2 If not indicated otherwise, quotations refer to the Politics. The Nicomachean Ethics I abbreviate as NE. I use Robinson's, R. translation for the Politics, Books III and IV (Oxford: Clarendon Press 1995)Google Scholar; Rackham's, H. translations for the remainder of the Politics (Cambridge, MA: Harvard University Press 1990)Google Scholar and the Nicomachean Ethics (Cambridge, MA: Harvard University Press 1994); and the translations in Barnes, J. The Complete Works of Aristotle Vol 1 and 2 (Princeton: Princeton University Press 1984)Google Scholar for other Aristotelian works.

3 Kahlenberg, K. ‘Zur Interpretation von Buch III der Politik,’ in Steinmetz, P. Schriften zu Politika des Aristoteles (Hildesheim: Olms Verlag 1973) 102–81Google Scholar, at 119, points out that this is what Athenian history has taught Aristotle, which is also reflected in his writings about the development of the Athenian constitution.

4 To kyrion does not have to be given to a person. It may belong to the law, cf. 1281a36. Thus translating kyrion as sovereign is misleading, since ‘sovereign’ is reserved for persons in modem political thought.

5 This approach to power is not restricted to just claims. Somebody with military power may have no just claim, but it would be imprudent to ignore him. Aristotle investigates how to accommodate all claims without jeopardizing the good life. Passages supporting this view are 1318b33-6 and 1319a2-4. (Cf. also 1268a25: parties not sharing in power are not friendly towards the state; and 1270b18: the state can only persist if all of its sections want to preserve it; in 1329a23-25 a happy state is a one in which all sections are happy; in 1297a7-8 we learn that mixed constitutions ensure permanence, and mixed constitutions are those that try to balance different claims.)

6 I use the name ‘accumulation argument’ and not ‘summation argument’ (cf. Braun, E. ‘Die Summierungstheorie des Aristoteles,’ in Steinmetz, 396–424Google Scholar; Schuetrumpf, E. Die Analyse der Politik durch Aristoteles [Amsterdam: Gruener 1980]Google Scholar) to have a term covering ‘summation arguments,’ but not restricted to conceptualizing the ascription of the property in terms of some summation. Examples of AAs that are summation arguments are those in which the property is measurable wealth (estate, gold), voting power measured in number of votes, or military power. In summation contexts, the first guiding question is easy to answer. For that reason, accumulation arguments that are not summation arguments are more interesting conceptually.

7 Waldron, J. ‘The Wisdom of the Multitude,’ Political Theory 23 (1995) 563–84CrossRefGoogle Scholar

8 For recent discussion cf. Bookman, J. ‘The Wisdom of the Many: An Analysis of the Arguments of Book III and IV of Arisotle's Politics,’ History of Political Thought 13 (1992) 1–12Google Scholar, Garrett, J.E. ‘The Moral Status of “the Many” in Aristotle,’ journal of the History of Philosophy 31 (1993) 171–89CrossRefGoogle Scholar, J. Waldron, ‘The Wisdom of the Multitude'; and for earlier discussion cf. K. Kahlenberg and E. Braun. Other literature relevant to the Politics in general will be mentioned as we proceed. The most extensive recent analytical study of the Politics, Miller, F.D. Nature, justice, and Rights in Aristotle's ‘Politics’ (Oxford: Clarendon Press 1995)Google Scholar, only briefly touches on the issues relevant for this paper; cf. 261-2.

9 It has been pointed out to me that pursuing the second guiding question is not required, since Aristotle simply takes for granted that executive functions are filled by individuals elected from the citizen body (cf. chapter 5.4 in Miller on rights to office). However, once we raise the question of whether groups can make claims to power on the same grounds as individuals, we also need to ask whether there are other considerations addressing differences between individuals and groups that may interfere with such arguments. That inquiry is introduced by the second guiding question.

10 The multitude may also claim power on quantitative grounds: if they were excluded from office, the city would be full of enemies (cf. 1281a28-33, 1281b30). The multitude can also make claims on military grounds (cf. 1279a40-b5).

11 In 1332a33-5 we learn that the state is spoudaios, because of the virtuous citizens who share in government, cf. also 1283a25-26. Even though the list of qualities that Aristotle considers as supporting claims to power is long, virtue occupies a special place among them; cf. 1280b7-8, 1283a25, 1301a39-b1, 1303b15-16.

12 In a presentation to the Princeton Political Philosophy Colloquium, chaired by George Kateb, in November 1998, Richard Kraut (presenting his unpublished paper ‘Realism and Idealism in Aristotle's Politics’) argued that Aristotle uses a lower standard of evaluation for groups than the one he uses when praising the rule of virtuous individuals. As the use of the vocabulary discussed in this subsection in particular and my discussion of Aristotle's argument in general should make clear, this reading is ruled out by what Aristotle actually says about groups. This point, however, by no means affects Kraut's overall project in his paper.

13 The situation is more involved, but that should not matter here; cf. Broadie, S. Ethics with Aristotle (Oxford: Oxford University Press 1991), 67–71Google Scholar.

14 In our passage, that is confirmed because the group may be better, beltios, than the spoudaios anēr. Beltios, however, is the comparative of agathos. So agathos and spoudaios are at least very close, otherwise this comparison would not make sense 1276b and 1277a support this claim.

15 Literally, this passage speaks of the free men and the multitude. But it is safe to assume that Aristotle does not think of the multitude as including people other than freemen. In 1326a18-29 he points out if he talks about the multitude, he does not include slaves, resident aliens, and foreigners. It seems that this holds true for our passage as well.

16 Aristotle expresses reservations concerning liberty: living as one likes is not good, cf.1310a30-6, 1316b24-5, 1317b10-12, 1319al-2; it cannot guard the evil in everybody. Yet people like living in such a way, 1319b31-33. For Aristotle on liberty, cf. Barnes, J. ‘Aristotle and Political Liberty; in Patzig, G. ed., Aristoteles’ ‘Politik'-Akten des XI. Symposium Aristotelicum (Gottingen: Vandenhoek und Ruprecht 1990) 249–63.Google Scholar

17 Note that, in 1308b28, Aristotle contrasts the multitude with the respectable, epieikeis.

18 For a discussion of ‘the many’ and related notions, cf. Garrett, who argues that Aristotle uses a ‘ladder of characters’ (189) in which the good or virtuous (i.e., the spoudaios anēr) stand on top, then we have the continent, the incontinent, ‘the many’ and the fully wretched (whom Garrett understands as a subgroup of ‘the many’).

19 I take this point to be a serious objection to Bookman, Schuetrumpf, and Waldron. Their discussion leaves the exploration of the argument at a fairly superficial level. That is, it leaves the reader wondering just what is happening in such an assembly that virtue and practical wisdom seem to appear out of nowhere. Leaving the argument at that rudimentary stage, moreover, is entirely unnecessary, because Aristotle provides resources to go much further.

20 Aristotle uses the analogy of art. A piece of art assembles properties that are normally scattered. Aristotle points out that there may well be an individual superior to what the painter has drawn in some regard. Yet overall, the painted motif is superior. So the good man does not have to achieve the maximum in all specific virtues; rather, he assembles many strengths.

21 One may argue that morion can only express the second possibility, whereas the first would be expressed by meros. But meros and morion are too close to warrant that, and Aristotle's usage treats them as quite close, cf. for instance 1326a21-2.

22 The questions of (a) how to think about Aristotle's usage of the concept of nous both in his ethical writings and in writings such as the Posterior Analytics, and of (b) how to lay out the relationship between nous and the two types of virtue go beyond the scope of this essay. NE 1144b10-17 is the central statement about nous that I need, i.e., if somebody possesses natural virtue he possesses something different from full virtue, and then it takes nous to transform natural virtue into full virtue. In a more detailed exposition of these notions the proper place of habituation versus insights of reason would have to be explained.

23 Just what is the work done by nous? Nous has both a theoretical and a practical role to play (cf. NE 114la2-8, 1143b 1-5): its theoretical role is the apprehension of general principles, and its practical role is the apprehension of particulars in deliberation. For a discussion cf. Broadie, chapter 4, sections X and XI. In the passages quoted here nous is thought of as a different state of mind or operation from phronēsis, no matter how much it is involved in phronēsis in practical contexts. Characteristically, nous is intuitive and immediate and not discursive and inferential.

24 Another passage that has implications for the question of whether virtues can come in degree is 1260a11-24, where Aristotle discusses how different kinds of people have virtues in different senses. He mentions that virtue can be ateles or teles, unfinished or finished, 1260al4-15. The word hosos is used in this context, too (1260al7, 1260a20). Here, Aristotle discusses both intellectual and moral virtue. So this passage supports the reasoning above. Cf. also 1260a34-6.

25 I run the second and the third account together and need to keep them apart from the first, as an answer to (b) only depends on how we understand what it is to be part of virtue.

26 It was done in Braun, E. and Schuetrumpf, E.; and Waldron, J. and Keyt, D. ‘Aristotle's Theory of Distributive Justice,’ in Keyt, D. and Miller, F.D. eds., A Companion to Aristotle's Politics (Oxford: Blackwell 1991)Google Scholar also criticized it. E. Braun criticizes Aristotle on the grounds that a summation cannot do, but does not explore other possibilities (175). Nichols, M. in Citizens and Statesmen: A Study of Aristotle's Politics (Savage, NJ: Rowman and Littlefield 1992)Google Scholar says: ‘Aristotle's argument in favor of the collective judgment of the many would be simply valid only if the city were not more than the sum of its parts’ (67). This underestimates the Argument from Virtue, as I want to show here.

27 A reference to Madison, J. Hamilton, A. and Jay's, J. Federalist Papers, Kramnick, I. ed. (London: Penguin 1987)Google Scholar might be of interest here: those authors were quite worried about the results of collective deliberation and therefore about its proper places in the American constitution. In the conclusion (No. LXXXV) we read that ‘the result of deliberation of all collective bodies must necessarily be a compound, as well of the errors and prejudices as of the good sense and wisdom of the individuals of whom they are composed’ (484). The topic comes up in particular in No. LXVIII, where the election of the president is discussed. There they point out that the immediate election should be made by ‘men most capable of analyzing the qualities adapted to the station and acting under circumstances favorable to deliberation.’ But even those distinguished men ‘are to assemble and vote in the State in which they are chosen, this detached and divided situation will expose them much less to heats and ferments, which might be communicated from them to the people, than if they were all to be convened at one time, in one place’ (393). For, as we read in No. LV, ‘Had every Athenian citizen been a Socrates, every Athenian assembly would still have been a mob’ (336). This pessimism is also shared in one of the earliest studies of social psychology, LeBon's, G. The Crowd (Atlanta: Cherokee 1982)Google Scholar: ‘In a crowd, men always tend to the same level, and, on general questions, a vote recorded by forty academicians is no better than that of forty water-carriers’ (191).

28 It is not clear that Aristotle means to say that the judgment of the many is equally valuable as the judgment of the wise, but 100b and 104a express a positive view on the judgment of the many. One needs to be careful, though, not to make too much of these passages and the others quoted in the text: Aristotle's confidence in some basic wisdom of the multitude tends to be expressed in his characteristically guarded manner.

29 This does not prevent Aristotle from claiming elsewhere that the majority speaks conventional wisdom, whereas the wise speak according to the standards of nature and truth, cf. Topics 173a25-30. He also points out that we should be careful not to believe in those arguments meant to trick the multitude (1308al). He may think of an unassembled multitude here, or even an assembled group of people who do not listen to each other and are not co-deliberators.

30 Waldron argues that Aristotle's own practice of reviewing different views (endoxa) can be taken to be a model for what is supposed to go on in such assemblies. Yet we also read that one should not consider just any view, Rhetoric 1396a32, NE 1095a28. For interesting skeptical reflections on collective decision making, also cf. Plato's Demodocus, section I.

31 This passage says that both intellectual and moral virtues are produced and increased by instruction. It seems fair to take this to imply that nous can come in degrees. I am not aware of other passages that explicitly support this view. However, it seems that nothing Aristotle says about nous elsewhere is inconsistent with this reading of the introductory line of Book II.

32 Schuetrumpf points out that there is a similar result in Plato's Gorgias ( 488d9ff): the many who are stronger than the few are also better (179). However, the result is achieved only because Callikles prematurely accepts the equivalence of strength and goodness. Schuetrumpf provides a number of interesting cross references to other passages in Greek literature mentioning the subject matter of this essay (174-85).

33 On politics and education cf. Lord, C. ‘Politics and Education in Aristotle's ‘'Politics,“’ in Patzig, 202–15Google Scholar; the different senses of education involved here are discussed on 207.

34 Yet he is skeptical of the impact of education on the many, cf. NE 1179b11-18. He doubts that the logos can reach them (tous de toioutous tis an logos metarruthmisai, NE 1179bl6-17; cf also 1179b27-9). But the logos distinguishes humans from other social animals (logon de monon anthrōpos echei tōn Zōōn, 1253a10-11) and is needed for public deliberation.

35 The connection between phronēsis and experience, empeiria, is of interest here: In NE 1103a14-16 Aristotle points out that intellectual virtues require instruction and therefore time and experience. In NE 1142a14-15, we read that the knowledge of particulars required by practical wisdom comes to the agent ex empeirias. In NE 1143b 13-14 we read about experience as a quality of elder people: dia gar to echein ek tes empeirias omma horosin orthos, i.e., experience has given them an eye for things. With regard to this visual metaphor it is plausible that in terms of experience, certain groups do better than persons. Bodéüs, R. The Political Dimensions of Aristotle's Ethics (Albany: SUNY Press 1993)Google Scholar, chapter 5, elaborates on Aristotle's insistence that experience is also required of the reader of the Politics.

36 On this theorem, cf. Grofman, B. Owen, G. and Feld, S. ‘Thirteen Theorems in Search of the Truth,’ Theory and Decision 15 (1983) 261–79CrossRefGoogle Scholar. It has recently come up for discussion as part of an interpretation of Rousseau's General Will; cf. Grofman, B. and Feld, S. ‘Rousseau's General Will: A Condorcetian Perspective,’ American Political Science Review 82 (1988) 567–76CrossRefGoogle Scholar and Discussion Note in the section on ‘Democratic Theory and the Public Interest: Condorcet and Rousseau Revisited,’ American Political Science Review 83 (1989) 1328-35; Estlund, D. Discussion Note in the section on ‘Democratic Theory and the Public Interest,’ American Political Science Review 83 (1989) 1317–22CrossRefGoogle Scholar; and Waldron, J. Discussion Note in the section on ‘Democratic Theory and the Public Interest,’ American Political Science Review 83 (1989) 1322–8Google Scholar; J. Waldron, in ‘The Wisdom of the Multitude,’ discusses it in connection with the key passage. The version presented here is not the original version, which assumed that each individual has a probability bigger than 1/2 of being right.

37 Cf. also Miller, F.D. 83-4 and Yack, B. The Problems of a Political Animal (Berkeley: University of California Press 1993) on this important passage.Google Scholar

38 For an interpretation of the passages on the god-like individual as king in the context of myths relating to earlier (heroic) times, cf. Bodéüs, R. Aristote et Ia Theologie des Vivants Immortels (Paris: Les Belles Lettres 1992)Google Scholar, chapter III.4.

39 Schuetrumpf points out that Aristotle himself criticizes AAs applied to happiness (185). The context is Aristotle's point that Plato's state in the Republic is not happy unless the guardians are (1264b19-21). The point there is that guardians really are straightforwardly unhappy. This supports the view that there must be some tendency of the individuals towards having the property at stake for accumulation to make any sense.

40 The beginning lines of the key passage make it dubious that Aristotle means this to be an argument of his own; Larsen, J. ‘The Judgment of Antiquity on Democracy; Classical Philology 49 (1954) 1–414CrossRefGoogle Scholar quotes some other passages from the Politics showing that Aristotle was not a democrat (4). But if we understand the term in any of the contemporary senses, that should be clear anyway. Nothing here depends on whether Aristotle regarded this argument as his own.

41 Aristotle's ideas about what a polis should be like assign public deliberation an important place. On the one hand, he points out that a polis is different from a household in consisting of dissimilar people (1261a23-5, 1263b32-9, 1276b28-9 and 40, and cf. 1255b12-14 about the shared interest within a household, i.e., among slave and master). On the other hand, he emphasizes the existence of a noteworthy common interest in a polis (1278b22), pointing out that under good constitutions the common interest is fostered, 1279a29); that in a polis, people share an understanding of right and wrong (1253a17-19); that a polis is a partnership (koinōnia, 12S2a1, 1276bl); that factions destroy a city (1272b14-15); that a city is a community of similar people, 1328a36 and not a random multitude, 1328b16-17; and that Babylon is too big (1265a14-15, 1276a28-9) to fit with Aristotle's ideas about how large a city should be (d. 1325b33-1326b25). The extent of homogeneity of the citizens is large enough to make successful deliberation possible, but not so large that coordination would be superfluous (d. 1263b30-40 for a good passage on how to occupy a middle ground between being too similar and being too dissimilar). Recall that it is the use of speech, amply called for in public deliberation, that makes humans ‘political animals’ to a larger extent than any other ‘gregarious animals’ (politikos versus agelaios, 1253a7).

42 Mill seemed to think that books, newspapers, and pamphlets could to some extent replace personal acquaintance with regard to issues of political importance in territorial states Mill, O.S. ‘Considerations on Representative Government,’ in Robson, J.M. ed., Collected Works of John Stuart Mill [Toronto: University of Toronto Press 1977], 547)Google Scholar. A de Tocqueville in Democracy in America, Mayer, J.P.; ed. (Garden City, NY: Doubleday 1969)Google Scholar repeatedly pointed out the comparative homogeneity of America (!), which he thought was an important factor in favor of democracy. So, e.g., we read that ‘There is no French province where the inhabitants know each other as well as do the thirteen million men spread over the extent of the United States’ (385). The Americans ‘do not always agree about the best means of governing well … but they agree about the general principles which should rule human societies … Men have the same ideas concerning freedom and equality’ (373). Or: ‘In. the United States one fact wonderfully smooths the existence of the federal government, namely, that the different states have … the same level of civilization, so it is always an easy matter for them to agree’ (167).

43 Yet another way of understanding the difference between pure and impure food is in terms of highly valuable food and edible food that contains impurities. For instance, some people mix expensive pure maple syrup with a larger quantity of cheap syrup. The resulting mixture is more useful than only a small amount of the pure syrup. However, the problem with this version is that it is hard to see what usefulness would amount to in the political scenario. For the point of adding the cheap syrup is merely to have ‘more syrup,’ and the usefulness of this operation would consist in its leading to a larger supply of syrup. There is no plausible analogy in the political scenario. This reading is stimulated by Robinson's translation, from which I deviate in the statement of the Mixing Argument because I think it fails to sustain the analogy.

44 As opposed to this line, in 1277b29 Aristotle points out that subjects do not need wisdom, but right opinion; this may be taken to mean that they do indeed have that at least.

45 One might also worry about the following: as for the analogies Aristotle mentions in this passage, we need to recall that they all refer to some technē that one does not need to have in order to judge its quality. Politics, however, is a science (NE 1094a25-9), and in sciences this seems dubious. This worry, however, rests on there actually being a systematic distinction between arts and sciences, but it seems that Aristotle does not stick to a systematic distinction between the two of them in the field of politics. So this worry is unnecessary.

46 For an illuminating historical discussion of such ideas mostly in French writers between the Renaissance and the times of Rousseau, cf. Keohane, N. Philosophy and the State in France (Princeton: Princeton University Press 1980)CrossRefGoogle Scholar. Of special interest for our context are her discussions of Claude de Seyssel (chapter 1.4) and Blaise Pascal (chapters 9.2—9.4). For a discussion of the history of the notion of general will, cf. Riley, J. The General Will Before Rousseau (Princeton: Princeton University Press 1986)Google Scholar. For a contemporary discussion of parallels between groups and individuals cf. Elster, J. Solomonic Judgments: Studies in the Limitations of Rationality (Cambridge: Cambridge University Press 1989)Google Scholar, chapter 4.

47 It seems there is only one exception, namely the well-known passage in section 96 of the Second Treatise, where Locke points out that ‘it is necessary the Body should move that way whither the greater force carries it, which is the consent of the majority.'

48 My presentation follows G. Sreenivasan ('What is the General Will?’ Philosophical Review forthcoming), whose paper on the general will strikes me as the most illuminating discussion of this concept available. In particular, I follow him in interpreting many remarks about the general will in terms of constraints on public deliberation. This way of reading Rousseau is likely to be controversial, and space does not permit me to argue for it thoroughly. However, the relevant point with respect to Aristotle can still be made in many interpretations of Rousseau that differ from mine; all I need is the centrality of an account of public deliberation constrained by certain conditions. Yet I should explicitly mention one important and wellknown passage in the Social Contract that seems to contradict the interpretation of the general will in terms of public deliberation: ‘Si, quand the peuple suffisamment informé délibère, les citoyens n’ avaient aucune communication entre eux, du grand nombre de petites differences resulterait toujours la volonté génerale, et la délibération serait toujours bonne.’ Rousseau seems to say here that the general will results from the citizens’ deliberating without communicating among themselves, and this in turn suggests that the general will would not be part of a theory of public deliberation. However, there are numerous other passages in the Social Contract that strongly support the reading that the general will does indeed emerge from public deliberation. Time does not permit me to present a detailed exegetical discussion, but relevant passages include II, 3 (beginning), II, 4 (middle), III, 4 (middle), N, 1, IV, 2 (end). In the light of such passages, the passage quoted above would have to be read as saying that the individual citizens make up their minds independently before they vote, even though they went through a process of public deliberation earlier.

49 For evidence that Rousseau did indeed think of the political community as one large person, cf. II, 6 x and III, 5. For a history of the concept of general will cf. Riley, who argues that its original context was theological, i.e., God's intention to redeem everyone was his general will to redemption. From there, the concept was gradually transformed in to a political one as used by Rousseau, according to Riley.

50 The most important constraints are, first, that the subject matter of the deliberation is perfectly general (Social Contract, II, 6), second, that the conclusions of the deliberation apply equally to all the members of the community (II, 4), third, that all members of the community participate in the deliberation (II, 2; II, 4) and, fourth, that all parties to the deliberation think for themselves (II, 3). Serving a different purpose, not all these conditions may be applicable to the Aristotelian scenario.