , , \ 
1TEpL TaVTa Ta 1Ta TJ, EL uE f.LTJ, Ta 1T,,-ELUTa Kat KVptWTaTa' Eav 'Yap 

A-ln-JTai TE T£X 8VUXEP~ Kat KaTaAEi1f'TJTat Ta Evoo~a, BEBEL'YJ,LEvOV 
" "It ,... av ELTJ LKavW!). 

Next he lists seven beliefs about weakness held by many people 
(EvBo~a), concluding with the words Ttl /-LEV o-ov AE"Y0I-'EVa TavT' 
fUTlv. Then various problems arising from these beliefs in combina­
tion with one another or with other beliefs are set out (a1TopUxL). 
Finally, he gives a solution (AV(J"(,!) of these problems, i.e. a theory of 
weakness which 'saves' as many of the common beliefs as possible. It 
is important to save as many of the EvBo~a as possible, because the 
opinions of the wise and the many are not likely to be wholly without 
truth.19 Indeed, whatever all agree upon is true.20 

It is not only in studies such as ethics that dialectic is useful. Aris­
totle says that dialectic is the method to be used in clarifying the 
'starting points' (apxai) of every science, since no science can exam­
ine its own apxal (Top. 101a34-b4). These starting points include 
axioms and also general concepts, such as place (Ph. 208a27fi) and 
the eternity of the universe (Cael. 270b4ff). They will be examined 
on the basis of the opinions of the wise and of the many (EvBo~a). 
These include, for example, philosophical theories, concepts implied 
in our ordinary language, and the products of observation, both in 
our daily lives and in the pursuit of science (cpaLVOI-'EVa KaT' atcr(J"fJ­
crLV, e.g. Cael. 306a16-17); all these data can be referred to as cpaLVO­
/-LEva or EvBo~a, and they will be different, as the subjects to be in-

17 Eth.Nic. 1094bllff: because of the 7TAaVTJ and 8taqx,pa in political matters, one 
ought not to look for 'TaKpL/3e<;, but rather describe matters nfmp; cf 1104al-5. 

18 Cf J. Burnet, The Ethics of Aristotle (London 1900) xxxi-xliii. 
19 Cf Metaph. 993a30-b5, 'Ti<; av (Jupa<; a,.wp'TOL; 
20 Eth. Nic. 1173a 1: & yap mluL OOKEt, 'TUW' EZvai cJxx~ V' 0 8' a vmpwv 'TUV'TTJ v 'T-ry v 

7TiO''TLV ov 7Tavv 7TLO'T()'TEpa epEL. Cf 1143bll-14. 



BLANK, DAVID L., Dialectical Method in the Aristotelian "Athenaion Politeia" , Greek, Roman 
and Byzantine Studies, 25:3 (1984) p.275 

280 DIALECTICAL METHOD IN THE ATH.POL. 

vestigated will differ.21 We shall see that in history, too, various kinds 
of Evoo~a present themselves to the researcher for consideration. 

Insofar as any inquiry is about the starting points of science or 
begins from observations which are not invariable and which may 
conflict, that inquiry will be dialectical. Characteristics of the observa­
tions and authorities themselves will determine, when there is a 
conflict, whether any particular 'phenomenon' is 'saved' or not. Does 
the observation cohere with what we already know? Is the authority 
trustworthy? These are the questions that one must ask. Thus, in 
many cases the beliefs of the wise will clearly take precedence over 
the beliefs of others,22 and of these the beliefs of the wisest will 
count for most.23 Some of our beliefs are so secure that any datum or 
theory which contradicts them must be dismissed.24 

Political history is a dialectical inquiry in this sense. Its sources, 
especially for the earlier periods, were unreliable-oi p..V80AO'YOVVTE~, 
Aristotle calls them at one point25 -and often contradictory. Unreli­
able stories could at times be supported or contradicted by UTJp,EUx.,26 
such as survivals of ancient practices in contemporary language or 
custom. Thus, the fact that the eponymous archon was the last of the 
three major offices to be instituted is shown by the circumstance that 

21 Fundamental in this regard are: W. Wieland, Die aristotelische Physik2 (Gottingen 
1970) 65-95, and G. E. L. Owen, "Tithenai ta phainomena," in Aristote et les problemes 
de methode, ed. S. Mansion (Lou vain 1961) 83-103 (= Aristotle: a Collection oj Critical 
Essays, ed. J. Moravcsik [Garden City 1957] 167-90 [cited here]). A basic summary is 
found in J. L. Ackrill, Aristotle the Philosopher (Oxford 1981) 12ff, 107-15. M. C. Nuss­
baum has also made a valuable restatement of Owen's work and has carried it further: 
"Saving Aristotle's Phenomena," in Language and Logos, Studies in Ancient Philosophy 
Presented to G. E. L. Owen, edd. M. C. Nussbaum and M. Schofield (Cambridge 1982) 
267-93. Nussbaum in my view overstates the dichotomy seen by Owen between scien­
tifically observed phenomena and received opinions (note Owen's own linkage of the 
two at 174f), while underestimating the degree to which Aristotle actually differentiates 
these two sources of the pre-existing knowledge from which all understanding must 
come. 

22 E.g. Eth.Eud. 1214b28-15a4: 1Taua~ ,.uv o~v Ta~ &)~a~ E1TIUK01T6V ... 1TEPUP'YOV­
... Ta~ 8E TWV uocJ>Wv E1TIUKE1TTEOV J.d)Va~· ClT01TOV 'Yap 1TPOUcpEPELV AO'YOV TOL~ AO'YOV 
p.:r/8Ev 8EO~VOt<;, aAAO: 1TCx9o~. 

23 Top. l00b22f: TOL~ Uocf>oL~, Kat TOVTOL~ Tj 1Tamv Tj TOL~ 1TAEiuTOL~ Tj TOL~ ~AwTa 
'YVwpiJ.LOL~ Kat EV&)~L~. 

24 The most fundamental of all beliefs is the principle of non-contradiction (Metaph. 
l005bI9-34). One interpretation of Parmenides' theory of the One is dismissed (Ph. 
185b19fi) on the grounds that it conflicts with this most basic axiom. Cj. Nussbaum 
(supra n.20 283-89. 

25 Pol. 1312a3 on Sardanapalus: El aATj6ij TaVTa oi /.Lv6oAo'YoIIV'TE~ AE'YOVULV, El BE /.L~ 
E1T' EKEivov, aM' E1T' aAAov 'YE uv 'YEVOLTO TOWO aATj8E~. Compare Hecataeus FGrHist 
1 F 1: oi 'Yap 'EAA 7) VWV AO'YOL 1TOAAOi TE Kat 'YEAoun. 

26 Cj. Mathieu (supra n.7) 26; Rhodes (59) points out that this form of argument 
was not only Aristotelian. 



BLANK, DAVID L., Dialectical Method in the Aristotelian "Athenaion Politeia" , Greek, Roman 
and Byzantine Studies, 25:3 (1984) p.275 

DA VID L. BLANK 281 

he, unlike the basileus and pole march, does not administer any of 
the 1T'a'Tpta (Ath.Pol. 3.3, UT//LE'iov). The present-day venue of the 
marriage of the basileus' wife and Dionysus is used as a UT//LE'iov of 
the location of the basileus' original seat (3.5). That Peisistratus' 
party, the diakrioi, included those of impure descent is proved to 
Aristotle by the revision of the citizen's roll after the fall of the 
tyranny (13.5, UT/J-Li'wv). 

The less reliable the Evoo~a from which the investigation begins, 
the less basic are the philosophical beliefs or theses which will be 
allowed to contradict or alter them. The opinion of an illustrious 
predecessor will only be altered if it conficts with something of which 
the researcher is quite certain. The opinion of someone not well 
respected may be altered on less than compelling grounds. These 
alterations of Ev8o~a will include the "rein rationalistische Verbesser­
ungen" of Pokrowsky. 

If, therefore, the accounts which the historian is trying to 'save' 
produce a:TrOpia" because they conflict with one another or with 
beliefs to which the historian is more attached, then the accounts 
must be altered in order to be saved. Sometimes a story could be 
shown to be inconsistent with uncontested fact, as the report that 
Solon and Peisistratus had been lovers was inconsistent with the 
dates of those two politicians (Ath.Pol. 17.2).27 Sometimes a piece of 
evidence may be used silently to correct a datum. Thus, it is possible 
that the amount of Peisistratus' tax on produce is changed from five 
percent (Thuc. 6.54.5, ElKOU'TT,V ~vov) to ten percent (Ath.Po!. 16.4) 
on the basis of the story about the 'tax-free farm', in which a tithe 
was mentioned 06.6, 'TT,V 8EKa'TT/v).28 In other cases a generalization 
serves as evidence that a specific claim should be rejected. So the 
charge that Solon allowed some of his friends to enrich themselves, 
and may even have enriched himself, using advance information 
about the seisachtheia (6.2) must yield to the 'democratic' account, 
which is 1T'tOaVW'TEpOC;, because a man as moderate in all things as 
Solon is not likely to have sullied himself in such small and unworthy 
matters.29 Here Aristotle's strongly-held general characterization of 
Solon resolves the Ct.1T'Opia presented by two opposing accounts. 

27 Rhodes 26 gives a complete list of places where Aristotle says why he prefers one 
account to another. 

28 Day/Chambers (supra n.2) 95. Rhodes 215 accepts the suggestion of K. J. Dover 
in Gomme, et al., Historical Commentary on Thucydides IV (Oxford 1970) 329, that 
{jEKa~ be interpreted as a generic term, which could include the more specific Ei-

KOUTTJ· , , , \, , ......" H' I \ , 

29 Ath.Pol. 6.3: ov yap EtKO~ Ell /MV TOL~ aAA.OL~ OVTW /MTPWV YEvE(J(}aL KaL KOLVOII ... 
ell OVTW {jE f..UKpOis Kat (XvagwL'; KamppV7raillEtIl favTolI. 



BLANK, DAVID L., Dialectical Method in the Aristotelian "Athenaion Politeia" , Greek, Roman 
and Byzantine Studies, 25:3 (1984) p.275 

282 DIALECTICAL METHOD IN THE ATHPOL. 

In the light of these examples, it is possible to argue that all the 
changes made by Aristotle in Thucydides' account of the assassina­
tion of Hipparchus were introduced in order to bring the incident into 
line with a single Aristotelian thesis about stasis. Aristotle believes, 
namely, that revolutions happen from trifles, but about great issues.3o 
Political imbalances form the background for the political conspiracies 
which seize upon erotic happenstances and other incidents, using 
them as occasions for revolution. 

Aristotle states that Hippias and Hipparchus controlled Athens after 
the death of Peisistratus,31 makes Hipparchus out to be a poorly­
behaved fellow, lacking in seriousness, and then ascribes the insult 
against Harmodius' sister to the outrageous brother Thettalus (Ath. 
Pol. 18.1-2).32 Peisistratus had been a very moderate tyrant, and 
Hippias was well qualified to continue his father's long and peaceful 
rule (</>WE(. 1TOA.t,T(.KO~ Kat. lJ.L<Ppwv), but the presence of the other two 
brothers, with their dubious characters, is bound to lead to trouble 
for the tyranny. For Aristotle is insistent on the fact that a long-lived 
tyranny is very much a result of the tyrant's moderate character and 
avoidance of scandal.33 Aristotle's assignment of the actual insult to 
Thettalus, when Hipparchus was the victim of the ensuing conspiracy, 
shows that the conspiracy was, in his opinion, not designed primarily 
as revenge for the insult, but was rather aimed at all three tyrants 
and the end of their reign. Harmodius and Aristogeiton are made 
into the leaders of a political plot, in accordance with Aristotle's 
general theory about revolts. 

There are four other differences over matters of fact between 
Thucydides and Aristotle in regard to the murder of Hipparchus. 
These also make most sense in the context of a politically motivated 
conspiracy.34 First is the statement that Harmodius' sister was in-

30 Pol. 1303bI7-18: 'Y{:yvovTaL ~v ovv ai. UT(IUEL~ OV 1TEpt I-UKPWV aU' EK f.J,U<pWv, 
UTaUW'OVUt 8E 1TEpt lLE'YaAwv. 

31 Cf, Jacoby (supra n.4) 332 n.7, "a compromise in regard to a most essential point 
of divergence between the contrasting versions of the tradition." 

32 The literature on this passage is too large to note here; see the treatment in 
Rhodes ad loc. for a summary. I am unconvinced by attempts to delete parts of this 
passage (so 1. M. Stahl, "Thessalos der Sohn des Peisistratos," RhM 50 [1895] 383-
93, esp. 386-89, and C. Fornara, "The Tradition about the Murder of Hipparchus," 
Historia 17 [1968] 400-24, esp. 414-18), or to interpret away the differences from 
Thucydides (so U. Wilcken, "Thettalos," Hermes 27 [1897] 478-82). 

33 Cf, Pol. 1314b24: ETL 8E ILT, J,LOvov awov q,aivEu8aL wr,8Eva TWV apxo~vwv vf3pi­
'OVTa ... &Ua w,.,8' aUov 1LT/8EVa TWV 1TEpi awwv; see also 1315aI5-24. Immoderate 
conduct makes the tyrant's subjects resentful and disposes them to use any small 
incident as an excuse to overthrow the tyranny: 1311 b9-11. 

34 There were other accounts of Hipparchus' assassination in circulation, e.g. the 
pseudo-Platonic Hipparchus and (probably) the Atthis of Androtion, but there is no 



BLANK, DAVID L., Dialectical Method in the Aristotelian "Athenaion Politeia" , Greek, Roman 
and Byzantine Studies, 25:3 (1984) p.275 

DAVID L. BLANK 283 

suIted at the Panathenaia (18.2), when Thucydides says that this 
occurred "in some procession" (6.55.1). Since the murder of Hippar­
chus took place at the Greater Panathenaia (Ath.Pol. 18.3 TOtS llav­
a8T1vaio,s, Thuc. 6.56.2 llava87}vac.a Ttl ~E'YaAa), Aristotle's version 
makes the time elapsed between the insult and the murder four years 
(if both were at a different Greater Panathenaia) or one year (if the 
insult was at a Lesser Panathenaia, the murder at the Greater) or a 
few days (if both events occurred during the same festival). That the 
interval was relatively long is suggested by the second disagreement 
with Thucydides: Aristotle says that there were many co-conspirators 
with Harmodius and Aristogeiton (18.2), while Thucydides says there 
were not many, for security's sake (6.56.3). The third conflict pits 
Aristotle's statement that the Panathenaic parade was not, in those 
days, conducted in arms (18.4) against Thucydides (6.56.2-3, 58.2). 
Fourth, Thucydides states that Aristogeiton at first escaped but was 
later caught and roughly treated, while the other conspirators were 
discovered by Hippias' ploy of disarming the citizens and then search­
ing for those who retained their daggers (6.58.1-2). Aristotle dis­
agrees explicitly, saying that the AE'YO~EVO~ AO'YO~35 is untrue, but 
that Hippias was at first unable to discover the co-conspirators and 
was led to torture Aristogeiton for their names (18.4-6). 

All these positions taken by Aristotle work well in the context of a 
political conspiracy. In that context it is reasonable to believe that the 
insult took place one or even four years before the murder, providing 
the impetus for Harmodius and Aristogeiton to found their con­
spiracy and allowing time for them to mobilize those dissatisfied with 
the tyranny and gather their many helpers. About Thucydides' ac­
count, certain questions arise, which may have been seen by Aris­
totle. If the sole reason for the plot were the insult to Harmodius' 
sister, why would anyone not personally connected with Harmodius 
join in the plot? Yet Thucydides' statement allows not only that there 
were other conspirators, but that there would have been more, had 
security not dictated otherwise (6.56.3, ~uav 8' Oll 1TolloL oi (;vvoJ..U,U­
~K(hE~ aucJ>aAEiac; EVEKa). Furthermore, if life was so pleasant un­
der the tyrants, why should the men in the procession, if they were 
armed, have supported the conspirators against the tyrants' body-

reason to believe that Aristotle was following another source in his disagreements with 
Thucydides. Even if some or all of the A th. Pol. 's corrections of Thucydides' account 
did stem from another source, rather than from conjecture, Aristotle's reasons for 
preferring that source to Thucydides may have been the ones argued here. 

35 C/ Eth.Nic. 1145b20, cited supra 279: Tel ~eyo~Jla. 



BLANK, DAVID L., Dialectical Method in the Aristotelian "Athenaion Politeia" , Greek, Roman 
and Byzantine Studies, 25:3 (1984) p.275 

284 DIALECTICAL METHOD IN THE ATH.POL. 

guards (6.56.2)? Thucydides' account of the murder, as is clear from 
the overall similarity and coincidences in vocabulary, was the basis of 
the Ath.Pol. narration. It seems likely that Aristotle saw the problems 
in Thucydides' version and decided that the murder of Hipparchus 
was not just an aAo'YW"'TO~ 'ToAp,a EK 'TOV 1Tapaxpr,J.W 1TEp(.8EO~, but 
had been part of a comprehensive, politically-inspired conspiracy. This 
theory also was consistent with Aristotle's view that erotic happen­
stances give rise to, but are not really the issue in, revolutions. 

As a result of these two types of a1TOpw., internal inconsistency and 
disagreement with a basic Aristotelian theory, Aristotle 'saved the 
phenomena' (i.e. the bulk of the report he inherited from Thucydi­
des) by altering certain details of Thucydides' report, and thereby 
altering its tenor. According to the basic theory of dialectic, the meth­
od here employed, all one has to go on in history are one's own 
general principles and reports which may well be faulty. That the re­
ports and the principles, when combined, may contradict one another 
and have to be altered or partially dropped cannot be allowed to cast 
the resulting synthesis in an unfavorable light. The opinions of the 
wise and the many are unlikely to miss the truth completely, and we 
often have nothing better to offer.36 

THE UNIVERSITY OF CALIFORNIA, Los ANGELES 

August, 1984 

36 My early work on this topic benefited from the advice of Professors R. S. Brum­
baugh, H. von Staden, and D. Kagan; it is a pleasure to thank them for their help.