REVIEW ARTICLE: THE NATURE OF

ATHENIAN DEMOCRACY

JOSIAH OBER

Reprinted for priqte circulation from

CLASSICAL PHIWWGYVol. 84, No.4, October 1989

C 1989 by The UnivenilY of Chic:qo. All righls raervcd.

PRINTED IN U.S.A.

REVIEW ARTICLE

THE NATURE OF ATHENIAN DEMOCRACY*

The constitutional history of Athens is an increasingly popular topic amonghistorians, at least in part because of the ground-breaking work of MogensHerman Hansen. Since the mid-1970s H. has published an impressive series ofmonographs and articles that have clarified many aspects of Athenian legal andgovernmental practice, especially of the fourth century B.C. Furthermore, since itis possible to determine more about Athenian government in the fourth centurythan in any previous period, H.'s studies have contributed to reassessments offifth-century democracy and its sixth-century antecedents. I His previous workhas already earned H. a place in the history of ancient constitutional studies; thepublication of this book, written for both the specialist and the general reader,will bring him a wider audience and provides the opportunity to view his workin a broader perspective.

H. 's oeuvre is voluminous, and his studies are invariably closely reasoned,copiously documented, and provocatively argued. His titles, which sometimestake an interrogative form ("How often ... ?" "How many ... ?" "Whendid ... ?"), leave the reader in no doubt about the subject at hand, and hisconclusions leave no doubt as to the author's stand. In a typical study H. beginsby narrowly framing the question to be asked. He defines the terms of thequestion as accurately as possible, lists possible answers, and then presents allthe evidence that seems to bear on the issue. In his conclusions he sums up theevidence, discards the possible answers that were not supported, and states thesolution. The method is thus empiricai and exclusive; it is the approach that,taken to a logical extreme, was favored by Sherlock Holmes, who claimed that"when you have eliminated the impossible, whatever remains, however improb­able. must be the truth."z Like a Sherlockian solution to a criminal mystery, thetypical study by H. appears to be an impartial collection of all the evidence thatis accessible to anyone willing and able to make careful observations. Once it has

"The Athenian Assemhlr in the Age "rDemos/hene.>. By MOGENS HERMAN HANSEN. Oxford andNew York: Basil Blackwell. 1987. Pp. viii + 249.

I. On pp. 125 30 H. lists 35 theses lhat summarize his revisions and clarifications of the traditionalpiclure of the Athenian constitution. I tend to agree with 16 of these (nos. 1,3.5.10. II, 14-21. 26, 34.35). The 5 theses with which I tend 10 disagree in whole or in part (nos. 6, 23. 24, 29, 32) will be themain subject of this article. On the remaining 14 I suspend judgment. either because 1 regard them aspossible but unproven or because 1 have not thoroughly worked through the evidence. For thesignificance of H.'s work for reinterpreting earlier Athenian history, see, e.g.. M. Ostwald. FromPopular Sovereignty 10 the Sovereignt.r or La ... (Berkeley and Los Angeles, 1986): R. Sealey, TheAthenian Repllhlic (University Park. Pa.. 1987).

2. Arthur Conan Doyle. The Anno/ated Sherlock Holmes. vol. I (New York. 1967). p. 638 (= "TheSign of the Four," chap. 6).

Permission to reprint this review article may be obtained only from the author.

322Rerrinted from CLASSICAL PHILOLOGYVo .84. No.4, October 1989© 1989 by The University of Chicago. All rights reserved.Printed in U.S.A.

324 REVIEW ARTICLEREVIEW ARTICLE 325

H. 's most evident a priori position takes the form of a value judgment: herespects the Athenians and believes that Athenian democracy is an importantsubject with contemporary relevance (pp. 5-6). He defends the fourth-centurydemocracy (pp. 123-24) against those who have claimed that devotion to thepublic good declined in the decades after the Peloponnesian War.

4H. denies that

the democratic government was parasitically dependent upon a slave economy,and instead points to the (usually overlooked) economic contribution of Athenianwomen (p. 34 with n. 232, p. 123). He demonstrates that the state was notfinancially crippled by the provision of pay for widespread political participa­tion, noting (p. 48) that the military budget was far larger than the civilian.H. refuses to view Athenian government as an antiquarian relic and cautiouslyemploys modern parallels (especially the Swiss cantonal assemblies) to elucidateAthenian practice (p. 2).

Equally fundamental, but in my view more problematic, are H.'s convergingassumptions that Athenian democracy is best understood as a "constitution,"and that we can best understand this constitution, and the principles' on which itwas based, by analyzing the relations between formal institutions of government.Students of Athenian history must come to grips with the constraints that these"constitutionalist" assumptions place upon the analysis of political life. s

A good example of H.'s method is provided by his investigation of whetherthe EKld.llcrla or the oIKacr't:"pla (or both) "embodied" the 0i'i lJ.o<; , and hisunderstanding of how this issue bears on the locus of sovereignty in the Athenianstate. H. begins by considering the definitions and interrelationships of the terms0i'ilJ.0<;, EKKAllcrla, and oIKacr't:"plOv. He collects a large number of texts (mostlypassages from fourth-century orations) in which citizens gathered in Assemblyare addressed by speakers as 6 0i'ilJ.0<;; conversely, jurors were seldom (or never)addressed by litigants as 6 0i'ilJ.0<;. H. concludes that the EKKAllcrla did in factembody the 0i'iJ.l0<; of Athens, whereas the oIKacr't:"pla did not. This conclusionis then combined with the fact that the decision of jurors in a oIKacr't:"plOv couldoverturn (through conviction in a ypaq111 ltapav0lJ.(Ilv) a decision made in theEKKAllcr{a: ergo, the Athenian 0i'ilJ.0<; (embodied by the citizens gathered inAssembly, but not by jurors) was not sovereign, and the lawcourts were

• 6sovereign.The validity of H.'s interpretation is contingent, of course, on the meaning­

fulness of the terms in which the original question was posed. Do the concepts"embodiment" and "non-embodiment" (or "representation," or "manifestation,"two other constitutional concepts H. discusses in this context) best suit therelationship between 0i'i lJ.O<; , EKKAllcr{a, and oIKacr't:T]pla? Is political power inAthens best conceptualized as legal sovereignty? Perhaps not; but H.'s argumentis likely to convince those who are used to thinking of good scholarship as

4. See, e.g., R. Garner, Law and SoC'iny in Classical Athens (New York, 1987), p. 132; L. B. Carter,The Quiet Athenian (Oxford, 1986).

5. The constitutionalist model has been questioned by, e.g., W. R. Connor. "The Athenian Council:Method and Focus in Some Recent Literature," CJ 70 (1974): 32-40; M. I. Finley, Ancient History:Evidence and Models (London, 1985), pp. 99-103.

6. Pp. 101-3; evidence collected by H. in "Demos, Ecclesia, and Dicaslerion in Classical Athens,"GRBS 19 (1978): 127-46; H. also emphasizes the distinction between 'V'iqltOlia and VOIIO~ in thiscontext.

objective and without ideological bias, and of "the evidence" as transparent,neutral, and authoritative. If a scholar can show that most of the texts bearingon the question he or she has posed point to a certain conclusion, the tendencyof empirically-minded historians is to assume that the problem has been solved.Moreover, if a number of scholars have investigated a particular question, and ifthe majority of them agree about its solution, this may add to the convictionthat the broader issues have been settled. The authority of the communis opinioof "objective" scholars, in conjunction with the authority of "objective" texts,validates the assumption that the questions asked and answers offered by pastscholarship-and the form and terms in which those questions and answers werecast-are self-evidently meaningful. But if both the texts and the a prioriassumptions employed by scholars are ideological constructs, the issues ofauthority and meaning become much more complex.

As H. sees it, the fourth-century Athenian democracy was based on a set ofabstract principles: moderation was desirable; the law should exist separate fromand superior to social and political processes; the organs of government shouldbe distinct in their powers; and, in general, all forms and forums of publicdecision and action should be defined by clear boundaries and restrictions. Theseprinciples were reflected in laws and institutions, which were changed over timein order to conform more precisely to the principles. The sum of laws andinstitutions, which embody and conform to prior principles, is the constitution.H. acknowledges that political realities require institutions to be somewhat moreflexible than their ideal forms, so that there is always a "gap between theconstitution and how it works" (p. 62); but he regards these exceptions as lessimportant than the notion of how the institutions should have worked.

The abstract principles on which the constitution was built are, for H., insome sense exterior to the social matrix and to the realm of practical politics.Because of these guiding principles, one is to assume, constitutional developmentwas linear, teleologlcal, and "Whiggish": changes aimed at a definite end, andthat end was good; thus, fourth-century constitutional evolution proceededinexorably toward the creation of an orderly, legalistic, and moderate demo­cracy, These constitutional and teleological assumptions are always implicit inH.'s writing: for example, "Athenian democracy was, ideally, based on theprinciple of rotation .. ." (p. 61); "The debate [in Assembly] was a one-waycommunication from the speaker to the people. In principle there was nocommunication from the participants to the speakers,,;7 "The Athenians of403 wanted to replace the radical democracy of the fifth century with a moremoderate form of democracy," and as a result "the powers of the ekklesia weresomewhat reduced" (p. 94).

H. emphasizes the centrality of institutions in defining the political order whenhe argues against the existence of a ruling political elite that used constitutional

7. Page 71; pace H., the text of Aeschin. 3. 2 (a paraphrase of a "Solonic" law liepi tii~ tlilv P'1tOpwveUlCoollia~) does not prove that comments from the audience violated "the leller of the law" (althoughthis may be the impression Aeschines wished to convey): the clause <iiOllep 01 VOIIO. IlpootattOUOI needrefer only to tiii IIp£o~utatq> tlilv Ilo}"itwv, and not necessarily to the words aveu aopli~ou lCat tapaxii~.The actual text of the law is, of course, lost.

326 REVIEW ARTICLEREVIEW ARTICLE 327

institutions to mask its true power (pp. 85-86). H. formulates the nonconstitu­tionalist position as follows: "don't believe that you understand Athenian demo­cracy by learning about the boule, the ekklesia and the dikasteria. Look insteadfor the real power exercised by leading politicians, influential families andpolitical groups." H. argues that this approach is anachronistic; he outlines theconstitutional decision-making process and concludes (p. 86): "Behind this con­stitutional framework there are no traces of informal organizations correspond­ing to political parties or interest groups in modern democracies.,,8 Yet thisexclusionary argument is far from complete, since parties and interest groups ofthe sort found in modern democracies hardly exhaust the roster of extra­institutional forces that might have contributed to Athenian political life.

The Athenian elite need not have been a ruling elite, a political party, or aformally organized interest group in order to have influenced public decision­making. The distinction between elite and nonelite citizens was a matter of greatconcern to Athenian citizens of all classes. As both philosophical and rhetoricaltexts show, Athenians knew very well that differences in birth and economicclass had political significance.9 H. acknowledges in passing the "constant andimportant opposition between social groups" (p. 86); but he usually ignores theinfluence that social tension or conflict had on the public realm.

Although H. understands that social realities and political practice are relatedin important ways, he generally subordinates socio-political questions to institu­tional concerns. For example, when discussing how many citizens attended atypical meeting of the Assembly he does address the socio-political questionwhether pay for attendance was adequate to allow laboring citizens to partici­pate (pp. 11,47-48); yet H. (following Ath. Pol. 41. 3) assumes that the intro­duction of pay was motivated by the constitutional requirement of a quorum.One might argue instead that the reform occurred because the Athenian lowerclasses were' determined to prevent the domination of the Assembly by the upperclasses, who were seen (rightly or not) as likely supporters of oligarchy.

H. usually conceptualizes the Athenian citizenry as a political institution:citizens are defined in relation to the institutions of government, before whichthey were (for the most part, and at least in principle) equal. Political equalitywas clearly a chief component of an Athenian's self-image, but an exclusivelyinstitutional definition of the citizen obscures an important political factor: theways in which subgroups within the citizenry defined themselves in relation toeach other. A citizen of democratic Athens lived both in a constitutional realmwhere political equality was the norm and in a social matrix where inequalitypredominated. The voters in Assembly, lawcourt, and Council were institution­ally equal but socially unequal; the dissonance between the two spheres sig­nificantly influenced decision-making.

8. As an example of the nonconstitutionalist position H. cites w. R. Connor, New Politicians inFifth-Century Athens (Princeton, 1971), pp. 4-5; on the influence of political groups in fourth-centuryAthens, see B. S. Strauss, Athens after the Peloponnesian War (Ithaca, 1986). Neither Connor norStrauss sees Athenian political factions as parties or interest groups of the modern type.

9. The relevant texts are discussed in J. Ober, Mass and Elite in Democratic Athens (Princeton,1989); cf. R. Seager, "Elitism and Democracy in Classical Athens," in The Rich. the Well-Born. and thePowerful. ed. F. C. Jaher (Urbana, 1973), pp. 7-26; E. C. Weiskopf, "Elitevorstellung und Elitebildungin der hellenischen Polis," Klio 43-44 (1965): 49-64.

Much of H.'s work has emphasized the separation of powers that obtainedamong the various bodies of government. 10 Central to The Athenian Assembly(esp. chap. 4) is the attempt to define the scope of the Assembly's powers and toshow how those powers were limited by the powers of other bodies, especiallythe people's courts. But H.'s strong arguments (pp. 61-62, 114) for the pre­valence of amateurism in Athenian government tend to weaken his thesis con­cerning the separation of powers. If each body within the Athenian governmenthad been controlled by an entrenched bureaucracy concerned to maintain itsown position vis avis other bureaucracies, the separation of institutional powerswould indeed be a key issue. 11 But since (as H. demonstrates) government bodieswere staffed by large numbers of amateurs-citizens chosen more or less atrandom from the whole social spectrum-the resolution of class tensions withineach institution (and ultimately within the society as a whole) may be for manyreaders-as it may have been for the Athenians-a more consequential matterthan the formal separation of powers between institutions.

Looking at the function of democracy as a social as well as a politicalstructure illuminates the ways in which the Athenians themselves thought aboutand used their government. Such an approach suggests alternatives to the linear,causal relationship that H. traces between formal principles and institutionalchange. By concentrating on the climate of opinion, and on the ways in whichopinions were formed and expressed, we can see how institutions operated andwhy they evolved. This is not to say that institutions were epiphenomenal;Athenian political ideology (the set of ideas about the public realm common tomost citizens, sufficiently coherent to lead to action but less formally organizedthan theoretical principles) evolved within institutional contexts. t2 But institu­tions and the relations between them are only part of the story, and notnecessarily the fundamental part. If we view Athenian democracy as a socio­political (rather than a constitutional) phenomenon, and if we take the primary"domestic policy goal" of the democracy to be the resolution of tensions result­ing from social inequality (rather than the achievement of an orderly, moderategovernment), we must ask a set of questions different from those H. poses.Consequently, we must take a different approach to rhetorical texts-an ap­proach that calls into question some of H.'s conclusions about the meaning ofthe two components of the word 011llolCpatia.

Although H. is conversant with epigraphy and is able to marshal complexarchitectural arguments (e.g., pp. 17-19), his most valuable source is the corpusof Attic orations. I) Most extant speeches were delivered by elite private litigantsor public speakers (pijtOP£~) attempting to influence large bodies of (mostly)nonelite citizens. Rhetorical strategies and topoi can reveal much about the

10. See esp. pp. 98-124; cf. H.'s "Initiative and Decision: The Separation of Powers in Fourth­Century Athens," GRBS22(1981): 345-70.

II. As it is in modern governmental systems; see, e.g.. J. K. Tulis, The Rhetorical Presidenq(Princeton. 1987), pp. 9-12, with the literature cited there.

12. On ideology and institutions in ancient and modern democracies, respectively, see M. J. Finley,Po/it;('$ in the Ancient World (Cambridge, 1983), and S. Bowles and H. Gintis, Demoaacy andCapitalism (New York, 1986).

13. Cf. p. 3 and pp. 230-41 (the "Index of Passales Cited;; citations of orations are about equal innumber to citations of all other sources combined.

328 REVIEW ARTICLE

REVIEW ARTICLE 329

social background of Athenian politics. But even a scholar primarily interestedin institutional procedure must be conscious of rhetorical context. Athenianpublic speeches were addressed to enthusiastic amateurs, not to judges or pro­fessional lawyers. The context demanded a rhetoric based on personalities andon appeals to class and status-group interests, to patriotic aspirations, to pre­judices, hopes, and fears. The discourse of the IltKUO'ti)plOv and tKKATJoia,therefore, relied primarily, not on the language of constitutional law, but onmore subtle and less logically rigorous appeals to popular ideology.14

Descriptions of political and legal practice offered by the orators are seldomimpartial accounts of constitutional procedure and precedent; and yet they areoften treated as such by H. He is not unaware of the problem: in a discussion ofa passage from Hyperides' For Euxenippus H. notes that the speaker uses theterm pi)'twp in two ways, as a legal term and in the sense belonging to "commonusage" (p. 62). But if the ordinary usage of key words differed from their legaldefinitions (as established by their use in sometimes-archaic laws), and if theorators deliberately confused common usage and legal definition to suit theirown purposes, then decontextualized rhetorical passages must be handled verycautiously as clues for restoring the details of a constitutional order.

A few examples point out the dangers of divorcing rhetorical passages from theirideological context. H. states (p. 39) that the nine npOEOpOt who presided over meetings ofthe Assembly "seem to have had wide powers. They could refuse to put a proposal to thevote.... " The source cited in support of this contention is Aeschines 2. 84; Aeschinesdescribes a meeting of the Assembly held a few years earlier-a meeting he was sure thejurors would recall-at which Aleximachus proposed a motion. After the motion had beenread out, "Demosthenes arose from among the npoEopol and refused to bring the motionto a vote." So far so good; but Aeschines' description of the aftermath disproves H.'sconclusion; "But you [i.e., the jurors, who are here simply identified with the Assembly­men] shouted and called the npoEopOl to the Pii~Hl, and so against [Demosthenes'] will themotion was put to a vote."

Aeschines relates the incident to demonstrate Demosthenes' undemocratic attitude andto prove his opposition to a treaty. He casts Demosthenes in the role of the arrogant andheadstrong individual who attempted to subvert the will of the people-and failed. Thereis no implication here that Demosthenes, as np6Eopo~, could legitimately withhold from avote a proposal on which the majority of Assemblymen wanted to vote. The incident israther an example of the power of the people to override any attempt by a magistrate touse his position in an arbitrary way. If the passage proves anything about "constitutional"authority, it is that npoEopol had no independent power to block proposals, and that ifthey attempted to do so, the people could summon them to the Pii~a and force the vote.The "summoning" itself is an example of the sort of communication from audience tospeaker that H. says was forbidden "in principle." In this case, emphasizing the putativeconstitutional principle at the expense of practice (which is discussed on pp. 69-72)obscures the fundamental importance of two-way communication in the praxis of theAssembly.

H. states elsewhere (p. 139, n. 51) that "we know from Dem. 21. 193 that garrison troopswere not allowed to attend the assembly." In the passage cited, Demosthenes describes his

14. I do not mean to imply that public rhetoric was empty of content, or that speakers and audienceswere unconcerned with arriving at good and just decisions, but thaI jusl decisions are not necessarilydependent upon legal language and precedent; cf. Ober, Mass and Elite, pp. 43-50; S. C. Humphreys,"Law as Discourse," His/ory and Anthropology 1 (1985); 241-64.

opponent Meidias' intention to repeat the "slander" he had made at the npOpOATt, defam­ing both the oii~o~ and the tJeKATJoia by claiming that the Assembly in which he wasconvicted "was composed of men who had stayed home when they ought to have marchedout, and who had left the fortresses deserted" (ro~ 0001 otov f:~ltVQl Katt~EvoV Kai 0001

to. <pPOUPI' i]oav EPTJ~a AEAOI1tOtE~, E~EKATJcriacrav). Here H. finds a formal legal pro­hibition in a claim put by a partisan speaker into his opponent's mouth about what oughtto have occurred but did not (the citizens in question ought to have marched out andmanned the foilresses but instead were in the city and so were able to attend an Assembly).Even if we assume that Meidias really made the claim Demosthenes attributes to him,there is a significant difference between "should not have" (according to one speaker) and"were not allowed to" (according to law). Perhaps there was a law prohibiting garrisontroops from attending the Assembly; but this passage does not demonstrate its existence,language, or purpose. Rather, the passage illustrates Demosthenes' tactic of attributing toMeidias a hatred of the oii~o~, here allegedly manifested in Meidias' willingness to insultthe people by claiming that they shirked their military duties.

In support of his theory that the powers of the lawcourts were separate from andsuperior to the powers of the Assembly, H. points out that "when ekk/esia (or demos) anddikasterion are used as contrasting terms and mentioned in order of precedence, dikaste­rion invariably takes precedence over ekk/esia (demos)" (p. 105; cf. p. 91). In five of the sixpassages that H. cites, the speakers do indeed praise the quality of decisions made byjuries, as opposed to decisions made in Assembly.ls But must we conclude that thesecomments were based on a constitutional separation and elevation of judicial power? Eachof the passages that H. adduces is a statement made by a litigant facing a popular jury.One might hypothesize that the speakers deliberately flattered their audience by makingthe most of the jurors' responsibilities. An Athenian olKacrtTt~ might well look kindlyupon, and pay close attention to, a litigant who had just informed him that his voterepresented the most authoritative decision made in the public realm.

The passages that emphasize the separateness and superiority of olKacrtTtPla (vs. oii~o~

or EKKATJcria) should be read in conjunction with other passages that assume a congruitybetween decisions made by jurymen and those made by Assemblymen. Litigants some­times warned jurors that their decisions would be closely monitored by the oii~o~, andsuggested that jurors should make a decision that would please the oii~o~. 16 The tendencyof litigants to remind jurors of the decisions "you [sc. jurors]" had formerly made inAssembly likewise argues against a strict separationist interpretation. 17 But, as we haveseen, H. believes (p. 97) that "a proper understanding of the term demos disposes of theview that ... the dikasteria were manifestations of the people (the demos) . ... Only theekk/esia was a manifestation of the demos. ... [N]either the nomothetai nor the dikastaiwere believed to embody the demos (Le. the whole of the people)."

H. has perhaps put too much weight on the habit of public speakers of referringto Assemblymen, and not jurors, as 6 OTtj.lOC;. The primary meaning of 0Ttj.lOC; tothe Athenians was not "Assemblymen," but the "the whole of the Athenian citizenbody." This latter meaning, which we might represent by the word "Demos," wasan ideological construct. Demos was real, in that there were indeed some twentythousand or (p. 8) thirty thousand individuals living in fourth-century Attica whoenjoyed full citizen rights; but Demos could not be perceived by the senses. No

15. Page lOS, n. 675; in the sixth passage (Oem. 24. 78) neither the lerm oiil1o<; nor Ihe termtl<l<A'1oia is used, and il is not cerlain thaI either term is implied.

16. Oem. 21. 2, 227; Lys. 12.91,22. 19; Dinarch. I. 3, 106,2. 19,3. 16; cf. I. 84.17. E.g., Oem. 19.224,21. 214-16; Hyper. 5.17; Lys. 19.14; Aeschin. 2.84,3.125; Dinarch. 3. 19;

Isae. 5. 38. H. discusses Ihis issue in "Demos. £Celesia, and Dicas/erion." pp. 135-36.

18. Citizen population: M. H. Hansen, Demography and Democracy (Herning, 1985). Face-to-facecommunity: Finley, Politics in the Ancient World. pp. 28-29, 82-83. "Imagined community": B.Anderson. Imagined Communities (London, 1983). On the origins and development of "the mull i­purpose word demos," see D. Whitehead, The Demes of Allica (Princeton, 1986), pp. 364-68.

19. Shown being crowned by the female deity Democratia on the stele-relief adorning the v61'0~against tyranny: SEG 7. 87: cf. A. Raubitschek, "Demokratia," Hesperia 31 (1962): 238. Since the lawwas passed and the stele authorized by VOl'06ttQI, not by an Assembly, the figure must representDemos qua imagined community, notlii\l'o~ qua ElCKA'1olQ.

20. See Ober, Mass and Elite. p. 299 with n. 14; R. Sealey, "The Origins of Demokratia, " CSCA 6

(1973): 283.21. t.lIcQotQi: see the passages cited above, nn. 16 and 17; that Demos' will was symbolized by the

decisions of VOl'06EtQl is demonstrated by the stele described in n. 19 above.22. See Webster's Third New International Dictionary, s.v. "representative": " ... constituting the

agent for another esp. through delegated authority; ... of. based upon, or constituting a form ofgovernment in which the many are represented by persons chosen among them usu[ally] by election."

331REVIEW ARTICLE

23. The Invention ofAthens. trans. A. Sheridan (Cambridge, Mass., 1986), p. 176.24. Praising the divinity of '1'1\1''1, Aeschines (2. 145) defined it as what resulls "when to nAi\60~ twv

no),lt<JlV of their own will ... say that something is the case."

the various institutional "parts" of the citizen body (EKKA11oiu, OIKOO'tT]pta,VOlloOt'tUt, ~OUAT]) could stand for and refer to the whole citizen body. Oratorscould speak of jurors as having made decisions in Assembly because both a juryand an Assembly were parts of the whole. The words Of}Il0C; and 'A011vutOt(which properly denoted the whole of the citizen body) could be used to refer,respectively, to the "part" of the citizen body that attended a given Assembly orsat on a given jury. Synecdoche takes us from the vocabulary of constitution tothat of signification, but this should cause difficulties only for those who assumethat the Athenians were concerned more with legal definition than with symbolicreference. Close reading of speeches by fourth-century orators, which NicoleLoraux has described as "the only [Athenian] texts genuinely inspired by demo­cratic thinking," reveals the importance of symbolic reference in the Athenianpublic realm. 23

H. tends to downplay the significance of the ideological construct Demos,because his approach necessarily emphasizes institutions-thus, of}ll0C; as EK­KA11oiu. But without the concept of Demos as the implied authority behind alldemocratic bodies, an entity that transcended and sustained all public institu­tions, Athenian political life is incomprehensible. In the passages I have citedabove, otKuo'tUi were reminded by public speakers of their duties to Demos bothas jurors and as Assemblymen. The procedure of 'YPuq>i) nupuvol103V gaveDemos a chance to consider, at some remove, decisions made in Assembly.When a jury overturned a 1jIT] q>tOllU by convicting its proposer, this was a resultof Demos' having turned against the proposer or having reconsidered the origi­nal decision; the jury's action did not override Demos' authority. We must keepin mind the role of informal (extra-institutional) communication in Athens. 24 Ifrumors circulating in the agora suggested that Demos was unhappy with a1jIT]q>tOIlU, some ambitious citizen might be encouraged to take the risk oflodging a 'Ypuq>i) nUpUVOllli3V against the proposer, since a successful indictmentwould be likely to gain him the sympathy of Demos; and this sympathy would inturn stand him in good stead in future contests in the Assembly and thecourtroom.

The primary principle involved in democratic decision-making at Athens wasnot separation of institutional powers but demotic control of the public realm.Demosthenes makes this clear in his speech Against Timocrates. The speechemphasizes the important role of the laws and the lawcourts (and so is cited byH. some 85 times; cf. p. 235). But "what, then, is the only honest and trust­worthy safeguard of the laws? You: ot noHoi" (24. 37). Demos was a socio­logical construct (ot noHoi) as well as a political construct (ot nOAt'tUt), anddemocracy worked by balancing political equality against social inequality. Thearbitrary powers of the elite were regulated through the forms of debate anddecision in the courtroom and the Assembly. In both institutions the audiencerewarded and punished speakers through 06pu~0C; (cf. pp. 70-72) and by thefinal vote. The ordinary citizens were suspicious of elite pretensions; contestants

REVIEW ARTICLE330

one had ever seen the whole of Demos; it was too big ever to gather in anyoneplace. Therefore, the Athenian Demos was not, as M. I. Finley supposed, a"face-to-face society," but rather an example of what Benedict Anderson hascalled an "imagined community.,,18 This imagined Demos was, however, afundamental and vivid political concept: Demos could be personified as a mature,bearded man;19 an anti-democratic coup would result in Demos' being over­thrown: KU'tUAUEtV 'tov of}llov was the most common periphrasis for counter­revolution. 20 This imagined Demos was the of}Il0C; assumed in the word °11110­Kpu'tiu: the entity that held power in the state.

A meeting of the Assembly was open to all citizens, and decisions made bythose who attended-"the demos in the narrower institutional sense" (p. 97)­certainly symbolized the will of Demos; but the participants at a given Assemblywere not identical to Demos. Nor, certainly, were juries or boards of VOlloOt'tUt,bodies that were limited in size and that excluded citizens under the age ofthirty. These added restrictions may explain the convention of addressing jurorsas 'A011vutm rather than as of}Il0C;, but decisions of VOI.100t'tUt and OtKu(J'tui,like decisions of Assemblies, symbolized the will of Demos.

21H. argues that,

unlike the EKKA11oiu, OIKUO'tT]ptU did not manifest or embody 0 of}ll0C; but were"representative of the demos" and were "assembled to act on behalf of thedemos" (p. 104; H.'s emphasis). The distinction is, for H., significant, because"representation ... implies distinction and not identity."

H. points out (pp. 102-4) that the concepts of "delegation of powers" and"committee" cannot easily be applied to analyses of Athenian government; butwe should recognize that "representation," "embodiment," and "manifestation"may be equally troublesome. "Representation," in a constitutional sense, impliesdelegation of authority and formal appointment or election.

22Since H. argues

that the authority of the lawcourts was not delegated, and since the Athenianjuror was not appointed or elected to represent a larger constituency, "represen­tation" does not seem the best way to conceptualize the relationship betweenlawcourts and Demos. "Embodiment" and "manifestation" imply identity, butthe Athenians at least sometimes perceived a distinction between Demos andthose who attended a given Assembly (e.g., Dem. 21. 193, noted above; otlier

examples cited p. 138, n. 40).I would suggest, as an alternative. the concept of "synecdoche," a figure of

speech in which a part stands for and refers to a whole, or vice versa. Each of

25. These ellamples, taken from pp. 107-8, could be multiplied.26. The problem is not solved by replacing the question "What institution is sovereign?" with "What

institution is KUPLO~?"; for as H. points out (pp. 106-7), the answer to the latter question (as supplied byrhetoricaltellts) varies with the contellt. NOIlOI are sometimes said to be KUPIOI (p. 106 with n. 681); butVOIlOI had no independent power of action, and no effective enforcement apparatus ellisted apart fromthe will of the people: cr. Oem. 21. 223-24.

27. Pp. 105-7; cf. J. W. Allen, A History of Political Thought in the Sixteenth Century (London,1941), pp. 247-70 (English theorists), 394-444 (Jean Bodin).

28. uviathan (London, 1839), pp. 153-84 (chaps. 17-19),297-307 (chap. 28),322-43 (chap. 30).

in the Assembly and the courtroom were constrained to devise rhetorical formsthat would be acceptable to their mass audiences. Elite citizens' involvement inthese institutions entailed their acknowledgment-in both the content and thestructure of their discourse-that the will of the many was the will of Demosand that Demos was the legitimate political authority. The ultimate lCpatOC; ofDemos lay in this demotic control of the public processes of signification: thepower of the many to participate actively in assigning political meaning tosymbols deployed in public speech.

The imagined community, Demos, provides the missing subject that wouldallow H. 's many passive clauses to be recast in the active voice: "legislation wasconferred on the nomothetai ... "; "the ekklesia was deprived of jurisdiction ...";"the people were restricted to the passing of decrees and the election of offi­cials ..."; "the people were entrusted with the ad hoc election of envoys....,,25Without the concept Demos there is no active agent for these passive verbs,other than the term "polis," which is unsatisfactory in this context (cf. p. 178,n. 664: "It is not the demos, but the polis which appoints the dikasleria, " citingDem. 21. 233); the polis cannot be characterized as a political agent distinctfrom "the will of the citizen body." Through the operations of its parts, Demoseffected the various institutional changes described by H. These reforms, whichH. has done much to clarify, changed the external forms in which power wasdeployed but never limited the collective power of the Athenian citizenry.

H. has often conceptualized political power as "sovereignty," a term he hasused frequently over the years and retains in the conclusions of the present book(p. 129: "In fourth-century Athens the ekklesia was no longer 'the sovereignbody of Government' "), even after he has pointed out some of the difficultiesassociated with its use (pp. 105-7, esp. p. 106: "Many problems are avoided if wedismiss the concept of sovereignty ..."). The reasons for this appaient ambiva­lence are not clear. But we must face the issue squarely: if "sovereignty," definedas unitary power with a specific institutional locus, is a concept without meaningfor Athenian democracy, then the question "Was the ElClCATJoia sovereign?" isitself no longer meaningful. 26

The concept of sovereignty was developed in the sixteenth and seventeenthcenturies by Western European political theorists writing on the institution ofmonarchy.27 Monarchical power is by definition unitary, since it is located in theperson of the monarch. Absolute monarchs-those whose sphere of publicaction is unlimited-are therefore unitary sovereigns. The theory of sovereigntywas given its classical formulation in English by Thomas Hobbes, who specifi­cally defined sovereign power as an institution that monopolizes the legitimateuse of coercive force and that is legitimately maintained by force. 28 Neither

333REVIEW ARTICLE

Hobbes nor any other early-modern political theorist considered democracy, aspopular sovereignty, to be either desirable or practical; they conceived of sover­eignty, properly so called, as unitary state power that resided, preferably, in theperson of the monarch, or else in a representative assembly. The traditionaltheory of sovereignty does not encompass the idea that legitimate power couldreside in an abstraction such as "the People"; consequently, it is of very limiteduse in explaining democracy.29

The concept of sovereignty can be applied usefully to democracy only byreplacing the idea of "sovereignty as located in institutions" with "sovereignty asthe ability to change institutions." Adam Przeworski suggests that "people aresovereign to the extent that they can alter the existing institutions, including thestate and property, and if they can allocate resources to all feasible uses. ,,30 Iwould add: "and so long as they also control the ideological symbols throughwhich political agents make decisions." By this extended definition the Atheniancitizenry was indeed sovereign: there was no entrenched ruling elite at Athensthat could successfully oppose the desire of Demos to alter institutions orallocate resources, and the people controlled the forums of public communica­tion (especially the Assembly and the courts) in which ideological symbols wereforged and deployed in open debate. 31

Redefining sovereignty to include power over discourse residing in the citizenry,as opposed to legal power residing in an institution, mitigates the problem raisedby the concept "separation of powers." The latter concept, an offshoot of thetraditional institutional theory of sovereignty, was first enunciated in the seven­teenth and eighteenth centuries, when English and French aristocrats attemptedto limit the powers of absolute monarchs. 32 Since the separation of powers ispredicated on a subdivision of unitary sovereignty and on the assumption thatsome of the sovereign's powers would pass into the hands of a ruling elite, theconcept as traditionally defined is not much more useful for our purposes than thetraditional theory of sovereignty. And once we dispense with separation of powersas a central issue, we can de-emphasize restriction and limitation as the operativeterms in which we conceptualize the nature of Athenian democracy.33

None of the preceding discussion is intended to cast doubt on the importanceof H.'s overall achievement, which is well represented (if not fully embodied) byThe Athenian Assembly. Individual democratic institutions certainly had dis­tinctly defined responsibilities in Athens, and these responsibilities were redefinedover time, H. 's question concerning "the extent of the sphere regulated by

29. Cf. Hobbes, uviathan. pp. 155 (chap. 17), 171-84 (chap. 19); Allen, History of PoliticalThought. pp. 436-48; Bowles and Gintis, Democracy and Capitalism. pp. 22-23, 167.

30. On popular sovereignty, see Bowles and Gintis, Democracy and Capitalism. pp. 181-83, quotingPrzeworski on p. 182.

31. Cf. Ober, Mass and Elite. esp. pp. 293-339.32. For the classic formulation, see Montesquieu, The Spirit of the Laws. trans. T. Nugent, vol. I

(Cincinnati, 1886), pp. 171-209 (Book II).

33. M. FoucaUlt, The History of Sexuality, vol. I: An Introduction, trans. R. Hurley (New York,1980), pp. 81-102, attacks the emphasis that is placed on restriction and limitation when power isConceived as legal sovereignty. S. L. Collins, From Divine Cosmos to the Sovereign State: Conscious­ness and Order in Tudor England (Ollford, 1989), argues persuasively that political order was firstasSociated with separateness and difference in the seventeenth century, and he demonstrates how thisnew aSsociation led to the idea of representative sovereignty.

REVIEW ARTICLE332

334 REVIEW ARTICLE

decrees of the people" (p. 109) or by other constitutional instruments is valid,even if it does not fully define the nature of Athenian democracy. No student ofclassical Athenian political life (even one who has read all of H.'s other, morespecialized articles and monographs) can afford to ignore this seminal book.Those who remain unconvinced by some of H.'s conclusions will learn muchfrom the attempt to refute them. It should be obvious to readers of this articlethat thinking about the implications of H.'s theses has substantially aided myown attempt to understand Athenian democracy. Whether or not H.'s institu­tional model of political life becomes the accepted paradigm, The AthenianAssembly will be the standard reference manual on the EKKA.lloia for theforeseeable future. It should quickly establish itself as a classic in the field ofancient constitutional and legal history and deserves to find a broad audience.

JOSIAH OBER

Montana State University