Aristotle on the Virtue of the Multitude

It is generally believed that one argument advanced by Aristotle in favor of the political authority of the multitude is that large groups can make better decisions by pooling their knowledge than individuals or small groups can make alone. This is supported by two analogies, one apparently involving a “potluck dinner” and the other aesthetic judgment. This article suggests that that interpretation of Aristotle’s argument is implausible given the historical context and several features of the text. It argues that Aristotle’s support for the rule of the multitude rested not on its superior knowledge but rather on his belief that the virtue of individuals can be aggregated and even amplified when they act collectively. This significantly alters our understanding of Aristotle’s political thought and presents a powerful alternative to the epistemic defenses of mass political activity popular today.


176
Political Theory 41 (2) which many contribute (ta symphorēta deipna) is better than a dinner provided out of a single purse (ek mias dapanēs). For each individual among the many has a share (morion) of excellence (aretē) and practical wisdom (phronēsis), and when they meet together, just as they become in a manner one man, who has many feet, and hands, and senses, so too with regard to their character (ta ēthē) and thought (dianoia). Hence the many are better judges than a single man of music and poetry; for some understand one part, and some another, and among them, they understand the whole (allo gar alloi ti morion, panta de pantes). 2 Waldron dubbed this argument "the doctrine of the wisdom of the multitude," or "DWM," and explicated it as follows. Aristotle thought that "the people acting as a body" were "capable of making better decisions, by pooling their knowledge, experience, and insight," than any single member of the body, however excellent, was capable of making on his own. 3 Mass political activity thus resembled a "potluck dinner" 4 or an audience judging a play: a greater variety of contributions could produce a better result. This reflected decision making in the Athenian assembly. Debating the invasion of Sicily, Waldron hypothesized, one citizen might know something about the coastline, another Sicilian military capacities, another the costs of such expeditions, and so on. From the diverse knowledge of the group there could thus emerge the "widest possible acquaintance with the pros and cons," and hence the best possible decision. 5 Similar accounts had appeared before, 6 but Waldron's was easily the fullest and most elegant and has continued to influence students of democracy and of Aristotle alike. 7 Certain details remain disputed: for example, whether the aggregation of views was fully "dialectical" or more "mechanical," 8 and whether citizens themselves possessed knowledge relevant to the subject at hand or simply knew who did. 9 But Waldron's three core claims are widely accepted. First, that Aristotle's "feast" is a "potluck dinner." This is supported by a later appearance of the same analogy, in which "a feast to which all the guests contribute" (hestiasis symphorētos) is contrasted with "a banquet furnished by a single man" (mias kai haplēs). 10 Second, that the multitude's contribution to political activity is diverse knowledge, wisdom, insight or expertise. 11 And third, that this reflects debate in an assembly, as in classical Athens. 12 Accordingly, Aristotle is widely regarded as a moderate supporter of deliberative democracy on epistemic grounds. 13 Yet there are reasons to doubt this view. The "pooling of knowledge" paradigm certainly fits the intellectual aspect of his argument, denoted in the text by phronēsis, "prudence" or "good sense," and dianoia, "thought." But how about its ethical aspect, denoted by aretē, "virtue" or "excellence," and ta ēthē, "moral characteristics"? Also, Athenian assemblygoers seem to have been better known even among democrats for ignorance rather than knowledge (especially in relation to the invasion of Sicily). 14 Is an epistemic account really probable in this context? What about the second analogy: can an epistemic approach adequately explain audience responses to artistic works? And could Aristotle really have found potluck dinners so appealing? Did the ancient Greeks even have potluck dinners?
None of these questions has been fully explored; those who doubt the argument have simply dismissed it as weak. 15 I think we can do better. Aristotle was not interested in the benefits of pooling diverse knowledge, but in the political authority of aretē, "virtue," understood in its general sense as a capacity for right action encompassing both ethical and intellectual qualities. He was concerned with the quantity of aretē that could be possessed by different agents, and his claim in this passage was that some multitudes, when they act collectively, can exhibit more aretē than even highly virtuous individuals. Specifically, I will suggest, he believed that all forms of virtueperhaps especially courage and justice, the two that he associated most with large numbers-are easier to practice in groups than alone, and this supported the view that a multitude could be an effective political agent.
This account fits recognizably within the familiar framework of Aristotelian "virtue ethics" (albeit with a twist that connects it to wider issues of collective action), which seems preferable to reading the passage as an outlier in Aristotle's writings. 16 But if my interpretation is sound, what explains the staying power of the former account? One factor is the difficulty of the text: it is highly truncated, leading many translators to expand upon it, inserting misleading terms in the process-thus giving false preconceptions to readers who have Greek, and precluding understanding among those who do not. 17 Another is the seeming plausibility of the assumption that the contributions made to the "feast" are individual dishes. This naturally suggests a "potluck dinner," valued for its "variety," and the rest follows. A third is the familiarity of epistemic arguments for political authority. Since at least Plato, a leading criterion of fitness for rule has been appropriate expertise; the current enthusiasm for epistemic defenses of democracy is only the latest incarnation of this approach. 18 That Aristotle should have advanced a similar argument may thus seem unsurprising. If I am correct, however, he was engaged in a profoundly different project, which deserves attention not only for the sake of getting Aristotle right, but also because it provides a way of defending certain forms of mass political activity that may ultimately prove more powerful than the epistemic arguments favored today.

The Conventional Account and Its Weaknesses
The core elements of the accepted account are the "potluck dinner" analogy, the claim that the multitude's contribution to political activity is fuller knowledge, and the supposition that this reflects decision making in the Athenian assembly. None of these elements is free from difficulty, however, as an examination of the text and its context reveals.
The essential feature of Aristotle's collectively provided feast is that it is expected to be better than one provided by a single man. Yet, as others have noted, this is hardly a common view of potluck dinners. 19 Indeed, in the Straussian tradition, it is precisely the improbability of this depiction (along with the reference to the "one man" with "many feet, and hands, and senses," which is regarded as "monstrous") that is taken to show that the whole argument is ironic. 20 Still, most commentators have proven willing to suspend disbelief on this point, since potluck dinners would seem to have a natural affinity with democracy. Everyone can participate; participation is on equal terms; contributions are small; and the results unpredictable. An analogy between dishes at a potluck and views in political debate is visible even at the linguistic level: in the words of Josiah Ober, benefits arise when different people bring different things "to the table." 21 Yet no matter how natural the "potluck dinner" analogy may seem today, there is no evidence that potluck dinners existed in ancient Greece. We have plenty of evidence of communal dining 22 but none of dinners where guests supplied dishes of their own choosing, and the idea fits badly with what we know of Greek domestic culture. Gender segregation suggests that any such dinner would have been all-male, yet for the same reason, attendees could not have cooked their own contributions; that would have been the task of a wife or slave. 23 So we must imagine a series of Greek men arriving for dinner, each bringing a dish prepared by a dependent-and then suppose that this could have struck Aristotle as analogous to the production of the citizenry's autonomous political decisions. If we take the point of the analogy to be "variety," further difficulties arise. Classical Greece was not twenty-first century America; its culinary traditions were not notably diverse. The Greek diet comprised mainly bread, olive oil, garlic, figs, and wild greens, with a little cheese, meat, or fish. Could a "finer" (kalliōn) meal really have been produced by the "potluck" process when the options were anyway so limited? 24 Moreover, on the subject of meat, sacrificing and roasting whole animals was normal procedure-a tricky thing to prepare in advance and take to a neighbor's in a dish. 25 Ancient standards of hospitality present another problem. The guest-host relationship was literally sacred, to Hestia, goddess of hearth and feast, and Zeus Xenios, protector of travelers, strangers, and guests. No self-respecting Greek householder could have proposed an evening's entertainment on the condition that guests help out with the food and drink. 26 Finally, whatever relation this passage bears to potluck dinners, another passage in the Politics suggests that Aristotle would have disapproved of them. At least, he disapproved of the Spartan requirement that all citizens, however poor, contribute to its public messes (syssitia); he preferred the Cretan system, in which such meals were provided from public funds, to which the poorest did not contribute. 27 These points raise doubts that a potluck dinner is the relevant paradigm. The text does nothing to dispel them, though that is not obvious from most English translations. Common renderings of the feast terms include "a feast [or 'feasts'] to which many contribute," 28 "a feast to which all contribute," 29 "a feast to which all the guests contribute," 30 "a banquet to which many contribute dishes," 31 and "potluck dinner" itself. 32 Yet Aristotle says merely that the dinners are symphorēta, "collectively provided" (sym, "collectively" or "jointly"; phērō, "bring" or "provide"). This term is rarely attested: the only examples we have of it in connection with dining are the two under discussion, which brings us to a dead end. 33 Nonetheless, it certainly does not imply that those who eat the dinner also help to provide it, or that the contributions made are individual dishes. All we can say is that Aristotle thinks a better meal will be supplied by more than one person, and that he is concerned with cost: the contrast he draws is with a meal "from a single purse" (ek mias dapanēs), not one cooked "by a single chef." 34 The next claim is that the multitude's contribution to political activity is knowledge. The main evidence for this appears in the second analogy, on aesthetic judgment. Most English translations feature the verb "understand" here, though "appreciate" is also seen: namely, "some appreciate one part, some another, and all together appreciate all." 35 Both suggest that Aristotle is concerned with some form of critical intellectual comprehension, but in the Greek there is actually no verb here at all. The relevant clause is highly truncated, reading simply "some a certain part, others another, and everyone the whole." An appropriate verb must be supplied by the translator. Many have opted for "understand" or "appreciate," but it need be neither.
The suggestion that Aristotle is interested in the production of betterinformed decisions is also doubtful. What is described as "better" in the Greek text are not decisions, but rather the agents who make them and the act of judging itself. Aristotle says that many people can be "better" (beltious) than a single man, and that they judge "better" (ameinon), but he never describes a decision in this way. This might seem ultimately to come to the same thing, 36 but a deeper issue is also at stake. The word better used here is 180 Political Theory 41 (2) not epistemic but moral: beltios, like the adjective spoudaios that denotes the "good man" in the preceding clause, is an explicitly ethical term, suggesting "decent" or "morally serious," not "well-informed" or "correct." 37 Consequently, there is nothing in the passage to suggest that Aristotle is arguing on epistemic lines. 38 The final element of the accepted account is the link to the Athenian assembly. The "potluck dinner" paradigm certainly looks plausible here: different people contribute different points of view, and every contribution helps to shape the final decision. But the "potluck dinner" reading is dubious, as we have seen: it remains to consider the political context more closely.
First, there is no reason to think that Aristotle is referring to classical Athens. He was acquainted with hundreds of poleis, seldom mentioned Athens directly (as Eugene Garver has observed, he cited Sparta considerably more often) 39 and is here making a general claim. 40 It might be assumed that Aristotle is defending Athens if he is defending democracy, but there were many democracies in the ancient Greek world, 41 and anyway, it is not clear that he is defending democracy, at least on his definition of it. 42 He is certainly defending the political authority of a multitude, but this could as easily be a reference to politeia (the first of the two types of rule by a multitude that he identifies) as to dēmokratia. 43 And since he supported politeia, but not dēmokratia, this would seem very probable; in which case classical Athens is irrelevant, for he certainly took it as an example of dēmokratia, indeed one of the most developed of its kind. 44 It is thus unlikely that Aristotle was thinking of debates in the Athenian assembly. Indeed, he was most likely not thinking of debates in an assembly at all. He cites three specific examples of political activity in connection with this argument: elections to important offices such as generalships or treasury positions; euthynai or "audits," the process of holding office-holders to account for their records; and judging cases in court. 45 All of these tasks were often performed by large numbers of ordinary citizens-assemblygoers, in the case of elections, citizen-judges (dikastai) or councilors in the case of audits, and dikastai again in the courts-so Aristotle did not need to look to assembly debates for an example of mass political activity. Indeed, the only suggestion that debates are relevant here is his use of the verb bouleuomai, usually translated "deliberate" and taken to suggest group discussion. 46 But this term could equally apply to decision making in elections, audits, and in court, so this does not get us very far. 47 This leads to a crucial point. As we have seen, the accepted account of Aristotle's argument affords an important role to speech as the medium through which diverse knowledge is shared; it is this that enables the best decision to emerge. This comes through especially clearly in Waldron's account: for him, the "doctrine of the wisdom of the multitude" forms "a kind of model or paradigm of our nature as speaking beings." 48 If the relevant political context were debate in an assembly, this might seem convincing. However, since Aristotle specifically cites elections, audits, and judging court cases, we need an interpretation that works in these situations, and here the accepted account falls short, since none of these activities-common throughout the ancient Greek world-ever, as far as we know, involved speech-making among the decision makers. In each case, decisions were made simply by voting, without prior discussion. In the case of elections, it is possible, though unlikely, that candidates or their proposers made a short speech before the vote, but it is highly unlikely that anyone else did so. 49 Canvassing occurred, if at all, outside the assembly; 50 at meetings, decisions were made speedily, usually by show of hands. 51 The situation in the courts is also clear. Aristotle reports that most legislators prohibited consultation (koinologountes) among judges prior to a verdict. 52 Aristotle thus cannot have supported the political authority of the multitude on the basis that it allowed diverse knowledge to be shared through public speech. Further confirmation of this is supplied by the text. Immediately after introducing this argument, Aristotle notes that it "would also apply to animals" (ta thēria). He sees this as a problem, and clarifies that the benefits of mass participation will arise only if the individuals involved have already achieved a certain level of virtue (aretē). 53 But the striking fact, easily overlooked, is that Aristotle thought that the argument, as it stood, could apply to animals. 54 Hence, it cannot depend on something that animals cannot do. Yet the possession of logos, "articulate reason" or "speech," was to Aristotle the crucial difference between humans and divine beings on the one hand, and all other living creatures on the other. 55 Under no circumstances could he have supposed that his argument would also apply to animals if speech played any part in it. It follows that speech cannot be a key feature of the political situations that he has in mind.
The accepted account is thus surprisingly insecure. We should return to the analogies on which it rests and try to establish an alternative.

The First Analogy Reexamined
The "feast" analogy appears twice in the Politics, denoted first by ta symphorēta deipna, "collectively provided dinners," and second, hestiasis symphorētos, "a collectively provided feast." 56 Given the context, these terms are presumably intended to be synonymous, yet the second formulation is 182 Political Theory 41 (2) considerably more informative than the first. Deipnon, "dinner" or "meal," was a common term in Greek, but hestiasis had a narrower meaning. It indicated a significant event, often privately financed, but held for communal political, diplomatic, or religious purposes. 57 In Athens, hestiasis had a particularly limited referent: it denoted the tribal dinner enjoyed during a major festival. 58 Providing the dinner was a significant public service (leitourgia), akin to sponsoring a chorus or team of athletes, and as with the offices of chorēgos and gymnasiarchon, each hestiatōr ("public host") was nominated from a list of 1200 individuals deemed wealthy enough to cover the event, and was supposed to be gratified by the honor of being asked. 59 It is immediately obvious that such an event could not have resembled a potluck dinner. There were ten tribes in Athens, each consisting of several thousand citizens. 60 Even if some stayed at home, inviting culinary contributions from attendees would have been a logistical nightmare, and rather than redounding to the honor of the organizer would have made him look contemptibly cheap. 61 Nonetheless, that Aristotle had a feast of this kind in mind is suggested by three points in the text. First is his mention of cost (dapanē), which implies that the event posed a significant financial burden and that funding, rather than culinary excellence, was the crucial factor in its provision. 62 Second is his use of the verb chorēgeō to describe the act of putting the dinner together. 63 This is derived from the noun chorēgos, "chorus-sponsor," and while it cannot prove that he had festival feasts in mind, it does make it seem likely. The third point concerns the adjective symphorētos, "collectively provided". As we have seen, this term is obscure, yet one possible interpretation suggests itself. We know that during the fourth century, the Athenians moved away from funding public services exclusively through single, very wealthy individuals and began to draw on panels of fifteen or more moderately wealthy individuals instead. Aristotle may be referring to a feast supplied in this way: that is, provided not by a single sponsor (nor by each attendee bringing a dish) but jointly funded by a committee.
Circumstantial evidence provides some support for this interpretation. Joint financing panels, or symmoriai, first appeared in Athens in 378/7 as a way of systematizing the payment of the war-tax, which fell only on the richest citizens. 64 From 358 a similar system was used to fund warships. 65 We have no direct evidence that festival feasts were ever provided this way, but Demosthenes floated the idea in 355, and a "great revival and reorganization" of festival ceremonies occurred twenty years later, right around the time that Aristotle returned to Athens and began teaching the material in the Politics. 66 Moreover, not only Athenian citizens but also wealthy foreign residents were liable to the performance of festival offices; Aristotle may even have participated himself. 67 The major obstacle to this reading is the suggestion of Mogens Hansen that festival offices were always performed by single men. 68 However, the evidence for this is not decisive. Hansen's source, a speech of Demosthenes from 350 or 348 which mentions the relevant nomination procedures, does, to be sure, list each office in the singular, which we might not expect if multiple nominations were the norm. 69 But since Greek lacks an indefinite article, we cannot tell whether "the" or "a" hestiatōr, chorēgos, and gymnasiarchon are under discussion here, and the appearance of the word trierarchon, "warship-sponsor," also in the singular a few lines later-despite the fact that this speech was made several years after that office was opened to joint funding-suggests that "a" hestiatōr must be a plausible reading. 70 In any case, even if festival feasts were sponsored singly down to 348, the reform might have occurred at a later date. On this evidence, the possibility that hestiaseis were jointly funded in the later fourth century cannot be ruled out.
If this interpretation is on the right lines, two points follow. First, contributions to the collectively provided feast will not have been individual dishes, but rather money (or conceivably resources in kind, such as livestock or grain), from which the entire feast was then provided. Just as modern gala events are organized by committees that take joint responsibility for raising and disbursing the entire budget (even if particular individuals take responsibility for specific items), so too a hestiasis symphorētos was a collective undertaking performed by a joint agent disposing of common resources. This is significantly different from a "potluck dinner," which is a collective undertaking performed by multiple single agents whose individual contributions remain distinct in the final outcome ("Paula's special pasta salad," and so on).
Second, though it is possible that a collectively provided feast might be preferred because of the greater variety of dishes, it is unlikely in this context. Festival feasts were not especially varied: they comprised chiefly meat and bread. 71 Additionally, collective funding for warships was introduced in order to build more ships, not to produce a more diverse fleet, and the goal in this case may have been similar. 72 A collectively provided feast could be "finer" simply because it was bigger. 73 More sponsors could mean more money spent overall, even if each gave considerably less than a single man under the previous system. In turn, this would mean more animals sacrificed and ultimately more meat to eat, the major attraction of such events. 74 It has been estimated that poorer households in Athens consumed 70-75 percent of their calorie intake in the form of grain; in these circumstances, it may have seemed obvious that what would make a meal "finer" was less "variety" than "abundance," and especially the abundance of meat. 75 The analogy itself supports this reading. A more familiar example of a collectively provided good would 184 Political Theory 41 (2) have been a warship; why should Aristotle have cited a feast instead? The relative novelty of the funding system may have been a factor, but more important may be the lack of correspondence between the money spent on a ship and the amount of ship produced. Extra resources would go to improving the ship in other ways: sheer size was not a relevant issue. 76 If size was the relevant issue for Aristotle, however, we can see why the "feast" analogy might have seemed apt.
The Virtue of the Multitude "Quantity," not "variety," thus lies at the heart of Aristotle's "feast" analogy and motivates his support for the political authority of the multitude. Many relatively small amounts of something-in the "feast" case, money-can add up to a large total amount, surpassing even what could be provided by a single man or a few with a very large stock. 77 What is contributed in the political context is stated in the next line. Every individual has a "portion" (morion) of aretē, "virtue" or "excellence," and phronēsis, "wisdom," "practical reason," or "prudence," or as Aristotle also puts it, "moral characteristics" (ta ēthē) and "thought" (tēn dianoian). These "portions" are then united when many men come together, just as their feet, hands, and senses are united, making them "better" (beltious) than one man or a few. 78 Aretē, in this context, presents little difficulty. It seems clear that it refers to the complex of moral virtues explored in Aristotle's ethical writings, such as courage, justice, moderation, liberality, and so on. 79 Aretē can also signify virtue in a general sense, including both moral and intellectual qualities, but when these aspects are distinguished, as here, it denotes the moral aspect alone. 80 Additionally, the only moral virtues to feature regularly in the Politics are courage, justice, and moderation, so we may take these to be central here. 81 Phronēsis is trickier to explicate. Though certainly an intellectual quality, 82 it differs from technē (skill), epistēmē (scientific knowledge), sophia (wisdom), and nous (intelligence) in also being implicated in ethical activity. Both moral aretē and phronēsis are necessary for right action: moral virtue "ensures the rightness of the end we aim at," phronēsis "the rightness of the means we adopt to gain that end." 83 At a deeper level, phronēsis is required to develop full moral virtue, while phronēsis without moral virtue is mere "cleverness." 84 Accordingly, when fully developed, moral virtue and phronēsis are inextricable, 85 but this connection is difficult to convey in English. "Wisdom" is not an ideal translation, because while that term can signify a faculty of the intellect, which phronēsis is, it can also signify "that which is known" (as in, e.g., "a book full of wisdom"), which phronēsis is not. Particular knowledge does play a role in phronēsis, but the terms Cammack 185 are not synonymous. 86 "Practical reason" is thus a better rendering, since it unambiguously suggests a faculty rather than an item or repository of knowledge, yet it too is not ideal, since the relation between phronēsis and logos, "reason," is incomplete in Aristotle's usage. Aristotle consistently describes human intellectual (and ethical) capacities as "involving reason" (meta logou), 87 but he also ascribes phronēsis (or at any rate a form of phronēsis) to certain animals, and animals do not possess logos. 88 Given this, it seems better to avoid strictly intellectual terms altogether and opt instead for "prudence" or "good sense," with "sensible" for the related adjective phronimos. 89 What a multitude contributes to political activity is thus each individual's "portion" of moral virtue and prudence, or every form of aretē (in the general sense) except the higher intellectual virtues. Exactly what Aristotle meant by "portion" is uncertain: perhaps a single element of virtue, such as "courage" or "justice," or perhaps a package of all combined. 90 The analogy drawn between virtues and body parts (moriōn) in the Eudemian Ethics supports the latter: 91 the correct reading of the reference to "one man" with "many feet, hands, and senses" would then be that just as each man puts both his feet, both his hands and each of his senses at the service of the group, so too he contributes his entire stock of justice, courage, moderation, and prudence. This question is relatively unimportant for our purposes, however; 92 the main thing is that the result is clear. What Aristotle is talking about is the coming together of individual aretē when groups act collectively.
We may focus initially on one key point. Aristotle is discussing the aggregation not of knowledge but of moral and intellectual capacities such as courage, justice, moderation, and good sense. He is concerned not with what any given agent knows, but rather what he is like: how brave, just, moderate, sensible, and so on. This is a crucial distinction, because on Aristotle's ethical theory, it is these qualities that determine what agents can do and how well they can do it. Plato argued that virtue was knowledge, but Aristotle explicitly rejected that view. 93 To Aristotle, aretē was action-centered: it was a "power" (dynamis) of "providing and preserving good things." 94 Specifically, it had the "twofold effect" of rendering its possessor "good" (or, we might say, a good specimen of its kind) and causing it to "perform its function well." 95 It followed that an agent with more aretē could outperform one with less, and this, I suggest, is the foundation of Aristotle's support for the political authority of the multitude.
The idea that a group of people acting together can outperform even highly capable individuals is relatively common in ancient Greek political thought. Xenophon's Hiero remarks that "nothing equals an organized body of men (syntetagmenoi), whether for protecting the property of friends or thwarting 186 Political Theory 41 (2) the plans of enemies," and one of Aesop's fables suggests the same. 96 Demosthenes also expressed this view. The fabulously wealthy Meidias had "mercenaries to look after him" and "witnesses to come running . . . when he asks," which was naturally "alarming to the rest of you as individuals, depending each upon his own resources." The solution was to "band yourselves together" (syllegesth' hymeis): so that when you find yourselves individually inferior (elattōn) to others, whether in wealth or in friends or in any other respect, you may together prove stronger (kreittous) than any one of your enemies and so check his insolence. 97 The obvious arena in which bands of individually inferior men can outperform their superiors is war, and this forms the crucial backdrop to Aristotle's argument. In archaic Greek, aretē meant above all military prowess, and Aristotle also used it in this sense, as in the suggestion that conquerors possess more aretē by definition and hence have a just claim to rule. 98 Yet war not only illustrated the link between greater numbers and greater power, it also indicated the greater efficacy achieved when a mass of men acts together. Though the figure of "one man" with "many feet, hands and senses" has been deemed "monstrous" by some, Aristotle may well have been thinking of a body of hoplites, whose strength lay precisely in their capacity to fight as one. 99 Significantly, hoplites also formed the citizen body in a politeia, which Aristotle defended on the grounds that military virtue, unlike other forms of virtue, could exist in large numbers. 100 Indeed, he even posited an etymological connection between politeia and polemos ("war") on this basis. 101 Even if a mass of men can defeat their superiors in war, however, what does this have to do with voting in elections, audits, and court cases, when more than brute force is required? First, even military virtue, on Aristotle's view, involved more than brute force. It included many elements of aretē, 102 particularly justice and courage. 103 More broadly, Aristotle characterized all practical virtues-that is, every moral virtue, plus prudence-as useful in both war and politics. 104 Indeed, these virtues sufficed for the "complete fulfillment of man's proper function," that is, the realization of a flourishing life in a polis. 105 It is also possible that the moral virtues played a larger part in Aristotle's conception of mass political activity. Not only were courage and justice, as key elements in military virtue, particularly associated with large numbers, but his confidence in the intellectual capacities of the multitude was typically low. 106 Yet this did not necessarily make such men bad or useless citizens. Indeed, Aristotle observed that "some people seem to form opinions better, and yet choose the wrong things from wickedness." 107 The same capacities that made a multitude useful in war could thus be expected to make it useful in politics. But a hard question follows. Even granting that more aretē might lead to greater success in these contexts, why should the "portions" of courage or justice or any other virtue of a mass of men be cumulative across a group? Why should a mass prove to be better than a single good man, when every individual falls short? Why should the group not prove as mediocre as its median member-or more vicious than any one of them?
This question has long puzzled readers who have seen the role of moral aretē in Aristotle's account. Newman, in 1887, spotted that Aristotle's principle would "justify the inference that the larger the gathering is, the greater its capacity will be," and responded: "Aristotle forgets that bad qualities will be thrown into the common stock no less than good . . . he forgets also the special liability of great gatherings of men to be mastered by feeling." 108 Susemihl and Hicks likewise supposed that crowd emotion would inhibit rather than support virtue. 109 Yet Aristotle does seem to have thought that aretē and crowd size were simply correlated. He likens aretē to wealth, strength, and weight, all of which increase arithmetically across groups. 110 He states that a mass can be "stronger," "richer," and "better" than a few, apparently all for the same reason; 111 his conviction that aretē is purely beneficial may also have persuaded him that it was summative. 112 He also explicitly believed that crowds are less likely to be swayed by emotion or other corrupting influences than smaller groups. At the second appearance of the "feast" analogy, he states that crowds make the best judges, both for the "feastrelated" reason and because they are more stable: "it is difficult to make a mass of men get angry and go wrong at the same time." 113 To be sure, this view directly contradicts modern notions of "mob justice," but Aristotle's presentation suggests that it was widely accepted among the Greeks. 114 There is certainly a puzzle here, but a final twist in Aristotle's argument may help to resolve it. Apparently, aretē can not only be aggregated in group contexts; it can also be amplified. Acting alone, Aristotle says, most men's injustice and folly will lead them to behave badly and make mistakes: 115 each is "immature in judgment." 116 But by "mingling with their betters" (mignumenoi tois beltoisi), the same men will become useful to the polis. 117 This suggests that acting collectively can lead to both the aggregation of aretē within groups and its amplification within individuals. If so, mass political activity could be expected to produce something more than the sum of its parts-just as Waldron, in fact, suggested. 118 In attributing this result to the dialectical quality of debate in the assembly, Waldron merely misdiagnosed the mechanism through which this phenomenon occurs.
One part of the puzzle is thus solved. If acting with others can be expected to increase the aretē of each individual, then the sum total of aretē across the 188 Political Theory 41 (2) group will obviously increase when it acts together. Moreover, if this action includes voting, as in elections, audits, and court cases, we can even see how the aggregation of aretē would occur: the votes are simply added together.
Yet the underlying mystery remains. Why should we become more courageous, just, moderate, or sensible when we act with others? Here we must return to the wider intellectual context. The idea that acting with or alongside others can strengthen individual virtue was another relatively common feature of Greek political thought. The effect could be gained in multiple ways: through moral support, 119 emulation, 120 rivalry or competition, 121 the desire to impress an audience, 122 or the fear of shame. 123 Each mechanism pointed in the same direction: the presence of others would cause individuals to amplify socially valued characteristics and repress those that are disapproved, leading groups to act better and do more than single men.
Aristotle was an innovator in several respects, but not this one. 124 He saw the value of rivalry and competition: "if all men vied with each other in moral nobility (pros ton kalon) and strove to perform the noblest deeds, the common welfare would be fully realized." 125 He recognized the importance of shame, especially for stimulating "civic courage," 126 and that the eyes of others were its "abode." 127 Most significantly, he believed that it was easier to practice aretē in groups than alone. This emerges most clearly in his discussion of friendship. 128 The life of virtuous activity ought to be supremely pleasant, "yet a solitary man has a hard life, for it is not easy to keep up continuous activity by oneself; it is easier to do so with the aid of and in relation to other people" (meth' heterōn kai pros allous). The life-work of the "good man" (ho spoudaios) will thus be more "continuous" if "practiced with friends"; moreover, his society will supply "a sort of training in goodness" for others. 129 This recalls the claim that men become useful by "mingling with their betters"; evidently, good models are important, and this is confirmed elsewhere. Friendship between "inferior people" is evil, for they take part in inferior pursuits, but friendship between the good is good, for they "become better by practising friendship and correcting each other's faults, as each takes the impress from the other of those traits in him that give him pleasure." 130 This suggests that the benefits of collective political activity will emerge only if the parties involved have already achieved a certain level of aretē-which is precisely the point Aristotle makes about the non-applicability of this argument to animals. 131 Notably, however, this condition is met in a politeia, where all citizens possess at least military virtue. In that context, acting "with the aid of others" could certainly be expected to increase aretē across the board. This interpretation may not be self-evident. Yet neither is it particularly far-fetched. Indeed, especially in relation to the positive effects of moral support and the desire to avoid shame, it is arguable that something like Aristotle's reasoning is accepted even today. Courage presents the most obvious example. Individuals are commonly braver in groups than alone: this is frequently observed in wars and revolutions, and the same dynamic can be imagined in the political situations that Aristotle has in mind. Recall that elections in ancient Greece were normally decided publicly, by show of hands: individuals might well have felt bolder in their choice-making if they could see that others shared their view. Or take the case of a general accused of treason: one-on-one, an ordinary citizen might have felt too awed to demand punishment, though he could do so as one of a crowd. 132 Justice may seem harder to interpret on these lines, given the modern skepticism of "mob justice" (which, as we saw above, however, Aristotle did not share). Yet we should remember that justice, for the ancient Greeks, was intrinsically otherregarding, and that judicial activity explicitly concerned what is due the polis as a whole; the presence of other members of the community might well have helped individual judges to keep that end in view. 133 Moderation, also, can be boosted by moral support, as anyone who has dieted knows, and its role in the use of wealth could plausibly be enhanced if the eyes of others encouraged self-restraint. 134 Finally, it is significant that Aristotle specifically defines phronēsis in the political sphere as the capacity to discern the common interest (to koinē sympheron). 135 It is surely easier to focus on what is common, rather than one's private interests, when acting as one of a crowd. Moreover, there may even be an internal connection between phronēsis and crowd size: at least, a larger decision-making group will by definition give a better indication of what the whole community takes the common interest to be.
The theme here is thus not the positive effects of diverse knowledge on decision making but the benefits of collective activity on individual and group aretē. Collective action by sufficiently virtuous individuals can boost each man's stock of justice, courage, moderation, and prudence, thus increasing the effectiveness of the entire group. As Aristotle notes, this argument does not support men of the multitude acting as generals or treasury officials, since those tasks must be performed by single men. 136 But they can certainly participate in any tasks open to collective authorship, such as electing men to office and judging them in court. 137 The only condition is that the aretē of the group reaches the necessary standard. A group of animals will never reach it (and, Aristotle asserts, some men are not so dissimilar from animals). Also, if the polis includes some man or men of exceptional virtue, the multitude will fall short in comparison (which is why Aristotle turns next to the discussion of ostracism and kingship). But if and when the aretē of a multitude exceeds that of other subsets within the polis, Aristotle accepts that it will have a claim to political authority.

190
Political Theory 41 (2) It may be asked how the "amplificatory effect" discussed here fits with the interpretation of the "feast" analogy given above. Surely the same dynamic cannot be seen in that case? Actually, it is not implausible that competition for honor might lead each sponsor to contribute more than he had originally anticipated. If so, the superiority of the collectively provided feast would result not only from the fact that the aggregated funds of a group can exceed those supplied by a single man. It would also reflect the increased size of each contribution made when men act together.

The Second Analogy Reexamined
We may now ask how this account fits with the second analogy, that concerning aesthetic judgment. Aristotle says that just as a multitude, when it comes together, becomes "one man with many feet and many hands and many senses," so too it unites its ethical qualities (ta ēthē) and thought (dianoia), and adds that "this is also why many people judge better of both musical productions and the works of the poets: for some a certain part (morion), others another, and everyone the whole." 138 The idea that a multitude can judge artistic productions better than a small number-for instance, the ten men appointed to judge dramatic competitions at the Dionysia-is certainly historically plausible. 139 Aristotle may have been thinking of Euripides, who won few prizes in his own day but was always a great favorite with audiences (and especially the poor), and by Aristotle's time was accepted as one of the great tragic poets alongside Aeschylus and Sophocles. 140 Or he may have had Aristophanes in mind, whose Clouds was placed third by the judges, despite the audience having "noisily demanded that it be put first on the list." 141 How do these cases support the account given so far?
The crux of the matter is how to supply the missing verb in the clause "some a certain part, some another, and everyone the whole." As we saw above, a common solution is "understand," although "appreciate" is also seen; a third possibility is "judge." 142 Of these, "judge" is preferable on syntactical grounds, since this verb appears in the previous clause, and Greek verbs are more commonly ellipsed than repeated. However, another approach is also possible. Arguably the most striking feature of the second analogy is the term morion, "portion" or "part." This term was used in the previous line to describe the "portion" of moral virtue and prudence supplied by each member of the multitude when the group acts together, but in the second analogy it has always been understood to have a different referent: the "parts" or "aspects" of the music or play judged by the multitude. That referent is implied whether the 192 Political Theory 41 (2) prudence, was entirely acceptable, even expected. 146 For another, it is not clear that a "division of labor" model adequately captures how multitudes judge either political actors or works of art. Certainly, different individuals will weigh different aspects of the judged object differently, but it does not follow that each responds only to a single aspect: more plausibly, each responds to the object as a whole, though in different ways. Certainly, this was assumed in Greek dramatic competitions, where there was no tradition of scoring acting, dancing, writing, costumes and so on separately. Rather, productions were judged against each other as wholes. 147 This reading confirms the interpretation of the "feast" analogy advanced above. If Aristotle was indeed thinking in terms of collective agency, it would logically have been impossible for him to draw a strict one-to-one correspondence between each man's contribution to a collective act and a specific "portion" of the task at hand. On the interpretation of the argument offered here, something is added when groups act collectively that cannot be traced back to any single individual, but is rather an emergent property of the mass. This is a very different conception of popular political activity from that implied in the usual account of this argument. On that view, the collectively produced outcome can be traced back to individual contributions in a straightforward way: one spectator understands one part of a performance, another another; hence together they understand the whole. Nothing is incorporated into the final outcome that is not already possessed by each contributor, conceived separately. But when Aristotle states the argument, he does not suggest that what happens when a multitude comes together is that each person is individually responsible for a discrete part of the collectively produced outcome. Rather, he says that everyone contributes his "portion" of ethical virtue and prudence to the collective agent, the "one man" with many hands, feet, and senses, by whom the outcome is ultimately produced, and this is a profoundly different idea. 148 The easiest way to illustrate this difference is to revert to the contrast between a potluck dinner and a collectively provided feast. In the potluck, each part of the dinner-each dish-can be traced back to the individual who originally contributed it: the meal is the sum of numerous distinct acts. In the collectively provided feast, however, the whole feast is the result of a single collective act, enabled by numerous contributions to a single fund. It would be impossible to assign any "part" of this feast to an individual sponsor: rather, the whole panel is collectively responsible for the whole. Arguably, Aristotle has a similar conception of political activity. At a fundamental level, he does not imagine that the activity of the polis can be broken down into several distinct parts, attributable to each single citizen within the community. Rather, the whole polis is ultimately implicated in the flourishing of the whole, because the aretē of each part is involved in the successful self-development of the rest. 149 Cammack 193

Conclusion
On the usual account of the argument explored here, Aristotle thought that a multitude can make better decisions by pooling its knowledge than individuals or small groups can make alone. Mass political activity thus resembled a potluck dinner or an audience judging a play: a greater variety of contributions could produce a better result. I have suggested that Aristotle's support for the political authority of some multitudes rested not on their access to diverse knowledge but rather on his belief that the ethical qualities and good sense of single men can be aggregated and even amplified when they act together. A multitude thus has a claim to rule whenever its aretē exceeds that of other groups within the polis, so long as its acts can be jointly authored.
This account detaches Aristotle from those who would wish to cite him in support of democracy on epistemic grounds. However, it need not diminish his significance for democrats today. I will close by suggesting where this significance lies.
Let us return to Waldron's suggestion that Aristotle's argument "stands as a kind of model or paradigm of our nature as speaking beings." As we have seen, this must be false, since speech does not feature in Aristotle's examples of political activity, nor does it fit his contention that his argument would also apply to animals. Yet there is deeper misconception at work in Waldron's interpretation, which takes us to the very heart of Aristotle's political thought. The problem lies in the supposition that Aristotle regarded logos, "articulate reason" or "speech," as the defining mark of man's nature as a "political animal." 150 He certainly believed that logos made man more "political" than any other animal, but it cannot define political animals as such, because while other animals-bees, ants, and cranes, for example-are also "political" on Aristotle's account, man alone possesses logos. 151 What defined a political animal was in fact not the capacity to speak, but being engaged in "some one common activity" (hen ti kai koinon . . . to ergon). 152 In other words, acting collectively, not reasoning, was to Aristotle the fundamental political activity.
I suggest that this perspective may prove salutary for contemporary political theory. The current enthusiasm for epistemic defenses of democratic authority faces a potential problem: what if the decisions of the multitude turn out not to be "better," on epistemic grounds, after all? What if mistakes are made that could have been avoided under "expert" guidance? Does that mean that ordinary people should be excluded from political authority? To Plato, the obvious answer to this question was yes, but democrats will find that less appealing. There are ways of getting around this problem in its own terms, 153 but on Aristotle's view, this is simply the wrong way to think about political activity. The great interest of Aristotle's account is that he does not 194 Political Theory 41 (2) base political authority on the contingent possession of potentially useful information but rather on a range of ethical and intellectual capacities that, given the right training and external conditions, can be widely shared. 154 To be sure, the benefits of collective political action will emerge only if most members of a community actually value such qualities as courage, justice, moderation, and good sense; if cowardice, viciousness, intemperance, and folly are widely tolerated, this cannot happen. However, this merely underscores the significance of a point made immediately prior to this argument in the Politics, which is that true poleis are distinguished from mere "alliances" precisely by their attention to aretē. 155 It does not follow that those who participate in political authority on this basis must be a small group; to the contrary, Aristotle's conception of what is politically valuable might well prove more hospitable to democratic aspirations today than the currently admired epistemic approach.
The question remains what this might mean in practice, and here we can reconsider the historical context that Aristotle had in mind. As observed earlier, there is no reason to think that his argument applied particularly to classical Athens. He does, however, link it clearly to Athens in an earlier era, that of Solon in the early sixth century. Solon is cited approvingly as the lawgiver who gave the Athenian plēthos the right to take part in elections and audits, while his judicial reforms, which gave ordinary Athenians control over the administration of justice, are discussed elsewhere in a positive light. 156 This ought to be very striking to modern democrats. Aristotle evidently believed that his argument supported the kind of mass political participation introduced by Solon. This system was certainly less democratic than classical Athens: for example, there was no general right to speak in the assembly. Yet it was still considerably more democratic than any modern democracy, judged by the extent of the powers enjoyed by ordinary citizens. Then as now, citizens had the power to elect their leaders, but the Greek practice of routinely holding those leaders to account has no equivalent today, and the fact that supreme judicial power was wielded by large bodies of ordinary citizens without any legal training is also unparalleled. Most people today would probably agree that hearing court cases, pronouncing final verdicts, and setting appropriate penalties requires special knowledge and expertise, and thus cannot be done satisfactorily by a mass of ordinary citizens. This confident sense of the limits of mass political activity was shared by Plato, and is readily justified by an epistemic account of political competence. It was not shared by Aristotle, however, and the failure of some of his recent readers to see this may reveal not only a deficiency in the contemporary interpretation of Aristotle but in the contemporary conception of democracy as well.