Hostname: page-component-7b7567c8c-kwt9d Total loading time: 0 Render date: 2026-11-13T20:09:06.874Z Has data issue: false hasContentIssue false
  • English
  • Français

Aristotelian Realism: Political Friendship and the Problem of Stability

Published online by Cambridge University Press:  11 September 2019

Eric Cheng
Rights & Permissions [Opens in a new window]

Abstract

This paper argues that the pursuit of stability is the primary concern of Aristotle's understanding of political friendship. Specifically, I argue that Aristotle chooses to understand political friendship to be a “special sort” of utility/advantage friendship, applicable to multiple regime types of varying degrees of (in)equality, because he fears what might happen when citizens in any polity develop mutual animosity. Turning to the contemporary liberal democratic context, I note that Aristotle provides us with a strong positive argument for why we ought to take political friendship seriously. However, I stipulate that contemporary liberal democracies present obstacles to the realization of classical political friendship. I thereby conclude by suggesting that citizens can potentially be political friends when they understand politics and their social relations through the “metaphor” of political friendship.

Information

Type
Research Article
Information
The Review of Politics , Volume 81 , Issue 4 , Fall 2019, pp. 549 - 571
Copyright
Copyright © University of Notre Dame 2019 

In this article, I argue that the pursuit of stability is the central concern of Aristotle's understanding of political friendship. Aristotle understands political friendship to be a solution to the problem of stability—the problem of how to prevent citizens who adhere to a plurality of interests and conceptions of justice from clashing in manners that threaten the survival of the regime, if not the existence of the polity itself. Specifically, I demonstrate that Aristotle understands political friendship to be a “special” sort of utility/advantage friendship. Political friends share a mutual affection that is rooted in a common recognition of political virtue; they recognize that they are each (1) committed to the common advantage of the regime, and (2) able and willing to do what it takes to further that common advantage. The concept of political friendship can thus be used to describe the healthy social bonds shared by citizens in multiple “correct” regime types of varying degrees of (in)equality—kingship, aristocracy, and polity.

Two implications follow from this. First, even though it is rooted in advantage, political friendship cannot be understood in terms of mere marketplace exchange; the affection shared by political friends is deeper than that shared by ordinary friends of utility (e.g., “buyers” and “sellers”), and the demands placed on political friends more vigorous. Second, political friendship cannot be understood as synonymous with virtue friendship—the sort of friendship that Aristotle deems necessary for the pursuit of the good life and the fulfillment of man's telos. In fact, in some regimes, the development of political virtue decreases the likelihood that citizens are capable of participating in virtue friendships. I argue that Aristotle understands political friendship in this manner—as a sort of compromise concept that is not necessarily conducive to human flourishing—because he fears what might happen when citizens in any regime view one another with animosity.

I conclude by suggesting that Aristotle provides us with a strong positive argument for why we should appreciate the importance of political friendship for the contemporary liberal democratic context. However, I also stipulate that contemporary liberal democracies present obstacles to the practical realization of Aristotle's understanding of political friendship. First, it is unlikely that citizens in liberal democracies would be able to recognize political virtue in one another, for these societies tend to be large; citizens in liberal democracies tend to be strangers. Second, liberal democracy imposes a normative constraint, that the stability secured by political friendship must not undermine the pursuit of liberal democratic justice; Aristotle's understanding of political friendship is not necessarily democratic—let alone liberal. What this means is that unless it is shared among political equals, political friendship might actually destabilize society. When political friendship is conceived as being shared among unequals, it serves as an unduly conservative source of social cohesion and, as a result, may prompt the proponents of equality to purposefully destabilize society in the name of justice. Therefore, I offer a “metaphoric” notion of political friendship that might allow citizens to be political friends despite these difficulties: specifically, citizens can potentially be political friends on the basis of the cognitive lenses or conceptual metaphors through which they consciously and unconsciously understand politics and their social relations. If we accept that the use of metaphors is inevitable insofar as we find ourselves in linguistic communities and that metaphors have the power to structure our experiences, emotional responses, and social practices, then we can say that citizens who are strangers can be political friends if they interpret the world through the assimilated metaphor of political friendship.

In what follows, I first briefly review the basics of Aristotle's understanding of friendship. Second, I survey a range of scholarly interpretations of Aristotle's understanding of political friendship. Third, I show that Aristotle understands political friendship to be a special sort of utility friendship and argue that Aristotle develops the concept as he does out of a concern for the problem of stability. Fourth, I suggest that citizens in contemporary liberal democratic societies can potentially be political friends when they understand politics and their social relations through the metaphor of political friendship.

Foundations of Aristotle's Philia

In the Nicomachean Ethics (NE),Footnote 1 Aristotle characterizes a variety of human relationships as types of “friendship” (philia). First, Aristotle describes “equal” friendships where the friends involved share the same underlying sort(s) of affection and function(s). They love (philein) one another for the same reason(s) and to the same degree. There exist three primary sorts of equal friendships: pleasure, utility, and virtue friendships (NE 1155b20–30). Friends of utility are friends because they derive advantage from one another, friends of pleasure because they derive pleasure from one another, and friends of virtue because they love one another for their intrinsic goodness (NE 1156b10). Second, Aristotle describes “unequal” friendships, such as those between fathers and sons and between husbands and wives, where the friends involved have different functions, exhibit different sorts of affection, and merit different degrees of affection (NE 1158.b15–25).

Aristotle's notion of philia thereby encapsulates a variety of relationships that we might not be inclined to describe as friendships today; the love implicit in philia refers to the affection one feels towards someone with whom he/she shares, could share, or would like to share in any of a range of reciprocal relationships, from the superficial to the profound. What links these types of friendship together, however, is the presence of some sort of characteristic or lasting sense of reciprocity and mutual friendly affection among the friends involved (NE 1155b30). Friends do not merely have good will towards one another; they also act on that mutual friendly affection in a manner that maintains equality, properly understood as proportionality according to merit (NE 1166b30–1167a10). So, while equal friends must exhibit the same degrees and sorts of friendly affection because they share the same function (within the context of the friendship), unequal friends must exhibit different degrees and sorts of friendly affection because their functions differ. In the context of friendship, however, considerations of “a certain quantity” trump those of merit: if the differences of wealth, virtue, and/or vice among unequal friends become excessive, then those people can no longer be friends (NE 1159a1).

Tempting Interpretations of Political Friendship

At the outset of NE, Aristotle declares that friendship is the greatest of “external goods,” but it soon becomes apparent that only virtue friendship truly qualifies. The happy life of human flourishing (eudaimonia), according to Aristotle, is a self-sufficient life: it is “in itself … choiceworthy and in need of nothing” (NE 1097b16). The happy person must certainly be virtuous. However, even he must also have certain goods that are external to the self. First, he must have instrumental external goods (e.g., money, power) in order to survive and perform many actions. Second, he must have intrinsic external goods, without which supreme happiness is “disfigured”—even if instrumental goods are somehow unnecessary (NE 1099b1–5).

Why can only virtue friendship fully serve as the greatest of external goods? First, as an instrumental good, virtue friendship is more reliable than utility and pleasure friendships because it is grounded in the intrinsic goodness and durability of character, not in fleeting, incidental qualities that change and vary from person to person. Second, Aristotle considers friends external goods of the intrinsic sort, not just the instrumental sort. Even if we can each achieve material self-sufficiency without the aid of others, we still need friends because “we do not mean by self-sufficient what suffices for someone by himself, living a solitary life, but what is sufficient also with respect to parents, offspring, a wife, and, in general, one's friends and fellow citizens, since by nature a human being is political” (NE 1097b9–13). Happiness thus involves loving others as extensions of oneself (as “other selves”) (NE 1170b8), and this is only possible when the friends involved are good in themselves and deeply familiar with one another. Therefore, the sort of friendship Aristotle has in mind when describing friendship as the greatest external good is the sort that only virtuous people, capable of loving others as “other selves,” can access: virtue friendship.Footnote 2

It is therefore tempting to conclude that political friendship (politikē philia) must be virtue friendship: if man is a social-political being, and if virtue friendship is required by the good life, then political friendship seems to consist of the joint pursuit of the good life. Alexander Nehamas argues that this might be the case even if only an excellent few share in such friendship. According to him, politics was regarded in the ancient world as a “sphere of higher human activity,” and “the connection between virtue and public accomplishment … was close enough that the relationships of the virtuous were played out, for the most part, in the open.”Footnote 3 So, friendship furthers a healthy politics insofar as it is present among virtuous statesmen like Pericles. Likewise, although Ann Ward notes that political friendship is “initially a friendship of utility,” she insists that “the reduction of political friendship to utility friendship … causes an isolated individualism, if not passionate class warfare … [and] over time, tear[s] it [the city] apart.”Footnote 4 She goes on to argue that in order for a political community to sustain itself, its citizens must be aware that political friendship aims at something higher than mutual material advantage; in the case of timocracy (the regime most conducive to friendship), political friendship can aim at “nobility … the universal good.”Footnote 5

Similarly, even though John von Heyking distinguishes between virtue friendship and political friendship, he argues that the two are analogous: political friendship is the “civic version” of virtue friendship. Just as virtue friendship consists of the joint perception of the good, political friendship consists of the sharing of stories in whose name political action is conducted. So, not unlike how virtue friendship is a “condition of inferior friendships,”Footnote 6 political friendship is a precondition of politics: “political friendship resides in the backdrop of ‘regular’ political deliberations and activities…. Ambition counteracting ambition is constrained by agreement on constitutional fundamentals, expressed as a social friendship that prevents such conflict from degenerating into fratricidal war.”Footnote 7

In contrast, Sibyl Schwarzenbach and John Cooper argue that political friendship replicates or extends the social bonds of the family. However, not unlike Nehamas, Ward, and von Heyking, Schwarzenbach and Cooper argue that political friendship can only be sustained if it replicates or extends the social bonds of the family precisely because the survival of familial bonds demands an attention to virtue. According to Schwarzenbach, in conditions of political friendship, legislators realize an “ethical reproduction” of appropriate relations and habits among citizens reminiscent of that which a mother seeks to realize in her child.Footnote 8 According to Cooper, fellow citizens exhibit a concern for one another's character reminiscent of that which is exhibited by fellow members of a family. What follows is an “extension to a whole city of the kinds of psychological bonds that tie together a family.”Footnote 9

I argue instead that political friendship is a form of utility friendship. However, I also insist that political friendship is a special type of utility friendship which imposes greater demands on the parties involved than other forms of utility friendship. Political friendship is structurally unique; it is neither straightforwardly an example of utility friendship nor a utility friendship in name only. My argument thus has more in common with Jill Frank's than with Kazutara Inamura's, Douglas Den Uyl and Douglas Rasmussen's, and Bernard Yack's.Footnote 10 Inamura and Den Uyl and Rasmussen understand political friendship to be a straightforward example of utility friendship, not so different from that which characterizes marketplace transactions. Meanwhile, even though he does not characterize political friendships as de facto marketplace transactions, Yack argues that political friendship must not be confused with virtue friendship or even with “moral” utility friendships (to be discussed). With that, he insists that the difference between political friendship and marketplace utility friendship is one of degree rather than of kind. Frank, by contrast, argues that Aristotle understands political friendship to be a mix of virtue friendship and utility friendship. It is modeled on utility friendship yet involves the practice of virtue—a mixed good concerned with acting and judging well that maintains a view to the common good. Political friendship involves virtue to the extent that good judgment and moderation are required to sustain the fundamental terms of cooperation. This means that political friendship demands more virtue than does utility friendship; political friendship concerns the polity's enduring constitution and way of life, whereas utility friendship concerns whatever represents the immediate, mutual advantage(s) of the parties involved.Footnote 11

My interpretation of Aristotle's understanding of political friendship, however, diverges from Frank's on several key points. As we shall see, Frank and I disagree about whether political friendship demands actual familiarity among the friends involved. Frank understands political friendship to be a set of practices that facilitates cooperation by accommodating the means for resolving conflict.Footnote 12 In contrast, I insist that Aristotle's understanding of political friendship must remain a sort of social unity, grounded in a kind of mutual recognition among citizens, even if it is definitively modeled on utility friendship. Likewise, we advance different accounts of why Aristotle is moved to develop such a conception of political friendship. Frank's interpretation can be described as moral or ethical: Aristotle engages in an endoxic investigation for the purposes of discerning the character or attitude of friendliness required to sustain the individual and shared practices of property, justice, law, and so forth that make a good polity possible—one which consists of a unity of persons who differ in kind.Footnote 13 My interpretation leans instead towards the political: political friendship is a compromise concept that Aristotle develops to render plausible the possibility that fellow citizens in multiple sorts of cities can indeed be friends with one another. That is, I emphasize Aristotle's concerns about social-political instability. Therefore, on my account, Aristotle's understanding of political friendship expresses what can potentially be called an Aristotelian realism.Footnote 14

A Special Sort of Utility Friendship in Pursuit of Political Stability

For the Sake of Advantage, or for the Sake of Living Well?

Those who have argued most forcefully that political friendship is a type of utility friendship, such as Yack, Inamura, and Den Uyl and Rasmussen, point to Aristotle's characterization of political friendship in his earlier work on ethics, the Eudemian Ethics (EE): simply put, political friendship is utility friendship. Even though they would have come together in any case for the sake of living together, men come together because they cannot achieve self-sufficiency by themselves (EE 1242a1–9). Political friendship consists of an exchange of goods among equals; these friendships dissolve when the friends involved, be they cities or citizens, no longer find one another useful (EE 1242b25). So, political friends treat one another in a similar manner to how equal buyers and sellers treat one another. In theory, political friendships can proceed in one of two ways: (1) as legal friendships via contract, where money/payment, the measure of equality, clearly indicates when the friendship in question has ended; and (2) as moral friendships, where the friends involved trust one another to return favors. Moral friendships are rather similar to virtue friendships. In practice, however, moral friendships are bound to end acrimoniously: they are “unnatural; for friendships based on utility are different; but these wish to have both together, associating together really for the sake of utility, but representing their friendship as moral, like that of good men; pretending to trust one another they make out their friendship to be not merely legal” (EE 1242b38–1243a2).

This is a striking passage, for Aristotle skirts around the notion that man is a social-political being in order to emphasize our individualistic impulses, lack of trust, and seeming preoccupation with the economic. The point here is not so much that people are incurably selfish and untrustworthy as that the capacity to trust one another and the inclination to pursue the noble (and not merely the economic) are only plausible after people have secured some requisite level of “necessary” goods. Stated alternatively, political friendship, a form of utility friendship, is a means to realize the material condition required to access or pursue virtue friendship. The propensity to mistakenly characterize political friendship as “moral” friendship—as something closer to virtue friendship—reflects a desire for virtue friendship.Footnote 15 Aristotle remarks,

Civic friendship, then, looks to the agreement and thing [e.g., am I receiving a return which befits my tremendous sacrifice and/or recognizes the value of my provided service?], moral friendship to the choice…. Men start … by proposing to be moral friends, i.e., friends through excellence; but as soon as some private interest arises, they show clearly they were not so. For the multitude aim at the noble only when they have plenty of everything else; at noble friendship similarly. (EE 1243a33–1243b1, emphasis added)

NE lacks such definitive claims that political friendship is a sort of utility friendship. This is because EE and NE advance different understandings of political friendship. In NE, Aristotle does not align political friendship nearly as closely with marketplace utility friendships. Still, as Yack, Inamura, and Den Uyl and Rasmussen eagerly note, NE provides plenty of evidence to suggest that political friendship remains a sort of utility friendship. First, a matter of numbers: it is not possible to share in many virtue friendships. The virtue required to participate in virtue friendships is rare; few people are capable of viewing others as “other selves.” Moreover, the formation of virtue friendship requires time and intimacy—that people develop love, trust, and the “habits formed by living together” (sunētheia) (NE 1156b25, 1158a11–21). So, if political friendship were a form of virtue friendship, then the pursuit of political friendship anywhere other than in the smallest of social contexts would be a wholly utopian endeavor. Not only is it impossible for any city to consist entirely of excellent people; according to Aristotle, a city, by definition, cannot be so small as to lack full material self-sufficiency (Pol. 1252b27). Virtue friendship, as a result, cannot describe the healthy social bonds shared by a city's forty thousand inhabitants.

Second, Aristotle maintains that political friendship is oriented towards the pursuit of the advantageous. When a city is characterized by political friendship, its citizens are like-minded (homonoia) (NE 1155a25–8). They do not simply agree on just anything; they agree on how practical issues of common concern must be approached, and they act on those shared judgments—for instance, whether a political office should be created, or whether an alliance should be formed (NE 1167a20–30). When like-mindedness is absent, factional conflict ensues. For instance, Aristotle notes that the most common regime types of his day—democracy and oligarchy—are prone to factional conflict, in part because their citizens do not agree on what constitutes justice (Pol. 1301b30–1302a5). In democracies, the poor rule on the basis of numerical equality, while those denied a share in office (the wealthy) demand that the regime be ruled on the basis of meritocratic equality. Meanwhile, in oligarchies, the opposite holds: the wealthy rule on the basis of meritocratic equality, while those shut out of office (the poor) demand that the regime be ruled on the basis of numerical equality (Pol. 1301b30–1302a5).

Similarly, Aristotle construes the political community to be a composite of smaller communities that is formed to further the partial advantages of those smaller communities. For example, the political community encompasses communities of sailors and communities of soldiers, so it “seems to come together from the onset, and to continue to exist, for the sake of what is advantageous, albeit that of life as a whole rather than merely that of the present” (NE 1160a11).

This depiction of political friendship as a sort of utility or advantage friendship seems to be in tension with Aristotle's core thesis that man is a social-political being who would choose to live with others for the sake of living well, even if he could maintain material self-sufficiency without the help of others (Pol. 1252b30). Indeed, Cooper emphasizes Politics III.9 instead of NE.Footnote 16 There, Aristotle declares that the city exists for the sake of living well: “Living well, then, is the end of the city…. A city is a partnership of families and villages in a complete and self-sufficient life. The political partnership must be regarded, therefore, as being for the sake of noble action, not for the sake of living together” (Pol. 1281a2–3). For a city to be a city, rulers must pay attention to political virtue and vice; a city's inhabitants must care about virtue (Pol. 1280b5–8). Accordingly, Aristotle rejects several conceptions of the city that would align it with utility or advantage. The city is not a mere alliance formed for the sake of mutual protection, exchange, or use. If a city were formed for those purposes, then the act of reaching trade agreements or military alliances would serve to merge two cities into one—even if those cities do not share any common offices, and even if the inhabitants of either city do not care about the character of those in the other. Nor is a city simply the location where people happen to conduct business. Certainly, business transactions do occur in a city. However, business concerns necessary things—living rather than living well (Pol. 1280b30–34). So, a series of ongoing market exchanges, regulated by a common set of laws ensuring just transactions, does not make a city (Pol. 1280b20).

We can interpret this argument in a few ways. It can mean that citizens must be educated so as to be virtuous. Alternatively, it can mean that citizens must care about one another's character, or it can mean that a city is only viable when it has a virtuous citizenry. Regardless, what is clear is that Aristotle here binds political partnership to a concern for virtue and to living well, not to merely securing advantage. The Aristotle of Politics seems to contradict the Aristotle of NE: the former aligns political partnership with virtue, living well, and the natural sociality of man; the latter, with advantage.

A Special Sort of Utility Friendship

I argue, however, that closer consideration reveals that the characterization of political friendship in NE and in Politics are closer than they seem prima facie. First, according to Aristotle, even though the city is not wholly concerned with what is advantageous, its concern with advantage is in fact an expression of man's social-political nature. Man is a social-political being because of his unique capacity for speech, and speech “serves to reveal the advantageous and the harmful, and hence also the just and the unjust … and partnership in these things is what makes a household and a city” (Pol. 1253a1–19, emphasis added). Second, in both NE and Politics, the political community can only be understood within a context of difference; the people involved are not identical. We have already noted that in NE Aristotle construes the political community to be a composite of smaller communities. Likewise, in Politics Aristotle criticizes Socrates for conflating a city with a unity that can be understood as analogous to the household or, even worse, to an individual person. The city must rather be understood as a sort of multitude of households and villages where human beings—different in kind—live together (Pol. 1252b10–20, 1261a15–30). Indeed, according to Aristotle, a city risks ruin should it be overly unified: “For a household is more self-sufficient than one person, and a city than a household; and a city tends to come into being when the partnership formed by a multitude is self-sufficient” (Pol. 1261b10–14).

Third, like-mindedness demands a meaningful degree of virtue. Aristotle implies this when he maintains that the virtuous can be like-minded in the manner that political friends are like-minded, whereas the base cannot. The base cannot be like-minded because they only look to maximize their own benefit; they obstruct the pursuits of their fellow citizens, and they refuse to perform their share of public service. Not only do the base neglect the commons; they also instigate factional conflict, for they shun just action and seek instead to compel others to act in various ways via force (NE 1167b10–16). Therefore, citizens can only be like-minded when they are capable of acting justly towards one another and when they understand that their respective individual advantages are constitutive of and constituted by the common advantage of the polity.

What this all suggests is that Aristotle understands political friendship to be a special sort of utility friendship with its own distinguishing features. First, political friendship requires political virtue: the capacity to recognize that one's individual advantage is bound up with the common advantage of the polity, and, with that (as we shall see), the disposition to further the common advantage in a manner befitting one's relative status within the particular political friendship in question. Political virtue is not the noble virtue required by virtue friendship—political friendship must be shared by citizens who are dissimilar and who are not necessarily capable of participating in virtue friendship—but it is some sort of virtue nonetheless.

Certainly, it is important to acknowledge that some degree of virtue is required to participate in utility friendships generally. After all, market exchanges too cannot sustain themselves if their participants are so deprived of virtue that they will attempt to renege at the earliest opportunity. However, the virtue demanded by Aristotle's understanding of political friendship is more vigorous than that demanded by other sorts of utility friendship. We can attribute this to the fact that political friendship concerns practical issues with long-term consequences like the survival of the polity and the way in which citizens should govern and be governed—not less significant and/or short-term issues like the exchange of economic goods. However, we can also attribute it to the fact that ordinary utility friendships are, in Aristotle's view, often inadequate even as guarantors of successful economic exchange, let alone the survival of the polity: “those who use each other with a view to some benefit always want more and suppose they obtain less than what is proper. And so they blame the other because they do not obtain as much as they want and think they merit, and those who perform the benefactions are not able to supply as much as the recipients want” (NE 1162b16–21).

Indeed, there is evidence that suggests that in ancient Greece, the introduction of coinage to decrease the prevalence of bartering and to minimize the transaction costs of marketplace utility friendships actually aggravated the personal distrust and antagonisms that existed between buyers and sellers. This is because the introduction of coinage generated information asymmetries. Given that buyers could only assess the quality of sellers’ goods after the transactions in question had already been conducted, buyers often suspected that sellers were trying to swindle them. As a result, economic transactions frequently devolved into conflict. When buyers and sellers did manage to reach agreements, they usually did so at the end of processes where they “negotiated … not … [only] by exchanging numbers … but by using gestures and language, both cunning and conventional, to cap and forestall each other.”Footnote 17

Second, according to Aristotle, political friends need not and cannot possibly love one another as “other selves,” but they must be friends with one another on the basis of their political virtue, regardless of their (necessary) differences. That is, political friends cannot simply be politically virtuous; they must also recognize in one another political virtue. Only so can they sustain both a commitment to the long-term common advantage and their respective proportional statuses within the friendship, for only so can they have a sense that others are also committed to the common advantage.

Now, many scholars who interpret Aristotle's understanding of political friendship to be a sort of utility friendship challenge the notion that citizens can only be political friends when they recognize one another as politically virtuous. According to Frank, Danielle Allen, and Çigdem Çidam, Aristotle's understanding of political friendship is more akin to a virtue or to a set of practices that enables citizens to more ably resolve conflict and act “as if they were friends.”Footnote 18 However, if this interpretation were to hold, then political friendship would be seemingly exempt from Aristotle's basic definition of friendship—the fact that friends have good will towards one another and act on that mutual friendly affection (NE 1166b30–1167a10). Moreover, insofar as political friendship is a sort of utility friendship whereby friends come together to pursue that which is advantageous for themselves individually, political friends are not “on account of their [complete] virtue … eager to benefit each other … surpass-ingly … [as a matter of] obtain[ing] what he aims at” (NE 1162b10). So, even though they might not be quite as inclined to hurl accusations and to blame one another as other sorts of friends of utility might be, political friends are still concerned about whether their fellow citizens are indeed doing what they ought to do to further the long-term common advantage of the polity. In short, political friends can only be political friends when each can verify that the others are not free riders. This helps to explain why Aristotle insists that even though it is possible for citizens to have many political friends (NE 1171a18), there is a limit to how many political friends one can have, beyond which one's political friendships fail to be conducive to the survival of the political community: “For a city could not come into being from ten human beings, yet when there are ten times ten thousand, it is no longer a city either” (NE 1170b31–2).

This raises the question whether citizens who are strangers can be political friends. Aristotle is ambiguous on this point. However, given that Aristotle himself notes that a city's population cannot exceed ten times ten thousand, it is unlikely that he intends for his understanding of political friendship to require that all citizens know one another personally. How, then, is it possible for citizens who are strangers to verify whether they are indeed doing what they ought to do to further the long-term common advantage?

One possibility is that political friendship can be realized through a series of networks, not unlike those present in democratic Athens as described by Josiah Ober. On Ober's account, Athens was organized into a series of participatory districts (demes), united by a city-wide council (a “master network of local networks”). As a result, Athens had the feel of a face-to-face society, even though it was never literally a face-to-face society: “virtually every Athenian had access to a network of personal contacts…. Every Athenian was connected to every other Athenian by only one or two ‘degrees of separation.’”Footnote 19 Perhaps, when such a series of intertwining networks exists, citizens who are strangers can use those networks to recognize in one another political virtue and a commitment to further the long-term common advantage—just as Athenians were able to use such networks to draw on the collective resources and knowledge of Athens as a whole.

Putting this question aside, what is clear is that political friendship cannot be understood, as can ordinary sorts of utility friendship, in terms of mere market exchange. The friendly affection that grounds political friendship is deeper than that which grounds other sorts of utility friendship, and political friends must be more virtuous than other friends of utility. As a result, political friendships will likely be more stable than other sorts of utility friendship. We can therefore understand political friendship to be a sort of utility friendship that can facilitate the formation of other sorts of marketplace utility friendships. This is not to say that political friendship is a precondition of marketplace utility friendship. Indeed, in Aristotle's time, trade took place among cities that did not share in political friendship. Rather, this is to say that people are more likely to stay in or to repeat market exchanges—to cooperate—when the mutual affection they share exists underneath and goes beyond the specific exchanges in question.Footnote 20 The political friendship of NE therefore should not be understood as one of the utility friendships or relations of advantage condemned by Aristotle in Politics III.9.

The Problem of Stability

The fact remains, however, that Aristotle still maintains that the city exists for the sake of living well. Aristotle's understanding of political friendship therefore seems to fit awkwardly with the rest of his moral-political thought. On the one hand, Aristotle argues that the city exists for the sake of human flourishing. On the other hand, he espouses notions that fall short of the high demands of human flourishing—for instance, that the city should care especially about political virtue and that political friendship is a sort of utility friendship. Moreover, as these notions are rooted in a sort of advantage or utility, they provide less stability and durability than would notions rooted in virtue proper—perhaps an odd thought, given that political friendship (as we shall see immediately below) is supposed to serve as a foundational source of stability. Why does Aristotle not recommend that the city focus on cultivating virtuous citizens who are capable of participating in virtue friendships?

I submit that Aristotle chooses the seemingly self-contradictory option because he is deeply concerned about how to foster stability amid real conditions of political complexity, uncertainty, and potential volatility. That is, Aristotle is so concerned about what might happen when citizens in any polity organize themselves into factions and develop mutual animosity that he is willing to have those citizens develop a sense of friendship that might actually compromise their ability to live lives of full human flourishing.

Aristotle famously presents a typology of regimes. This typology is not merely an empirical survey of all the different sorts of regimes that existed in his day; Aristotle declares three sorts of regimes (kingship, aristocracy, and polity) “correct” regimes on the grounds that they are ruled for the sake of the common advantage, and three other sorts of regimes (tyranny, oligarchy, and democracy) deviant on the grounds that they are ruled for the sake of their rulers’ private advantages. Yet just because a regime is ruled for the common advantage does not mean that it is a regime that is conducive to human flourishing—that it is an “arrangement under which anyone might act in the best manner and live blessedly” (Pol. 1324a24–5). Indeed, even though Aristotle argues that such a city must be a moderately sized, self-sufficient city where equal citizens—relieved of the need to deal with necessary tasks by the presence of laborer-slaves (Pol. 1324a26–40, 1325b8, 1329a13–34, 1332b25–7)—are able to rule and be ruled in turn, Aristotle deems regimes like kingship and aristocracy correct; these are regimes where some (or one) rule(s) and others are ruled, but where none rule and are ruled in turn.

This compromise can be accounted for by Aristotle's conviction that even if a best regime did or could exist, aiming to realize the best regime might not be the most useful or practical course of action (Pol. 1288b21–40). It is not enough merely to provide an account of some ideal city without providing viable alternatives for cities where that ideal is untenable or impractical, for what is to come of those cities if there exist no alternatives? Aristotle fears what might happen if any city becomes “filled with enemies” (Pol. 1281b30). So, he is intent on investigating what practicable actions can be undertaken to ensure that these cities are stable.

The citizens of any stable correct regime can thereby be described as political friends (NE 1161a10). Each of the three correct regimes can be ruled for the common advantage when their citizens are committed to the common advantage of those regimes, are capable of sustaining that commitment, and recognize in one another that commitment. When citizens are friends in this manner, rulers can rule with a view to the whole, and the ruled can obey accordingly. Herein lies another reason why political friendship is distinct from other sorts of utility friendship: political friendship can be shared by political unequals—not just political equals. Just as political friendship can be shared among equal citizens who satisfy the property requirements of polity, so too can it be shared between unequals—between the virtuous and nonvirtuous in aristocracy and between the king and his subjects in kingship (NE 1160b1–1161a29)Footnote 21—at least, theoretically.

What this means is that political friendship, as understood by Aristotle, is not merely a special sort of utility friendship; it is a sort of friendship that manifests itself differently in different contexts. This is because what it takes to further the common advantage varies from regime type to regime type. The actions required to further the common advantage of a kingship differ from those that are required to further the common advantage of an aristocracy, and those actions differ from those that are required to further the common advantage of a polity. So, the manner in which political friends exhibit political virtue also varies according to regime type and according to their respective statuses within those regime types; rulers and subjects who are political friends demonstrate to one another that they are indeed politically virtuous by furthering the common advantage in manners befitting their relative statuses within the particular friendship in question.Footnote 22 For instance, a king demonstrates political virtue by ruling over his subjects as a father cares for his children and as a shepherd cares for his sheep, whereas his subjects demonstrate political virtue by obeying him as good children obey (and honor) their fathers. That is, what it means for a king to be politically virtuous differs from what it means for his subjects to be politically virtuous: “what is just in these cases … is not the same for both, but it does accord with merit” (NE 1161a23). Likewise, what it means for a polity's citizens (who rule and are ruled in turn on an equal basis, together) to be politically virtuous differs from what it means for either a king or his subjects to be politically virtuous. Not only does this underscore the nature of Aristotle's understanding of political friendship as a sort of compromise, formed in the name of stability; it underscores why political friendship is a special sort of utility friendship rather than a special sort of virtue friendship: in certain contexts (but not all), the ability to participate in political friendship renders one incapable of participating in virtue friendships. A politically virtuous subject in a kingship who is capable of being ruled is not thereby capable of living “blessedly.” In fact, he might definitively be incapable of living blessedly and of forming virtue friendships as a result of his political virtue, for he is, as a result, incapable of ruling and being ruled in turn.

The fact that Aristotle theorizes ways to foster provisional stability should the bonds of political friendship fail to be realized and the perils of faction loom on the horizon underscores his concern with the problem of stability—and, ultimately, political friendship's importance. Indeed, Aristotle acknowledges that the two most common regime types of his day—oligarchies and democracies—are unstable precisely because their citizens lack political virtue and are not political friends. The rich tend to be arrogant and base on a grand scale, for they possess all the gifts of fortune (money, strength, friends), usually from birth; they probably only know how to rule. Meanwhile, the poor tend to be malicious and base on a petty scale, for they are excessively needy and humble; they are probably only capable of being ruled (Pol. 1295b6–20). As a result, democracies and oligarchies risk becoming cities of contemptuous masters and envious slaves, either extreme oligarchies or extreme democracies—tyrannies. Such cities do not qualify as a cities at all, for cities involve “the element of affection” (Pol. 1295b24).

Aristotle is therefore moved to consider ways in which some degree of stability can be attained when broad segments of the citizenry do not share in political friendship. For example, he proposes that the “middling element” be mobilized to ensure that those deviant regimes become less deviant: “where the multitude of middling persons predominates either over both the extremities together or over one alone, there a lasting polity is capable of existing” (Pol. 1296b38). The middling element helps deviant regimes become less deviant in two ways. First, the presence of the middling element induces an “element of affection” that can help ensure that the given city will still qualify as a city. This is because the middling element, according to Aristotle, is capable of participating in both virtue friendship and political friendship. Unlike the modern middle class—eager to acquire more and become rich, yet anxious about falling into poverty in a context of regularized social mobilityFootnote 23—Aristotle's middling element refers to a rather stable social and psychological state: members of the middling element are those citizens who are moderate in both possessions and character. The middling element is the “readiest to obey reason” (Pol. 1295b6), the most capable of ruling and being ruled, and “alone without factional conflict” (Pol. 1296a6–8). Second, when the middling element comprises the largest part of the city or is “added” to one of the others in order to prevent the regime from becoming an extreme oligarchy or democracy, some degree of stability can be secured: “when the middling element is numerous, splits over the regime occur least of all” (Pol. 1296a6–8).

This passage tells us two things. First, it reveals that political friendship is essential yet potentially elusive. On the one hand, correct regimes can only be stable when members of different factions are political friends with one another, for alternative mechanisms to foster stability are less enduring and reliable than political friendship. When the rich and the poor are not political friends, the stability of the city and of its prevailing regime is not grounded in a mutual commitment shared among citizens, but rather in the ability of “the middling person” to act as “a sort of arbitrator” between factions that would otherwise “not put up with ruling in turn on account of their distrust towards one another” (Pol. 1297a1–5). On the other hand, it appears that a political friendship that spans the entirety of a given polity—one that is not merely confined to the middling element—can only emerge when a variety of factors which help ensure that all (or most) citizens are politically virtuous happen to align. So, either the middling element must comprise the majority—an unlikely event, given that few Greek city states had middling elements large enough to even tip the scale—or the rich and the poor, contrary to their tendencies, must somehow be capable of feeling affection for those outside of their respective classes. This means that the possibility of political friendship might largely be a matter of luck. As is the case with the possibility of the happy life, the possibility of political friendship seems beholden to “many reversals and all manner of fortune” (NE 1100a5).

Second, even though both equal and unequal political friendships can exist in theory, not all sorts of political friendship are equally viable or durable in practice. Great inequality does not render political friendship impossible, but it does render political friendship less probable. So, even though it is possible for a king and his subjects to be political friends—indeed, a kingship can only be truly stable when subjects allow themselves to be led by their king—it is unlikely that an Agamemnon can truly act as a “shepherd for his people” (NE 1161b15)! This echoes Aristotle's insight, discussed earlier, that even though inequality can be present in friendships generally, that inequality cannot be excessive: “for not only are the parties involved [in excessively unequal friendships] no longer friends, but they do not even deem themselves worthy to be” (NE 1158b35).Footnote 24

Contemporary Liberal Democracy and the Metaphor of Political Friendship

Aristotle understands political friendship to be a special sort of utility friendship, shared among citizens who recognize in one another political virtue and a commitment to further the long-term common advantage. The fact that political friendship assumes different forms in different regime types suggests that Aristotle intends for the concept to describe the healthy social bonds of any stable polity. Taking Aristotle's discussion seriously, we are led to believe that politics should not be treated as a joint pursuit of the good life, but also that politics must not be reduced to economics—to a series of mere marketplace, utility friendships.

Aristotle thereby provides us with a strong positive argument for why those of us who live in contemporary liberal democratic societies should appreciate the importance of political friendship. Indeed, the need for political friendship in contemporary liberal democratic societies seems heightened by the present political moment. A quarter century following the demise of communism, the often ferocious narcissism of small differences that characterized liberal democratic politics seems to be making way for a more fundamental polarization. Both the capitalist-skeptic Left and the nationalist-nativist Right have surged at the polls, with the latter frequently challenging core tenets of the liberal democratic political order. So, at least prima facie, it seems plausible to suggest that this trend can be reversed when citizens become political friends by virtue of their shared commitment to the common advantage of their respective liberal democratic societies.Footnote 25

There are, however, normative and practical reasons why we cannot simply resolve our current predicaments by importing Aristotle's understanding of political friendship. First, political friendship must be realized in a manner that neither obstructs the pursuit of other normative liberal democratic goals nor hinders spirited debate over the content of those goals. Even though Aristotle places constraints on the degree to which political friends can be unequal, he insists that both equals and unequals can be political friends. A king can be friends with his subjects, just as aristocrats can be friends with one another. However, given that liberal democracy presupposes the freedom and moral equality of individuals, the political friendship realized in liberal democratic society must be shared among equals; at the very least, political friendship must not interfere with ongoing debates over the meaning of political equality and with efforts to more perfectly realize equality. If not realized in this manner, political friendship might become a source of instability. This is because political friendship would then serve as an unduly conservative source of social cohesion—to consolidate political inequality—and thus prompt the proponents of equality to purposefully destabilize society in the name of justice (e.g., through civil disobedience).

Second, the size of contemporary liberal democratic societies makes it unlikely that citizens will attain the familiarity that Aristotle's understanding of political friendship requires. Earlier, I argued that political friends must be able to recognize in one another political virtue and a commitment to further the common advantage because political friends can only be friends when they verify that they are each not free riders. I suggested that citizens who are strangers might be able to attain such mutual recognition by participating in a series of intertwining networks.

There are at least two reasons why it is unlikely that this proposal can succeed. First, if these networks are developed, but not to the extent that they collectively encompass the entirety of society, then those networks can end up serving as sources of misunderstanding and, indeed, faction. In Robert Putnam's terms, such networks would serve to bond, but not to bridge; they “reinforce exclusive identities and homogeneous groups … [thereby] undergirding specific reciprocity and mobilizing solidarity,” but without establishing links “to external assets … [and] generat[ing] broader identities and reciprocity.”Footnote 26 This problem of isolated networks and factionalism, of course, is not unique to large liberal democracies. However, the size of contemporary liberal democracies exacerbates the problem, as does the presence of a greater number of potential cleavages—racial, ethnic, religious, cultural, economic.

Second, it is unclear whether such networks can facilitate the mutual recognition of political virtue as fluidly as they can facilitate the transfer of knowledge generally. For one thing, these networks are intended to allow citizens to readily call upon specific pieces of information or to have access to expertise when necessary or desired, whereas friendship demands some degree of sustained attention: “many friendships the lack of contact dissolves” (NE 1157b13). For another, even though it is less vigorous than virtue friendship, political friendship does concern the common advantage of a polity—something that is supposed to last and that cannot be “rendered in a short time” (as can be the case with the trade of services) (NE 1158a18). Therefore, even though political friendship does not require citizens to develop the habit of living together, it likely requires citizens to “acquire [some sort of] experience of the other person” (NE 1158a14–5). So, the knowledge of whether a citizen or a group of citizens is politically virtuous cannot be transferred in the manner that knowledge of how to perform a certain task, or even of whether one can trust the provider of a certain service or good, can be transferred.

This does not mean, however, that we must condemn political friendship to irrelevance. I submit that we can begin to think about how to realize political friendship in the contemporary liberal democratic context by considering how citizens who are strangers might still be able to be political friends. That is, we should consider alternative understandings of political friendship whose structures might potentially enable citizens to be political friends, even if they might not be able to recognize political virtue in one another.

I term one such understanding of political friendship “political friendship as conceptual metaphor.” According to this understanding, citizens might be strangers, but they can nonetheless be political friends by virtue of the fact that they consciously or unconsciously understand citizenship in terms of political friendship; they believe that to think and behave as a citizen is to think and behave as a political friend would. When this occurs, citizens might still be strangers, and they might not even feel any affection for one another. However, they will treat one another with a newfound or strengthened sense of reciprocity and inclusion, both in their face-to-face meetings and, perhaps, at the deeper levels of public policy and the basic structure; and a culture of trust that can redress the problem of stability can thereby emerge.

Now, it might be tempting to dismiss this notion of political friendship as wildly idealistic—as placing faith in the magical powers of mere linguistic embellishments to reform citizens’ worldviews and behavior. However, there is considerable evidence that metaphors have cognitive power. Aristotle himself observes that the “perception of similarity into dissimilars” (Poet. 1459a5–7)—the “giving the thing a name that belongs to something else” (Poet. 1457b7–8)—can render visible a hitherto unnoticed and “nameless act” (Poet. 1457b28). With the benefit of modern cognitive science, we can make an even bigger claim: metaphors are not just powerful; they are pervasive and necessary. Metaphors are primarily matters of thought,Footnote 27 and language serves as an instrument of thought; we think in terms of the metaphors that we assimilate during our participation in a given culture. Whether we realize it or not, metaphors structure our experiences, our emotional responses to those experiences, and, most broadly, the rules and dynamics of social exchange. We cannot understand ourselves without them. When we lack a certain metaphor, we cannot make sense of that metaphor; and when we accept a certain metaphor, we ignore or hide away realities that contradict it.Footnote 28

This suggests that we should appeal to the metaphor of political friendship. If we do so, then a culture of trust might emerge in the following manner. Even though they largely employ the metaphors they assimilate, people have some degree of control over which metaphors to invoke via language. So, cultures can evolve through “many iterative cycles of building”Footnote 29 between language and people's habitual ways of thinking and behaving; the rules and dynamics of social exchange can revise themselves, and citizens can come to consider politics and their social relations through new metaphors. Therein lies the potential of a metaphor of political friendship. Should citizens come to understand politics and their social relations through the metaphor of political friendship, they will gain the disposition to behave as political friends would. That is, they will more likely do what they should do to further the common advantage on their own impetus without having first observed their fellow citizens doing the same; and the more they perform these activities, the more they will perform those acts of political friendship as a matter of habit. Certainly, this does not mean that the individual citizen will be content to keep on acting as a political friend when everyone around him/her clearly does not. However, it does mean that citizens’ inclination to act as political friends will not nearly depend so heavily on the verification that each and every one of their fellow citizens is doing what he/she should do to further the common advantage. As a result, so long as citizens who behave like political friends as a matter of habit have a general sense that their fellow citizens tend to also behave as political friends, a culture of trust will emerge and deepen. Citizens will thereby be willing to continue behaving like political friends even if they observe some of their fellow citizens shirking their civic responsibilities.

Put differently, when citizens understand politics and their social relations through the metaphor of political friendship, they can indeed, as Allen suggests, potentially “become friends simply by acting as if they were friends”Footnote 30—but they can only continue being friends if they can observe at least some of their flesh-and-blood fellow citizens acting as if they were friends. So, it is not enough, as Schwarzenbach argues, for citizens who lack “personal knowledge of (or intimacy with) all or even most of their fellow citizens … [to simply demonstrate] a general concern expressed in society's public and political norms … [and] through the public processes of the state's social and political, legal, and educative institutions.”Footnote 31

Obviously, much more needs to be said about this metaphoric notion of political friendship. For example, more needs to be said about what will motivate some citizens to begin to use the language associated with the metaphor of political friendship in the first place. However, what we can say definitively is that when constructing an understanding of political friendship in the liberal democratic context, we must never lose sight of the fact that mere stability is not our objective; our objective is a stability that conforms to the broad normative contours of liberal democracy. This means that the metaphor of political friendship must conceive of citizens as moral equals. It must not be a metaphorical version of the relationship between a king and his subjects in Aristotle's day, the relationship between lords and serfs under feudalism, or even the relationship between the traditional mother and her child. Such metaphorical understandings of political friendship would serve to consolidate relations of hierarchy and, perhaps, oppression. However, the metaphor may draw on the relationship among citizens in the regime Aristotle calls polity, so long as that notion of citizenship is expanded to include far more than property-owning men.

If metaphors do indeed have the power to structure our behavior, then it is of vital importance that we come to understand politics and our social relations through the metaphor of political friendship. If we fail to do so, then we are in danger of becoming captive to alternative metaphors that might eliminate the possibility of trust, let alone friendship, altogether. In particular, we are in danger of understanding politics to be a form of warfare. As Michael Ignatieff observes, when this occurs, “adversaries … are turned into enemies. An adversary has to be defeated, whereas an enemy must be destroyed. You cannot compromise with enemies…. Democracy depends on the idea that you might be able to win over an adversary today and turn him or her into an ally tomorrow…. [But] the war metaphor … legitimizes a ‘take no prisoners’ approach.”Footnote 32 In other words, the absence of political friendship jeopardizes both liberal democracy and stability.

Footnotes

I would like to thank Ruth Abbey, Samuel Bagg, Mario Feit, Michael Gillespie, Ruth Grant, Jack Knight, Antong Liu, Abdeslam Maghraoui, Elliot Mamet, Alexandra Oprea, Thomas Spagens, Brian Spisiak, Isak Tranvik, Baldwin Wong, and my anonymous reviewers. I would especially like to thank Gerald Mara for commenting on several drafts.

1 References to works of Aristotle will be given parenthetically and are to the following editions: Aristotle, , Nicomachean Ethics, trans. Bartlett, Robert C. and Collins, Susan D. (Chicago: University of Chicago Press, 2011); Aristotle, , Politics, trans. Lord, Carnes (Chicago: University of Chicago Press, 1984); Eudemian Ethics and Poetics, in The Complete Works of Aristotle: The Revised Oxford Translation, ed. Jonathan Barnes, vol. 2 (Princeton: Princeton University Press, 1984).

2 Similarly, Sherman, Nancy, “Aristotle on Friendship and the Shared Life,” Philosophy and Phenomenological Research 47, no. 4 (1987): 589–613.

3 Nehamas, Alexander, On Friendship (New York: Basic Books, 2016), 41.

4 Ward, Ann, Contemplating Friendship in Aristotle's Ethics (Albany: State University of New York Press, 2016), 111–13.

5 Ward, Contemplating Friendship, 115–23. Nehamas's and Ward's discussions are reviewed in Abbey, Ruth, “Review Essay: On Friendship,” Review of Politics 79, no. 4 (2017): 695–707, alongside Digeser, P. E., Friendship Reconsidered: What It Means and How It Matters to Politics (New York: Columbia University Press, 2016).

6 von Heyking, John, The Form of Politics: Aristotle and Plato on Friendship (Montreal and Kingston: McGill-Queen's University Press, 2016), 11. Von Heyking is here drawing on Aristotle's characterization of virtue friendship as “complete” or “perfect” (teleia) friendship and utility and pleasure friendships as incomplete or imperfect (NE 1156a6–1156b33).

7 Von Heyking, Form of Politics, 10–11.

8 Schwarzenbach, Sibyl A., On Civic Friendship: Including Women in the State (New York: Columbia University Press, 2009).

9 Cooper, John, “Political Animals and Civic Friendship,” in Aristotle's Politics: Critical Essays, ed. Kraut, Richard and Skultely, Steven (New York: Rowman and Littlefield, 2005), 71–76.

10 Frank, Jill, A Democracy of Distinction (Chicago: University of Chicago Press, 2005); Inamura, Kazutara, Justice and Reciprocity in Aristotle's Political Philosophy (Cambridge: Cambridge University Press, 2015); Rasmussen, Douglas B. and Den Uyl, Douglas J., “Aristotelianism, Commerce, and the Liberal Order,” in Aristotle and Modern Politics: The Persistence of Political Philosophy, ed. Tessitore, Aristide (Notre Dame, IN: University of Notre Dame Press, 2002), 278–306; Yack, Bernard, The Problems of a Political Animal: Community, Justice, and Conflict in Aristotelian Political Thought (Berkeley: University of California Press, 1993).

11 Frank, Democracy of Distinction, 161–62.

12 Çidam, Çigdem, “Unruly Practices: Gezi Protests and the Politics of Friendship,” New Political Science 39, no. 3 (July 2017): 379.

13 Frank, Democracy of Distinction, 140–48.

14 I thank one of my anonymous reviewers for coining this phrase.

15 Yack advances a slightly different interpretation of this passage. On Yack's account, conflating political friendship with anything more than legal utility friendship—with moral utility friendship, let alone virtue friendship—is dangerous because it raises the expectations political friends have for one another, thereby making it more likely that they will disappoint one another: “they will complain not only about the failure to return advantage for advantage, but about betrayal and treason” (Yack, Problems of a Political Animal, 121).

16 Cooper, “Political Animals,” 71–75.

17 Johnstone, Steven, A History of Trust in Ancient Greece (Chicago: University of Chicago Press, 2011), 12–34.

18 Allen, Danielle S., Talking to Strangers: Anxieties of Citizenship since Brown v. Board of Education (Chicago: University of Chicago Press, 2004); Çidam, “Unruly Practices,” 379; Jill Frank, Democracy of Distinction.

19 Ober, Josiah, “Classical Athenian Democracy and Democracy Today,” in Athenian Legacies: Essays on the Politics of Going Together (Princeton: Princeton University Press, 2005), 41.

20 This characterization of political friendship is supported by recent research that suggests that marketplace cooperation is more likely when there is a preexisting culture of trust. See Guiso, Luigi, Sapienza, Paola, and Zingales, Luigi, “Does Culture Affect Economic Outcomes?,” Journal of Economic Perspectives 20, no. 2 (Spring 2006): 23–48.

21 This explains why Aristotle remarks that a regime is just to the extent that friendship—equal or unequal—appears in that regime (NE 1161a10).

22 Accordingly, Aristotle also argues that “the virtue of the citizen must necessarily be with a view to the regime. If, then, there are indeed several forms of regime, it is clear that it is not possible for the virtue of the excellent citizen to be single, or complete virtue…. That it is possible for a citizen to be excellent yet not possess the virtue in accordance with which he is an excellent man, therefore, is evident” (Pol. 1276b31–6). I thank Ruth Abbey for reminding me of this passage.

23 I paraphrase the formulation offered in Alexis de Tocqueville, Democracy in America, ed. J. P. Mayer (New York: Harper Perennial, 2006), 636.

24 Similarly, Lorraine Pangle notes that even though timocracies and democracies may find it difficult to realize political friendship because their members tend to have lax morals, it is far more difficult to realize political friendship in monarchies, let alone tyrannies; given that they presume their citizens to be unequal, monarchies impose additional structural barriers to friendship (Pangle, Lorraine Smith, Aristotle and the Philosophy of Friendship [New York: Cambridge University Press, 2003]).

25 It is important to stress that the full case for why we must appeal to the resources of political friendship in order to adequately address the problem of stability requires both positive and negative arguments: a positive argument that demonstrates political friendship's importance; and a negative argument that demonstrates why alternative paths to stability, such as John Rawls's shared commitment to justice, Friedrich von Hayek's spontaneous order, or Chantal Mouffe's pluralist agonism, are either insufficient or wrongheaded. That is, even though Aristotle provides a strong positive argument for the importance of political friendship, a strong negative argument still needs to be made.

26 Putnam, Robert, Bowling Alone (New York: Simon and Schuster, 2000), 22–23.

27 Cognitive science suggests that around 98 percent of thought is unconscious.

28 For more, see Lakoff, George, The Political Mind: Why You Can't Understand 21st-Century American Politics with an 18th-Century Brain (New York: Viking Penguin, 2008); Lakoff, George and Johnson, Mark, Metaphors We Live By (Chicago: University of Chicago Press, 2003); and Boroditsky, Lera and Thibodeau, Paul H., “Metaphors We Think With: The Role of Metaphor in Reasoning,” PLoS ONE 6, no. 2 (2001): e16782, doi:10.1371/journal.pone.0016782.

29 “Economist Debates: Language: Statements,” The Economist, Dec. 22, 2014, accessed Dec. 23, 2010, http://www.economist.com/debate/days/view/628.

30 Allen, Talking to Strangers, 156.

31 Schwarzenbach, On Civic Friendship, 53–54. One immediate objection, which I do not have the space to rebut here, is that my proposed metaphoric notion of political friendship cannot succeed because it cannot overcome the need for political friends to recognize in one another political virtue and a commitment to the common advantage. For example, Digeser argues that even though the metaphor of political friendship may be able to overcome the problem of mutual recognition of motives, it still cannot generate trust, for trust demands that people recognize the correct sort of motivation or disposition in one another and can count on one another to act in a particular way. See Digeser, Friendship Reconsidered, 110ff.

32 Ignatieff, Michael, Fire and Ashes: Success and Failure in Politics (Cambridge, MA: Harvard University Press, 2013), 150–52.