What is the relation of ontology to political theory? Are statements about foundations, such as “being is constituted by difference (or power or language),” relevant to political analysis? The discipline of political theory is split, broadly speaking, between theorists who believe that foundations can be safely ignored and those who believe that mistakes and prejudices in ontology bias results. Debates about foundations have shaken political theory along with every humanistic field during the past five decades. Critiques of foundations have come from many quarters within the analytic tradition of philosophy, influencing the political theory discipline through Rorty (Reference Rorty1989) and others. The continental tradition of postfoundationalism during the same time period has been nearly synonymous with postmodernism: Derrida Reference Derrida and Spivak1976 and Foucault Reference Foucault and Smith1972, Reference Foucault and Sheridan1995, among others, popularized the combining of literary and social theory with foundational critique; related ideas later influenced political theory (e.g., Connolly Reference Connolly1991; Derrida Reference Derrida and Collins1997; White Reference White1991, Reference White2000).

The problem of foundations is intriguingly bound up with the problem of friendship. Friendship is a human good, and any politics wanting to promote friendship among citizens seems to presuppose a substantive, which is to say foundational, account of the good. At the same time, friendship appears necessary to self-knowledge, through which alone, if ever, foundations could adequately be investigated. For example, Sandel (Reference Sandel1998, 147–52, 178–83) famously attacked liberal foundations—in particular, John Rawls’ epistemology of the self—on the grounds that without constitutive attachments to friends, no self could ever know itself sufficiently to ground liberal choice. Yet, theorists who wish that liberal individualism could be supplemented or moderated by civic friendship within the confines of liberalism must be deterred by the deterministic character of communal beliefs. Strong ontological claims seem to go together with strong civic bonds. The most notorious example in the Western canon is Plato's Republic. Plato's substantive account of the good is believed to motivate his attempt to unify the citizens through civic friendship, raising for liberals the specter of authoritarianism. This reading of Plato has become the basis for one of the most enduring arguments that foundational inquiry is politically irrelevant, at best, and politically dangerous, at worst (e.g., Popper Reference Popper1950; see Mara Reference Mara1997, 4–8, for bibliography). Mainstream political theorists, particularly liberal theorists, may fairly be said to be wary of foundational inquiry within political theory (e.g., Rawls Reference Rawls and Freeman1999a).

Yet, as the example of the modern rejection of Platonic metaphysics shows, postfoundationalism is only the most recent in a long line of foundational critiques. The historical trajectory of thought has moved from rejecting mistaken foundations to rejecting the possibility of foundations at all.Footnote ¹ A striking corollary of both early modern and postmodern foundational critiques is their dark or skeptical view of human relations. Although Aristotle propounded the natural sociability of humans, ideas of “negative association” have governed much of political theory at least since Hobbes. Individual actors seek power and security. Groups coalesce in opposition to competitors—“outsiders”—who are needed to drive together and keep together associates who would otherwise find reasons to repel one another (cf. Evrigenis Reference Evrigenis2008). Love and friendship were relegated to the private realm by early liberal thinkers. But the barrier between public and private has not always proved strong enough to prevent encroachment by the new political account. Most recently, postmodern thinkers have reopened the private realm cordoned off by liberalism, bringing political assumptions such as power and enmity to bear on love and friendship. Liberal affections are private,Footnote ² whereas postmodern affections are public and political. In the hands of Foucault (Reference Foucault, Hurley and Rabinow1997), for example, power becomes a supple, generalizable concept at the heart of all human activity, including the sexuality on which he would base a new form of identity group friendship. Derrida (Reference Derrida and Collins1997), seeks a universal political friendship but believes that only antagonism creates meaning; as a result he structures his treatise on friendship around an extended dialog with Carl Schmitt's (1976) theories of enmity. Although conceived as the very antitheses of foundations, the postmodern categories of power and enmity take on the functions of new, negative foundations replacing the old positive ones. On the one hand, power and enmity are capable of refinement and, as such, can assume various guises in order to constitute many phenomena previously believed natural or ethical. On the other hand, power relations and antagonisms are capable of shifting and thus often undermine anything constructed on (or out of) them, a feature that adds to their attractiveness as tools of critique. In the limit, negative foundations become blanket propositions believed to underlie any and all phenomena.

Surprisingly, Plato's dialog Lysis depicts Socrates taking a foundationally skeptical approach to friendship of all kinds. Socrates appears to sever his reasoning from the good, or at least from a highest good. He auditions an ontology in which nature does not form a cosmos or ordered whole. And he implies that all friendships, even private ones, are best thought of as negative associations. Crucially, the Lysis exhibits flexibility in Socrates’ thinking about foundations. Perhaps because foundational questions remain the most difficult (and unsettled) of questions, Socrates appears to choose his foundational accounts opportunistically, letting the subject matter determine his ontology. Thus, in a conversation about friendship, Socrates shifts away from his usual account of foundations only with the specific purpose of illuminating the problem with friendship, namely, that friendship is both necessary to and a brake on philosophy. I argue that possible ontologies, for Socrates, arise out of his philosophic activity and remain subject, in principle, to revision. Although Socrates appears postmodern in his ontological and epistemological skepticism, his opportunism stands in contrast to the blanket application of foundational skepticism to any and all problems in much postmodern theory. Finally, I argue that political theorists could replicate Socrates’ procedure by considering themselves at liberty to choose different ontologies based on their utility for understanding different questions, thereby remaining aware of ontological dimensions while avoiding both the Scylla of rigidity and the Charybdis of nihilism in foundations.

INTRODUCTION TO THE LYSIS

The dialog's argument has five major sections. In the first section (207c–211d), Socrates drives a wedge between young Lysis and his friend, and tries to convince Lysis that his parents do not really love him. In the first blush of philosophic discovery, Lysis comes to believe that useful knowledge is the only thing people really love. In the second section (211d–216b), Socrates begins explicitly inquiring into the nature of friendship, examining reciprocity, likeness, attraction between opposites, and attraction between people who are both good, both bad, or one good and the other bad. Two people who are alike, for example, cannot be friends because they derive no utility from each other because each already has what the other has. For the same reason, good people cannot be friends. Socrates’ assumptions seem limited to instrumental goodness. In the third section (216c–218c), Socrates examines a number of formulas in which only people neither good nor bad can need friends. Friendship is motivated by fleeing evils or enemies as much as by seeking the good or useful. In the fourth section (218c–221d), Socrates’ instrumentality becomes thematic; he reasons that if there were a “first friend” or highest good toward which all our lesser friendships aimed, we would not love that highest good because it would not be “useful” to us (i.e., for procuring us any higher good). In the absence of a summum bonum, Socrates’ world falls into ontological evil, in which only bad things exist. The earlier utility seeking now becomes a search for remedies—lesser evils to assuage our greater evils. The fifth and final section (221d–222e) ends aporetically as Socrates tries out a bewildering array of hypotheses dominated by the hypothesis that friendship is a love of “one's own,” an idea that looks back to the parental love Lysis had trouble explaining.

Thus, the Lysis treats friendship in the widest sense, seeking the common denominator among such varied human relationships as parental affection, a ruler's esteem, and business relations, and even comparing them to nonhuman attractions such as the reciprocity of heat and cold, while pursuing a resolutely demystifying or debunking tack throughout. Instead of elevating politics to the level of friendship, as the Republic does, Socrates here lowers friendship to the level of the political, attempting to see friendship as a utility calculation and a negative association. As such, the dialog is a crucial supplement to political dialogs in which civic friendship is an ideal (e.g., Laws 5, 739c–e, Statesman 311b–c). Among the studies of the Lysis, none has explored its connections to postmodern theory, although several have pointed out its dark or troubling account of friendship.Footnote ³ Even as the major arguments of the Lysis explore themes associated with postmodernism, the dramatic frame of the dialog supplies motives for why Plato's Socrates might want to explore such themes. Far more than in most Platonic dialogues, Socrates has an “agenda” or extraphilosophical interest in conversing, known to his audience but not to his interlocutors. Briefly, Socrates is persuaded to join the conversation in a wrestling school where a pack of postadolescent youths hang out to look at and chat up schoolboys. The most attractive of the latter, Lysis, is the object of unrequited desire of the lovelorn Hippothales, on whose behalf Socrates agrees to act as Cyrano. At points early in the dialog, Socrates seems successful in procuring Lysis for himself (e.g., 211a); near the end, however, Socrates explicitly gets Lysis to agree that he loves Hippothales. The dramatic frame thus likens Socrates to a pander or go-between, assimilating Socrates’ own search for a philosophic friend (211d–212a) to this potentially predatory quest. I argue that Socrates’ foundational doubt about friendship serves the dual purposes of seducing Lysis away from his current friends and loved ones and of illuminating an intimate aspect of the philosophic experience, an aspect which the implicit “corruption of youth” in some respects parallels. The fact that Socrates goes to such lengths to prevent love of friends from taking precedence over love of wisdom makes the Lysis’ conversation analogous in several respects to postmodern themes emphasizing the weakness of rationality and the power and persistence of the irrational. The Lysis forces us to look at friendship as if it had no foundations in either nature or the good, a procedure that calls into question the goodness of being itself and enables us to expand our understanding of the place of foundations in Plato's political theory.

THE POLITICS OF PRIVATE LIFE

Socrates begins the first section by eliciting the disputed and contested aspects of Lysis’ friendship with young Menexenus. Who is older, who is nobler, who is better looking? He claims to draw the line at asking “Who is wealthier” because friends’ possessions ought to be held in common, an ideal to which the boys readily assent (207c). Under Socrates’ guidance, Lysis will come to realize that he has always found Menexenus to be “contentious” (211b8–9).Footnote ⁴ A childish hope of seeing Menexenus “punished,” or humiliated by Socratic dialectic, will initially attract Lysis into Socrates’ inquiry into the nature of friendship (211c3ff.). But in the first section of the dialog, Menexenus is temporarily called away, and Socrates focuses on Lysis alone. As he had pitted the two boys against each other, Socrates now begins a line of questioning that will pit Lysis against his parents. In a number of unfair ways, Socrates insinuates that Lysis’ parents do not love him. He points out that they will not let him do whatever he pleases—he cannot drive the chariot, beat the mule team, or touch his mother's weaving instruments. Even slaves can do some of these. The boy has less power than a slave (e.g., the responsibility of mule driving belongs to a slave). His mother, in fact, would beat Lysis if he touched her things (208e). Lysis, he concludes, does not get to “rule over” anyone and derives no advantage even from his own body, which has a slave attendant set over it (208e–9a).

Socrates thus begins by pointing out the political aspects of relationships that modern liberalism traditionally sees as personal or private. He introduces into friendship and family love the ideas of contestation, the power to do as one pleases, masters and slaves, and ruling. Ancient privacy was different from modern privacy; in particular, the latter two ideas, slavery and ruling, were staples of the Greek household. Aristotle, for example, finds within the household not only masterly rule, but also, in the marriage relation, reciprocal political rule, although he concedes that wives do not actually rule in their turn (Politics 1.1252a24–b15, 1259a38–b10). An important postmodern critique of liberalism has been that liberal theory too rigidly demarcates between public and private spheres, between a political realm of state power, on the one hand, and a private realm of free association on the other—a boundary that postmoderns claim is porous. Under the influence, in particular, of Foucault, many postmodern theorists see power inhabiting and shaping both realms. “The personal is political” captures this insight in a popular formula. Postmoderns thus seek to repoliticize the human relationships that they see as having been illegitimately privatized by liberal theory. Hence, Foucault's recourse to the illiberal social life of antiquity (e.g., Foucault Reference Foucault and Hurley1990). This affinity between the Lysis and postmodern theory owes as much to ancient society's lack of separation between state and society as it does to Socrates’ deliberate arguments. But we will see a much closer resemblance.

Although he takes his initial bearings from ruling and doing as one pleases, Socrates’ analytic tool of choice is not power, like Foucault's, but the related concept of utility. Instead of distinguishing between using one's slaves and loving one's children, Socrates claims that usefulness is what makes children loveable. “Now then, your father does not love you, nor does any person love another to the extent that [the loved one] is useless” (achrestos; 210c7–8). Although economic and utilitarian strands of classical liberalism characterize personal choices as “utility,” most modern liberals in the Kantian tradition find that considerations of utility taint the purity of love. Such readers, like Vlastos (Reference Vlastos1981), are more likely than ancient readers to find Socrates’ account objectionable. Ancient readers assumed as a matter of course that children were useful, especially as a crutch for parents in old age (Obdrzalek Reference Obdrzalek2006). For example, the dialog mentions how children read aloud to and write things for their parents, tasks that, in an era before bifocals, might otherwise have been fulfilled by a literate slave (209a7–b4). Lysis’ parents also permit him to do whatever he wants with his lyre (209b4ff.), presumably because they derive enjoyment from his music just as they derive advantage from his literacy.

Yet, Socrates’ reduction of love to utility will prove extreme even by ancient Greek standards. Lysis gives two defenses of his parents’ conduct: (1) he is not old enough yet to do as he likes, and (2) he is not knowledgeable enough yet (epistamai; 209a4, 209c2). Socrates treats the latter appeal to knowledge as a decisive advance in the argument. It is this knowledge that makes Lysis useful to his parents and others. He concludes that Lysis’ parents—and anyone else, for that matter—can only love him to the extent that wisdom makes him useful and good:

This argument employs the typical Socratic paradox that knowledge is the only true title to ownership (e.g., Republic 2, 376b; cf. Lysis 210b–c). Whatever the legal title may say, people only own what they know how to use. If Lysis does not know how to use the chariot, his wealth, or even his own body, then he does not really control them. But the argument goes on to associate ownership with love. Lysis will be of no use to others and hence not really owned by his own people; thus, he will be unloved.

Is useful knowledge the real key to parental love? Socrates claims that Lysis’ father will hand over his property to Lysis’ care not on the death of the father or when Lysis comes of age, but rather as soon as Lysis’ mental competency exceeds his father's. Such a legacy is not unthinkable, and fathers retiring in their sons’ favor was perhaps more usual in Athens than today (Penner and Rowe Reference Penner and Rowe2005, 20). But Socrates immediately generalizes his claim in a way that brings out its absurdity (Price Reference Price1989, 3). Not only parents but neighbors, too, will place their property under Lysis’ control as soon as they perceive his superior competence. So will all the Athenians. So will the Great King of Persia (209c6–d6). The elephant in the room here—the obvious thing Socrates is not admitting—is that no one cedes control of property on the basis of utility alone: parents hand things on to their own children. Lysis’ neighbors can be expected to admire (and envy) his brilliant competence, but the beneficiaries of their legacies will remain their own children, not a useful stranger. The passage quoted previously may be said tacitly to bear witness to this omission. The word that can be translated “relatives,” oikeioi, more literally means one's own (or “akin” in a broader sense than blood relation). The phrase “all people will be your friends and all will be your own [oikeioi]” makes friendship and “ownness” virtually synonymous.Footnote ⁵ The end of the dialog (221e3–4) will explicitly entertain the hypothesis that one's own [to oikeion] is the object of eros, friendship, and desire. But if most of us are mistaken about what is our own, lacking the knowledge to use it and thus truly to make it our own, then it begins to look like we are also mistaken about who or what are really our friends.

Young Lysis becomes intoxicated by the heady vision of his useful knowledge being a title to rule and do whatever he wants. Socrates serves up an image of the Athenians as a whole placing the city's affairs in Lysis’ hands (209d4). Even barbarians, like the King of Persia, will let Lysis do whatever he wants—such as throwing handfuls of salt in the soup, if he were the King's cook, or throwing ashes in the heir apparent's eyes to cure them, as Lysis and Socrates travel the world together as itinerant doctors (209d6–10a8). In the first blush of philosophic discovery, Lysis falls prey to a belief that people actually behave the way Socrates has argued is rational, pursuing utility even in their most emotional or passionate commitments. The truth would appear to be that love and friendship make most people temporarily forget their own utility. Again, one motive for Socrates’ misleading argument is known to the older listeners but not to the two boys. In keeping with the conventions of homoerotic seduction, Socrates must separate Lysis from his philoi—his relatives and friends—who are expected to oppose his entering a new, erotic relationship with Hippothales (cf. Faraone Reference Faraone1999, 86–87; Xenophon, Symposium 8.19). Parents set slave attendants over their sons precisely to prevent such conversations from happening, and the conventional assumption is that agemate friends like Menexenus will vociferously oppose all intercourse between the beloved and the older lover (see, e.g., Aeschines, Against Timarchus 9–12; Plato, Symposium 183c). But Socrates now has Lysis whispering in his ear “playfully and lovingly” (211a). Socrates gets Lysis to admit (210c) that some person or thing may be “more our own” than our own father and mother (oikeioteron, i.e., more closely related to us), a kind of corruption of youth from the parental point of view. Crucially, Lysis is going to need Socrates as his friend: only together can they accomplish his dreams. Socrates thus continues to push his seduction (although he will later “prove” that Hippothales is Lysis’ own; oikeios, 222a–b). The philosophic parallel to this seduction is Socrates’ success in making us wonder whether most friendships are ignorant loves of our own, whereas true ownness would have more to do with utility than most friends would care to acknowledge.

THE ABSENT “FIRST FRIEND”

Both utility and ownness, as accounts of friendship, are amoral in contrast with the accounts of erotic love given in the Platonic corpus. In the Lysis, the good is glossed as the useful. In Symposium, erotic love also has this useful or instrumental aspect—lower loves are rungs on a ladder to higher loves—but the top of the ladder holds out hope that love is ultimately oriented toward virtue and godliness (Symposium 212a5–7). A narrower, more selfish orientation is thus evident in the Lysis: the self is clearly at the center of ownness, just as utility seeking is a form of self-interest. Friendship seems to be the main target of Socrates’ skepticism. For example, a similar skepticism applied to the selves that compose a friendship might have permitted Socrates to avail himself of an important moral result also found in postmodern thought: namely, if the self cannot be known to be bounded, then the site of utility calculations is called into question. The self might well intersect with other selves. Instead, Socrates leaves the utility-seeking self or soul unproblematized and trains his skepticism on friendship. Socrates’ reduction of friendship to utility and ownness is extreme even by ancient Greek standards. Regard for others, in particular, was seen to be a requisite of friendship; Aristotle reports how contemporary opinion held that a friend should “love the friend for his or her own sake” (Nicomachean Ethics 9.1166a2–4; cf. Rhetoric 2.1380b36–37; Konstan Reference Konstan1997, 3–6). This was opposed to using the friend for some further gain (Nicomachean Ethics 8.1157a14–16, 26–33). Aristotle admits that some lower friendships are based on mutual utility. In fact, he holds that civic friendship is a friendship based on utility (Eudemian Ethics 7.1242a7–9, 1242b23–7). Citizens use one another. For the grounds of love (utility) to be different from the object of love (fellow citizens) obviously raises questions about who (or what) the real friend is. Socrates may be correct that people feel affectionate toward others on whom they depend for utility; civic examples might include public attitudes in the United States toward firefighters and police just after 9/11. It is possible that Kantians like Vlastos (Reference Vlastos1981) are too squeamish about egoistic motives in friendship. Yet, for Aristotle, civic friendship relies on utility for lack of a better motive. Aristotle considers utility friendships to be friendships only “by resemblance” to the primary friendship based on goodness (Nicomachean Ethics 8.1157a26–33; cf. 8.1156b10–12, 9.1171a18–21). Socrates, in contrast, claims that utility governs even private and personal philia. Socrates thus might be said to take a politicized view of friendship.

How does such utility relate to goodness? In the passage quoted previously, Socrates said Lysis could someday become “useful and good.” In the ensuing arguments, Socrates will sometimes refer to people and things as good without reference to the useful. Does the good remain instrumental, or does Socrates recur to the more familiar moral conception of goodness? To resolve this issue, we must look ahead in the dialog to the passage where the good becomes thematic, the passage on the “first friend” (218c–220d). Socrates works through an analogy with the medical art: a neither good nor bad body is a friend of the medical art for the sake of health. Such a body is neither totally healthy nor terminally ill, either of which would imply no need for medicine's friendship. In choosing an impersonal scientific discipline as a “friend,” Socrates exemplifies the depersonalized assumptions that Vlastos deprecated. But he also nudges the argument toward becoming a “friend,” philos, of sophia (cf. 212d) or wisdom, which cannot reciprocate love. The medical art and sickness appear to be stand-ins for philosophy and ignorance. Health is a good (219a) for the sake of which we become friends of medicine. But health, in its turn, is our friend—a greater friend than medicine. All friends, Socrates posits, have this character of being for the sake of higher friends (219c). But, Socrates says, this involves an infinite regress. Does there not have to be a “first friend” (219d1)—a highest good at which all the rest aim—to secure their status as good for something? The merely instrumental character of the lesser goods calls into question their status as friends. Socrates calls them “phantom friends.” Only the first friend is a real friend (220a–b).

Because the first friend is not loved for utility, Socrates must now reverse the earlier premise that all friends are only for the sake of some further friend (220b). The only real friend does not have this character. But instead of allowing that real friends are loved for their own sakes, Socrates goes back to the earlier formula, which gave two reasons why we love a friend. The full formula had both a stick and a carrot, something we flee as well as something we seek. In the formula, a neither good nor bad body loves medicine not only for the sake of future health, but also because of its current sickness. Stated abstractly, “The neither good nor bad is friend to the good [e.g., medicine] because of what is bad and an enemy [e.g., sickness], for the sake of what is good and a friend [e.g., health]” (219b). Because the second half of this formula, the “carrot,” is now the rejected premise (one friend is not for the sake of other friends), Socrates wonders why we continue to love the first friend at all. Is it only because of the first half of the formula—the stick? Is the ultimate good loved only “because of what is bad and an enemy” (220c)? In the absence of disease, no medicine is required (220d). But in the absence of the bad, would we love the good at all? Socrates now posits a cessation of evils (a theme to which he will return) in order to see whether evil is what makes us love the good, rather than our loving it for its own sake. He asks whether the ultimate good (all instrumental goods being mere phantoms) would still “be in any way useful to us, or would it instead be useless?” (220c1–7).

This is the strangest moment in a strange dialog. The first friend is a likely candidate for a foundational conception of the good. It would be that at which all loves aim. But after painstakingly establishing that utility is not real goodness at all (useful things are not good in themselves), Socrates turns around and asks what use is real goodness. The answer, of course, is that the ultimate good can be of no use. All use is defined relative to it. Therefore, its goodness, if it has any, must transcend mere use. Although he has ostensibly rejected utility, Socrates here explicitly continues to hold every good accountable to the sole criterion of utility. Socrates could not ask what use the nonutilitarian good is unless utility remained the salient ingredient in his concept of goodness. He repeats his criterion in the strongest terms: does the good “have no use [oudemian chreian] in itself and for its own sake?” (220d6–7). The tag “itself and for its own sake” is strangely superfluous when talking about use. Utility is a relative term, while “in itself and for its own sake” is an absolute.Footnote ⁶ Nothing in itself and for its own sake is useful. Its utility emerges only when we consider it in relation to other things. Socrates substitutes “use” (chreian) where only the good (agathon) is appropriate. Utility thus continues to stand in for the good so far as the Lysis is concerned.

Socrates extends his instrumental criterion further. Because the first friend cannot be instrumental to some higher friend, the only remaining possibility is that it is instrumental to some enemy (“for the sake of an enemy”; 220e). But this is hardly the only possibility, unless one assumes the first friend must be instrumental. Socrates thus adheres to instrumentality—use—even at the cost of making enmity more basic than friendship. The first—and only real—friend is not loved for its own sake, but rather for the sake of an enemy. The enemy comes first. Socrates’ merely instrumental rationality leads away from a foundational ground of friendship, such as natural attraction to the good, leaving a vacuum that must be filled by enemies.

Although the mysterious passage on the “first friend” is only one dialectical swell in a sea of hypotheses in the Lysis, the passage puts Plato's Socrates temporarily on the same playing field with early moderns who rejected a summum bonum.Footnote ⁷ We are enabled to see what Plato believed were important ramifications of positions he disagrees with elsewhere. What will emerge is that, for Plato, the “modern” assumptions about power and enmity in politics cannot stand but must inevitably refine themselves into postmodern assumptions. Hobbes’ famous passage connects the loss of the summum bonum with the restless search for power after power. Amoral instrumentality—utility in the absence of the good—is approximately what “power” means in the modern (although not postmodern) lexicon. Socrates and Lysis do not use the Greek word for power, dunamis, to express what they mean by ruling (archein) or being one's own master. The Greek word sometimes means mere “potential to act,” as opposed to the modern meaning, the ability to coerce or acquire. “Power” certainly captures what Lysis means by getting to do whatever he wants. Yet, Socrates predicates getting to do what one wants on knowing how to do things. As I argue, Socrates’ focus on Lysis’ internal preparation (before he can exercise power in the world), and his attempt to get Lysis to translate his desire for power into useful knowledge are a commonplace of Socratic thought that was, in turn, an influential source for Foucault's transformation of power into ascesis and his combined term “power/knowledge.” Knowledge in the absence of a highest good is a refined version of power, technical expertise that can be used for good or ill (cf. Republic 1, 333e–334a). The upshot of Plato's thinking through the absence of a summum bonum will be an internalizing, generalizing, and diversifying of the concept of utility, a transformation remarkably similar to Foucault's reconceptualization of power.

FOUCAULT'S “FRIENDSHIP AS A WAY OF LIFE” AND THE USE OF PLEASURE

Power is the first of two major substitutes for ontology through which postmoderns have sought to understand social phenomena, a tool of critique especially associated with Michel Foucault. Comparing the Lysis briefly to two of Foucault's arguments will show how the ancient dialog can help reveal strengths and weaknesses of this negative foundation or antifoundation for political theory. In “Friendship as a Way of Life,” Foucault argues that affection is constitutive of political institutions. In the form of an interview, Foucault (Reference Foucault, Hurley and Rabinow1997) argues that a sexual identity group, male homosexuals, ought to create a new ethics and culture of friendship based on their sexuality. Although his histories of sexuality study illiberal social practices of antiquity such as pederasty in part to see how their overt hierarchies and exploitations were negotiated, Foucault here repoliticizes friendship in an ostensibly nonexploitative way. He thus shows one way in which power, as a substitute for ontology, can be compatible with a normatively good social practice. Foucault is able to arrive at this conclusion due in part to his reconceptualization of power (cf. Foucault Reference Foucault and Gordon1980, 102). Where a Hobbesian conception of power might seem capable of describing only a subset of social phenomena, such as coercion by the state, Foucault argued that modernity is characterized by decentralized power, a “microphysics” of power, in which power never quite emanates from official hierarchies, but instead operates in private relations. This subtle power makes the allegedly separate spheres of state and society blend together: each takes on aspects formerly reserved for the other. As I argue, Foucault (Reference Foucault, Hurley and Rabinow1997) uses this reconceptualized power to explain love and friendship.

As the interview begins, the problem of old age appears to provide the motivation for stable, lasting friendships; the interviewer begins by pointing out that Foucault is in his fifties, whereas the readers of the magazine, Gai Pied, are men in their late twenties and early thirties (Foucault Reference Foucault, Hurley and Rabinow1997, 136). Foucault avoids this natural motive for friendship, however, by viewing friendship as a form of resistance that subverts the dominant paradigm of homosexuality. Brief, one-off sexual encounters are a “neat image of homosexuality” that cancels out the societal “unease” that a lasting male–male relationship would generate. Not sex but “[a]ffection, tenderness, friendship, fidelity, camaraderie, and companionship” among homosexuals are what is troubling to the larger society (136–37). In such circumstances, friendship is a rebellion, whereas brief encounters are a submission to the dominant discourse.

Foucault proceeds to illustrate how power constrains itself through the very exercise of it, and how submission to dominant modes of thought can paradoxically become a form of resistance. His one use for the brief encounters of “mustached ‘clone’” (138) homosexuals in the late seventies (the “neat image”) is that their ferocious pursuit of short-term pleasures showed an inverted kind of self-restraint. The sixties ideal, he says, had been to “liberate yourself from the hideous constraints” against free love. This liberation movement, for Foucault, was a mush of indistinctness and lack of constraint: everyone would love everyone, and every love would be identical. But the mustached clones by their very lack of restraint among themselves showed a new form of constraint by their ever-growing aversion to women (138), an aversion that made them proof against “flounder[ing] in this polysexuality where there are no longer any differences between the two.” By wanting one thing so much, the clones ended up with the equivalent of constraint or self-restraint in another area. In submitting so freely to the power of the dominant discourse, they ended up exercising power by resisting another portion of the dominant discourse.

A liberal reader might object that Foucault here equivocates on “constraint” in order to cast these men's actions in terms of power, submission, and dominance. What was really their sexual liberation Foucault reclassifies as a creative internalization of power—power constraining itself. The fact that he repairs to so recherché a meaning of constraint suggests that such power is not the true basis for the men's solidarity and, indeed, that power will not be the explanatory key to Foucault's proposed new friendship either. At the limit, power would merely be the path Foucault needed to take to arrive at friendship, a place traditional moralists end up via other paths. Yet, Foucault's ensuing discussion of “ascesis” (1997, 137) points to his histories of sexuality, where we find that power is indeed his basis for a restraint on behavior that must ultimately include friendship. Although “[a]sceticism is a renunciation of pleasures,” ascesis in contrast is “the work that one performs on oneself in order to transform oneself . . .” (137). In The Use of Pleasure (1990), Foucault derives this Greek concept (askesis) in part from Xenophon and Plato, particularly from their accounts of how Socrates dealt with power-hungry Athenians such as Alcibiades and Callicles, attempting to curb and channel their will to rule. The desire to govern others for one's own advantage is inseparable from the desire to be effective at governing. Socrates can therefore make Alcibiades and Callicles think through their desire for power. If they are not good at governing, their acquisition of power will give them no advantage. If they cannot govern themselves, they will be ineffective at governing others even for the purpose of exploiting them. The only way to exploit others is to exploit oneself first—exploit one's abilities by developing them to their fullest extent. Philosophy will thus come first, by which they will learn to govern themselves. Politics will follow—if they still want to enter politics, after practicing philosophy (Foucault Reference Foucault and Hurley1990, 72–73; Alcibiades I 123d; Gorgias 527d; Xenophon, Memorabilia 1.2). In this Socratic argument, the desire for an amoral, instrumental rule over others is transformed into a quasimoral pursuit of excellence. No teleological end or highest good is in sight. An open-ended quest for an unachievable and perhaps unimaginable best or real self (which, Foucault says, “happily, one never attains”; 1997, 137) is the result of purely instrumental assumptions. Crucially, to work on oneself, it is necessary to have useful friends such as Socrates.

Similarly in the Lysis, Socrates dazzles a young boy with an image of what he really wants—power over Athenians and barbarians—only to show him that self-transformation (with Socrates always by his side) is the road to such power. Although Socrates stresses utility with Lysis, elsewhere in the Platonic corpus the same argument emphasizes conquering and dominating. For example, Plato's Laws Book 1 contains the most power emphatic rendition of this Socratic argument, showing the evolution from a Hobbesian-style assumption of undeclared war among all peoples to a subtler war within each person's psyche. Seizing on Kleinias’ assumption that all cities are at war, the Athenian stranger gets him to agree that the same state of war must exist within cities, between their neighborhoods, and within neighborhoods between households (626c–d). Privacy is no barrier to war. Kleinias’ formula captures it best (626d): “all are enemies of all in public, as privately each [is an enemy] of himself.” Similarly, at Lysis 220e, all the phantom friends are “for the sake of an enemy.” In the context of this self-improvement argument, Socrates probably means that this enemy is oneself. One's self as currently constituted is one's own worst enemy, yet it is the enemy for which one nevertheless does everything (cf. Bolotin Reference Bolotin1979, 175). As a result of this internal enmity, “conquering oneself is of all victories the first and best” (Laws 626e). We cannot know for certain whether Plato intended for us to understand this strategy of Socrates’ for refining the amoral assumptions of the power hungry to be only protreptic rhetoric on his part (as appears from Laws 628c–d) or whether we are to understand that Socrates in his more philosophical thinking sometimes entered into such assumptions, thinking apart from foundations. The fact that the Lysis makes the eschewal of a highest good thematic, after Socrates’ rhetoric has already captivated Lysis, would argue that the latter was sometimes the case. Foucault relies on this Socratic argument for his characterization of Greek sexual ethics as self-mastery, wherein the subject “would be ‘stronger than himself’ even in the power he exercised over others” (Foucault Reference Foucault and Hurley1990, 250). He endorses the concept of ascesis in his own name in Foucault (Reference Foucault, Hurley and Rabinow1997, 137): “it's up to us to advance into a homosexual ascesis that would make us work on ourselves and invent . . . a manner of being that is still improbable.” In this way, power remains the basis for everything for Foucault, but it becomes refined.

How necessary was Foucault's foundations-level theorizing about power in order to arrive at this politics of friendship? Mainstream political theorists have often doubted whether power can bear the weight of being an explanatory key for every occasion. The sheer generality and flexibility of Foucault's conception of power, its myriad forms and guises underlying both friendship and politics, raise an epistemological question. Does not a concept defined broadly enough to encompass all human phenomena “lose all useful profile” (Taylor Reference Taylor1985, 175)? The following example of conceptual stretching required by Foucault's foundational critique will show both the long reach and uncertain grasp of foundations-level theorizing in political theory. Just as he finds friendship to be political, Foucault also finds politics to be friendly, claiming that affection is a basis of modern institutions. “The institution is caught in a contradiction; affective intensities traverse it which at one and the same time keep it going and shake it up” (137). He posits that homosexuality was one of the conditions in armies of the nineteenth and early twentieth centuries, especially in the trenches of World War I, on which was predicated soldiers’ ability to follow their captain into danger and flounder in mud, living in close quarters for weeks at a time. He concedes that sexuality is only one source of affection, but it is the affection itself that challenges the institution (or what modern liberals believe institutions to be). No love was supposed to exist in the institution of the modern army, Foucault claims. In place of love, there was “supposed to be only law, rule, or habit.” But the case was quite otherwise, he contends. Such institutions depend on affection for their survival. For Foucault, a kind of civic friendship is inevitable, even under the cold legality of modern institutions.

Yet, even if we assume that affection is only a refinement of power (as when Alcibiades’ will to self-discipline attracts him to Socrates in Foucault's reading), nevertheless affection is refined so much from what we normally think of as power, that Foucault's exposé loses its shocking character. Politics and affection are obviously at different removes from power understood unsubtly as the ability to coerce. Foucault could not base politics on affection unless he could also distinguish the two. Power may generate all things, but then we have not said very much if we show any particular thing to be generated by power. To say, “That, too, is power” gets us no further than saying “That, too, exists.” Foundations-level statements seem too general to have purchase on political arguments. The fact that ancient theorists of civic friendship such as Plato arrive at a similar conclusion about affection in politics from different foundational starting points, while Foucault gets there from no foundations, only highlights the irrelevance, not relevance, of foundations for political theory. At most, we could say that Foucault's attention to foundations helps him expose a blind spot in liberal assumptions: the normative ideal of leaving love and friendship unregulated by the state blends deceptively into an assumption that love and friendship are naturally private or without public ramifications.

The same question whether an overly refined concept of power loses useful profile also dogs Socrates’ analogous concept of utility in the Lysis. Utility is clearly normative for Socrates: it is the standard that he holds up to friendship. Can friendships be useful, or must friendship be an uncalculating, potentially even self-damaging, embrace of one's own? Although he has removed the highest good, Socratic utility still seems to assume the lesser good of self-interest. The notion of using others for oneself loses some of its shocking character when we reflect that, by winning Lysis for himself, Socrates would hope to “use” Lysis’ mind and heart in a shared pursuit of knowledge, in ways that might be eminently useful for Lysis as well as for Socrates. This shared utility arises out the peculiar character of philosophic conversation, however. The good of learning is nonrivalrous in that sharing it with others does not diminish—and may even augment—one's own utility. No such favorable result would necessarily arise out of, say, a sexual relationship that was comparably utilitarian. The thought that people should use friends is plainly liable to abuse. Only because philosophic friends share a love of something higher than each other can they permit themselves to be used. The more ordinary utility relations of citizens, in contrast, would continue to require strict regulation by law to prevent injustice. Citizens, in turn, would be justified in continuing to think of friendship as something above and beyond utility. Philosophic utility stretches the concept of utility. Here, as with Foucault's power, foundations-level theorizing seems to miss the political. What relevance for political philosophy, then, does Socrates see in foundational inquiry?

A PRESOCRATIC ONTOLOGY OF FRIENDSHIP

The remainder of the dialog will imply that Socrates regards inquiry into foundations to be a zetetic, unsettling discussion necessary to but also inimical to human togetherness, which tends to require settled foundations rather than open-ended searching. The conclusion to which the refinement of utility led—the primacy of the enemy over the friend—points us to a second major substitute for ontology: enmity or antagonism, a tool of critique especially associated with Derrida. In this postmodern view, difference or opposition is determinative of being. As in the case of power, the Lysis provides a broader context in which to view this staple of postmodern thought. As various commentators have pointed out, the Lysis has a distinctly presocratic flavor (e.g., Umphrey Reference Umphrey1982). For example, in contrast to his usual procedure of separating kinds or forms, Socrates find a common denominator underlying three types of love—desire, falling in love, and friendship, bringing all three together under the rubric of a single formula (221e3–5). His willingness temporarily to ignore natural or “formal” differences resembles attempts by presocratic natural philosophers to reduce the variety of phenomena to one or a few overarching principles. As the dialogs often do, the Lysis introduces the opinions of natural philosophers early in the conversation. Socrates draws hypotheses such as “Like loves like” and “Opposites attract” from the natural science and the cosmological poetry of the day (214a). Dry desires wet, cold desires hot, and even sharp desires blunt (215e). But instead of ascending from natural speculation, Socrates radicalizes some of their premises to show what is at stake.

Such presocratic accounts implicitly view nature as an undifferentiated mass or heap. Socrates calls the physical theories literally “writings of the wisest . . . about nature and the whole” (peri phuseos te kai tou holou; 214b2–5). If such a whole existed, the diverse forms within the whole could each experience attraction and repulsion very differently. Friendship might be a uniquely human experience, and there would be no reason to reduce friendship to more basic, elemental principles. But the literal translation does not capture how the presocratic writers on nature themselves must have understood it. If “nature and the whole” is construed as a hendiadys, then the two words go together to make one meaning: the whole nature, or “all of nature” (Penner and Rowe Reference Penner and Rowe2005, 74). All nature is susceptible to the same general principles; there is no whole within which the human realm is distinct.

At issue between presocratic natural science and the usual Socratic procedure of dividing by kinds is the question whether an ordered whole, or cosmos, exists. Dividing by kinds implicitly assumes a cosmos, in which diverse species or forms can coexist within the ordered whole. The attempts of presocratic science to regularize all phenomena under a few unifying principles, however, implicitly treat nature as uniform rather than articulated. The diverse species are not eternal but must eventually collapse back into the primal soup (e.g., “All things are water,” “All things are fire”).Footnote ⁸ Everything must be governed by the same principles even when parts temporarily seem to defy them. The most noteworthy ancient exponent of this approach was Empedocles, who regularized all natural motions under the twin laws of Love and Strife, attraction and repulsion (Diels and Kranz Reference Diels and Kranz1951–52). On such natural premises, human loves fall under larger, nonhuman categories: natural philosophers are tempted to conceive of friendship as a highly specialized case of the attractions (or repulsions) that govern the rest of nature. In the Symposium speech of the physician Eryximachus, Plato mimicked Empedoclean ideas, and he made Aristophanes poke gentle fun at Empedoclean natural science in the ensuing speech (185e–193e) in order, in part, to clear the way for Socrates’ more philosophic and religious speech on love. But the Lysis stays much longer at the natural science level and shows no clean break with it. If we assume that nature is uniform, like a heap or pile of identical elements, then no section of a heap is a true part the way a foot or a hand is an organic part of a whole body. From a foot or hand, one can infer something of what the rest of the body is like. In contrast with a heap of sand, the scientist is at liberty to divide it arbitrarily into smaller heaps and call those parts. The only real parts are the grains or elements themselves. For all discussions of aggregates, there is no natural order; the scientist imposes order on what is essentially chaotic.

In this way, presocratic science sowed radical doubt about the phenomenal world. If the apparent kinds or species in the world are not eternal, in what sense are they beings? At least according to various Platonic dialogs, fifth-century Sophists and rhetoricians picked up on this philosophic doubt and employed it to help tear down conventional morality. Socratic rationalism, in Plato's presentation, restored morality, in part by conceiving of nature as a cosmos. Desires orient the natural kinds, within a cosmos, to the good. In contrast, beings in the presocratic view are merely neutral, not mixtures of good and bad, but entirely lacking any goodness or badness in themselves. That is, they are neither good nor bad in a more radical sense than Socrates entertains at 216c–218c. Socrates will go on to consider this more radical neutrality briefly at 220e6–221b6, where it appears as a short slide into ontological evil. Hence, although it is strange to see the Socrates of the Lysis working under presocratic assumptions, his procedure is curiously in keeping both with the instrumental rationality of the first friend argument and with the amorality of his promise to show Hippothales how to “catch” Lysis (205e2; cf. 206b9–c7). The boys are unable to resist arguments in which anything can become anything because no natural limitations exist. In many Platonic dialogs, the Forms provide a way of preserving separate kinds and avoiding further implications of the presocratic assumption that nature does not form a whole. Why Socrates does not depend on the Forms to fulfill this role in the Lysis is a question of great interest. His interlocutors are young and impressionable. The boys certainly do not lack the intelligence to understand the Forms: the argument of the “first friend” contains ideas that are similar in substance to arguments on behalf of the Forms. Instead, Socrates appears to eschew the comfort of the Forms in his eagerness to problematize friendship at any cost. I argue that he does so because philosophy and friendship are so intertwined that the right characterization of the philosophic life must come into play in any deep analysis of friendship.

Socrates’ willingness to dwell on the presocratic worldview brings out its substantive epistemological difficulties, on which Sophists could base paradoxes.Footnote ⁹ How, if the world is really uniform and undifferentiated, do humans make practical distinctions in speech? If the world contained natural kinds, then the intelligibility of their differences to humans might be relatively straightforward. But if those differences are not natural, then the distinctions humans make in practical life must be more willful or conventional than the Platonic Socrates usually allows. This alternative epistemology relies on antagonism or opposition in order to create or impose kinds or differences where none naturally exist. Pitting one bit of the undifferentiated heap against another bit enables the viewer to name one part as distinct from another part. The most startling formulas to this effect in the Lysis are “It is because of the bad that the good is loved” (220b8) and “If the enemy were to go away, [the friend] is no longer, it seems, a friend to us” (220e). Thus, the two different ontologies (kinds vs. heap) are linked to two views of friendship. To say that humans have a nature and that it is sociable implies an ontology of natural kinds. The ontology of the heap, in contrast, gives no guarantee about human nature and its sociability. The opposed ontologies also give rise to two very different epistemologies, the latter of which is prima facie more problematic, in that humans must impose meaning. Instead of decrying this problem and basing a return to the ontology of natural kinds on the need for epistemological ease, Socrates tacitly assumes the alternative epistemology of antagonism, a premise that keeps returning in various guises throughout the rest of the dialog.

Socrates has a motive for embarrassing friendship to the furthest extent possible, namely, to pry Lysis away from his current friends and to confuse the boy about the intentions of his lover. He fails to separate out the natural kinds of attraction, instead grouping sexual desire and passionate love together with friendship, in part so that Lysis will come to believe that his despised lover, Hippothales, is his friend (222a–b, with 207b4–7). The presocratic assumptions further Socrates’ rhetorical case. But this secret with which Socrates titillates the older youths listening to the conversation may also mask or provide cover for a second motive to look at friendship from the most searching perspective possible. If Lysis’ philoi would protect him from Socrates and his philosophizing, no less than they would protect him from all older men, their love begins to look like a positive obstruction to self-improvement. Friends can hold you back. Such a problem with friendship makes Socrates’ own search for a philosophic friend paradoxical. Socrates cannot truly be friends with anyone whose loyalties will obstruct him in the philosophic quest. Such an unfriendly standard required by philosophic friendship explains why Socrates can claim he has never yet found a friend (212a4–7). This is especially intriguing in that conversing with friends would appear to be essential to the philosophic quest (cf. 223b3–8; cf. Nichols Reference Nichols2009). As we argue, these considerations give Socrates a second, deeper motive to entertain the disturbing epistemological premise that antagonism—“enmity”—is necessary to create friends. Friendship is about others who are one's own, while philosophy cannot let preference for one's own elide the otherness of the other. Finding a philosophic friend begins to look like squaring the circle. Most dangerous of all, to the philosopher, would be to mistake one's own for the good, and to make that false good a foundation beyond which inquiry could no longer go. Any foundation is potentially liable to this accusation. Thus, although Socrates’ emphases on utility and negative association may confuse Lysis about Hippothales and homoerotic love, his purposes are not exhausted by the rhetorical situation. Socrates will use the disruptive character of foundational inquiry to show the relation of friendship and philosophy in a truer light. Meanwhile, we can take advantage of Socrates’ foray into negative foundations to see how he might have viewed a major tenet of postmodernism. In political theories based on Derrida's poststructuralism, the enemy is prior to the friend in the sense that struggle or opposition (rather than concord) is what permits us to see distinct beings, like “friend,” in the first place.

DERRIDA'S POLITICS OF FRIENDSHIP

Enmity, the second major postmodern substitute for ontology that we examine, gained its currency mainly from readings of Derrida's poststructuralism. Comparison between the Lysis and Derrida's work can shed new light on the uses and disadvantages of this negative foundation for political theory. Poststructuralism is, broadly speaking, “antiessentialist” in that it does not wish to rely on the assumption that natural essences exist. Beings may well be fluid and therefore technically unknowable. For Derrida, the resulting epistemology goes under the name “différance,” a coinage capturing both deferral and the constitutive, polemical process of difference making (Derrida Reference Derrida and Bass1982, 8–9). Only by making distinctions—pitting one bit of undifferentiated mass against another bit—do languages make meaning. Derrida (Reference Derrida and Collins1997), Politics of Friendship, draws out the oppositional or “us versus them” character of friendship from Plato through Kant and beyond, basing Derrida's own hope for a universal political friendship on the need for a foundations-level overcoming of opposition. Although Derrida's path to his foundational critique ostensibly involves only linguistics (e.g., Derrida Reference Derrida and Spivak1976), its far-reaching ontological consequences seem to implicate ethics and politics as well. Part of the appeal of this ontology is that uncertainty about the way things are prevents theorists from making strong claims about the way things ought to be. From this perspective, strong values claims often find their ground in strong claims about being. Ontology can be used to justify prejudice and domination. But for Derrida, the “things” to which we refer are mental constructs and are hence always collapsing back into unknowing. The creation of meaning is tantamount to the creation of beings; the epistemology and the ontology go together. Nothing as solid as domination could be constructed on such fluidity. Derrida's ethical humility is thus only one variant of the general critique of foundations, namely, that ontological claims always mask value claims. Ethical humility is not the only possible response to this ontological critique: other postmoderns in the wake of Derrida throw off what they see as the façade of neutrality in modern foundations, explicitly embracing negative foundations that contain values they uphold.Footnote ¹⁰

Derrida (Reference Derrida and Collins1997) is digressive and complex, exhibiting a “delight in disorder” that defies summary. One important strand is the privatization of friendship. Derrida sees Graeco-Roman friendships as having been public and political. The Judaeo-Christian tradition begins a privatization of friendship, in particular a turn inward into the isolated self which the Cartesian ego exacerbates (cf. Derrida Reference Derrida1988, Reference Derrida and Collins1997, 290–93). Derrida explicitly demurs from returning to the Graeco-Roman model of political friendship. The repoliticization of friendship that he seeks must occur, if at all, between radically separate, interior selves. The problem with the ancient political friendship found in Aristotle and Cicero, according to Derrida, is that it tended to reduce the Other to the self (Derrida Reference Derrida and Collins1997, 4; cf. Dallmayer Reference Dallmayer2001, 160). As Aristotle famously said, “the friend is another self” (Nicomachean Ethics 9.1166a32). This reduction violates the respect and distance necessary for a true appreciation of otherness. Politically, the moral benefits of civic friendship—such as shame at breaking the law under the gaze of a fellow civic member—violate the autonomy and heteronomy of the subject.

I argue that Derrida's reading of the Lysis—that it assimilates the Other to the self—captures only one important side of the dialog. Plato's Socrates identifies friendship with a love of one's own, but by the time he does so, his focus on utility has already problematized ownness. As we began to show previously, Socrates will jump from the hypothesis that friendship is only a lesser enmity to another possible cause of friendship: we love what is oikeion to us. Although Derrida admits that this word can be translated “suitable,” he prefers to emphasize blood kinship on account of its cognates, house and home (oikos and oikia; Derrida Reference Derrida and Collins1997, 154). Familiarity then becomes Derrida's translation: “the familiarity of the near and the neighbour” (emphasis in original). As we have seen, it is precisely familial love (of parents) that Lysis had trouble explaining with Socrates’ utility hypothesis. If, instead of utility, all love ultimately stems from kinship or familiarity, then family love is immediately explicable. But friendships between strangers who meet and grow to like each other become harder to explain: how did they get to be each other's own? Derrida is willing to stretch the concept of oikeios to include presence, resemblance, and affinity. But these terms imply finding, in the Other, some links to oneself; it is precisely this linkage that Derrida hopes to eliminate from friendship (1997, 155). He asks whether there might be a friendship without sameness, only otherness. The otherness that Derrida emphasizes at the expense of sameness is somewhat in tension with the constant breaking down of same and other in his theory more generally. Although “[t]he value of the oikeiotes dominates the end of the dialogue” (1997, 154), Derrida wonders whether the aporetic character of the Lysis’ ending might not point to some version of this question: can there be a friendship without affinity?Footnote ¹¹ Derrida thus carries the “othering” of the friend to great lengths. In the limit, any similarity, even common humanity, might accuse friends of parochialism. As we have seen, however, Socrates stretches the concept of one's own in much the same way Derrida does. Socrates suggests that something may be “more one's own” (oikeioteron) than blood kin. It is knowledge of utility, for Socrates, that makes others into owns.

Derrida sees “familialism” as the main problem with ancient civic friendship, a problem with distant reflexes even in the Kantian friendship he partially endorses (1997, 92–93, 252–77). Even notions such as “the brotherhood of man” retain a residual sexism and natalism. Derrida argues on behalf of “another politics,” “another democracy,” a “democracy to come,” in which no trace of birth or family remains as a requirement for membership (104, 232, 306). Much evidence could be adduced to show that Derrida is correct about ancient philosophers’ willingness to countenance parochialism or familism in politics. Civic friendship, in ancient theories, was often an extension of family philia-ties to include fellow citizens, so that, for example, all older male citizens would seem like one's father.Footnote ¹² Socrates in the Republic explicitly seeks to make family-style controls on behavior societywide. Aristotle relies still more heavily on the love of one's own, accusing Socrates of watering down civic affection by replacing real family love with an idealized family-style love and by removing the private property with which citizens give aid and do each other favors (Politics 2.1261b8–25, 1263a41–b10–14). Thus, the debate between Plato and Aristotle is not about whether civic friendship relies on the family and on one's own, but about how it relies on them. In this context of ancient civic friendship, Derrida is probably right to think that the Lysis would not permit its normative account of friendship (we ought to seek utility) to stretch its descriptive concept (most of us seek our own) far enough to include the possibility of universal political friendship. The settled foundations of the Republic seem to go together with parochial civic friendship, while the unsettled foundations of the Lysis go together with the seeking and openness of philosophic friendship.

Both Plato and Derrida carry their analyses of friendship down to the ontological level. But while Plato's emphasis on the enemy is consistent with political parochialism, Derrida's ontological antagonism exhibits a tension with his desire for universal friendship. On the one hand, Derrida sees that familialism means bounded groups with insiders and outsiders, the reverse of universal friendship. On the other hand, Derrida's critique of foundations itself privileges opposition and antagonism at the linguistic level. What is the relation between these two types of opposition, political and linguistic? Political conflict might be inevitable if antagonism is written into the very structure of being—or knowing.Footnote ¹³ Derrida's task becomes to inquire whether différance entails political enmity. He thus turns to the philosopher of enmity, Carl Schmitt.

THE CONSTITUTIVE ENEMY

There is an analogy, at least, between poststructuralism and Carl Schmitt's (1976) concept of the political. In poststructural linguistics, words are defined by the “constitutive outside” (i.e., all the words that they are not). Only difference from other signifiers within a closed system gives a signifier its meaning. For Carl Schmitt, the collective identity of insiders is constituted by being not others. The outsiders are real or potential enemies. Only enemies create “the political”; hence, enemies are constitutive. Difference creates identity. Yet, Schmitt understood friends and enemies to be a political dichotomy rather than a logic that permeates all levels of being. In Schmitt (Reference Schmitt and Schwab1976), enemies are not logical antagonisms, but human enemies understood in the most literal sense, with the “real possibility of physical killing.” Schmitt's logic emphatically does not extend the concept of enmity into the ontological realm Derrida and the Lysis inhabit; instead, he distinguishes sharply, for example, between political enemies and the personal or private enemies that would have been a ready logical extension (28–29).

What, though, if Schmitt's primacy of the enemy is only a corroborating instance of the more general rule that opposition creates being? This postmodern possibility would decisively link political analysis with foundational critique. Derrida permits himself to extend Schmitt's ideas past their original political realm, arguing that Schmitt himself in later life came to expand and deepen his definition of the enemy, eventually seeing the enemy less as external and more as an internal facet of the self (Derrida Reference Derrida and Collins1997, 162–65; cf. Abizadeh Reference Abizadeh2005, 57). Although Derrida cites only a few examples, other scholarship bears out Derrida's claim. Changes in successive editions of the Concept of the Political from its earliest appearance show that Schmitt came more and more to conceive of internal, internecine struggles (especially religious strife) as equally constituting the political. Whenever two or more are gathered against a third, the political is with them (Meier Reference Meier and Brainard1998, 33–35). Schmitt also came to realize that his own predilections toward enmity were based on metaphysical commitments. For example, in his Glossarium, Schmitt wrote a new version of Descartes’ foundational statement, Cogito ergo sum, “I think therefore I am.” Schmitt's new version was Distinguo ergo sum, “I make distinctions, therefore I am” (Lilla Reference Lilla2006, 57; Schmitt Reference Schmitt and von Medem1991). Enmity became an increasingly metaphysical concept for Schmitt.

Yet, epistemological antagonism clearly does not simply entail political antagonism. Because only logical opposition is really necessary to create a constitutive outside, the “outsiders” need not be human: disease or racism could be sufficient scourges to create groups opposing them. Furthermore, such inhuman enemies could, in principle, be enough to constitute all humanity together as one universal group in opposition, as Abizadeh (Reference Abizadeh2005) points out. Thus, the tension in Derrida's book between defending particularism and “otherness” against the universal claims of grand narratives, on the one hand, and deconstructing the oppositions that separate humanity into particularisms, on the other hand, is resolved in favor of the grandest aspiration of all: universal civic friendship. His soaring political hope is paradoxical because it is explicitly based on the thinnest and most unattainable ground; it is grounded on or in harmony with an infinitely receding possibility of friendship. This “friendship beyond friendship” is of “featherweight vulnerability” (1997, 218) because it resides only in each person's desire to be heard.Footnote ¹⁴ Apparently, this conception of friendship is so thin that it can offer “no foothold for reversal of any kind, for any dialectical opposition” (218). Thus, the unassailability of his premise is explicitly based on its unattainability. Derrida's new concept would be an “impure” concept of friendship: “outside or on the near side of this oppositional or ‘polemological’ logic,” the logic of friend/enemy and friendship/war (246). Such an impure friendship will entail getting beyond “oppositionality itself, ontological adversity” (249).

Are such foundations-level statements relevant to political analysis? Derrida finds a means of getting beyond ontological adversity in his poststructuralist belief about the fluidity of opposites: their flow away from but also back into one another, which makes it possible to transcend oppositionality. Opposites tend to collapse back together. As Abizadeh (Reference Abizadeh2005, 57) puts it: “difference . . . constitutes identity . . . and so penetrates it . . . that, ultimately, not only is difference internal to the identity itself, but the distinction between what is internal to the identity and what is external to it is always breaking down” (emphasis in original). Differently put, because enemies are constitutive, they are no longer “other” to us, but intimately part of the self. Insiders cannot in good conscience attack outsiders once they have come to realize that the outsiders helped constitute them as what they are (cf. Derrida Reference Derrida and Collins1997, 136). We must embrace the Other who constituted us. This ethical result is debatable: presumably, Schmitt's Cromwell, for example, continued to hate Catholic Spain even while thanking God for this “providential enemy” that made England what it was (1976, 68).

Mainstream theorists may wonder whether Derrida's results are enough to justify delving into foundations in political theory. For example, McCarthy (Reference McCarthy1988) notes how Derrida's “postmetaphysical” thought still “remains at the level of metaphysics,” even though its purpose is “to disrupt and displace it.” He advises Derrida that “a better way to be postmetaphysical in ethics, law, and politics is to stop doing metaphysics, even of a negative sort.” The stretching of concepts inherent in foundational categories causes them to lose political traction. Derrida's resolve to see (almost) everything in terms of oppositions leads him to deconstruct Kant's fraternity but still permits him to cherish the ambitious goal of universal friendship. To permit an explicitly unattainable goal to supplant a more realistic but still quite high goal is to make the best an enemy of the good. In a similar way, we saw how Foucault's resolve to see everything in terms of power enabled him to expose a subtle liberal fallacy that affection is private and, on that basis, to propose a rather ambitious goal for identity group friendship. In each case, we observe a wide swing from the critical mode to the idealistic mode. Foundations-level treatments of politics appear to lack internal discipline because no natural boundaries govern the stretching of concepts. Is there a way of handling foundational analysis that ensures that it illuminates political theory?

The Lysis shows us a model for how to keep foundations-level analysis relevant to social and political problems. Like Derrida, the Socrates of the Lysis chooses to delve into the foundations of friendship in order to find the most minimal or pessimistic account of friendship possible. Derrida, however, brings his foundational critique readymade to friendship. His is a blanket procedure, applying the same foundational analysis to any and all problems. In contrast, Socrates lets the subject matter determine which foundational account to apply, choosing his foundation circumstantially. The presocratic ontology of adversity is not found many other places in the Platonic corpus, but it is particularly fruitful for criticizing friendship. Socrates’ procedure appears to be something like the following. He starts with a question or suspicion: can it be that most friendships blind us philosophically? Then he comes up with a test: how much of the ordinary phenomena of both friendship and philosophy would be saved by assuming that all relations with friends and with wisdom are negative associations rather than positive attractions? He then chooses his ontology accordingly. If enmity and utility clarify the relative standing of friendship and philosophy, then these negative foundations will have done their work even if more traditional foundations turn out to be true in the end. His inquiry into foundations is thus not solely for its own sake, but for the light it can shed on moral and political questions. In his zeteticism, Socrates now follows these presuppositions to their logical extreme, entering a realm into which arguably neither Foucault nor Derrida saw any need to go: to examine what friendship and philosophy would look like under the assumption that all beings are evil.

ONTOLOGICAL EVIL

Fleeing evils turns out to save many of the phenomena of friendship that we experience. Even the attraction we feel toward our friends and their comforting companionship may merely be effects of greater repulsion: ports in a storm, or areas of lower pressure to which we gravitate on account of the evils that beset us. Socrates pursues this argument at an ontological level. He asks not whether this or that being is good, bad, or neutral, but whether all beings in what they have in common—their very being—are good, bad, or neutral. Socrates’ medical analogy had already begun to treat all goods as remedies for illness. What are called goods now become merely lesser evils: they are good only in the sense of distasteful cures like medicine we would not want to take were it not for our dire condition (“as if the good is a drug for the bad”; 220d2–3). The lesser evils keep off the greater evils. Our sickness makes them relatively attractive.

Socrates pursues the possibility that the bad is the cause of friendship by stipulating for the second time a counterfactual “cessation of evils” (220e6–201d2; cf. 220c1–7). He wonders if desires like hunger and thirst would still exist in the absence of evil. Socrates at first throws up his hands and—like many modern theorists in the face of fundamental ontology—exclaims “Or is the question ludicrous, what will exist [if the bad is removed], for who knows?” But then he reasons: a hungry person can be benefited, so it is possible to desire beneficially. Desires have nonevil aspects that might remain in a world where evils ceased to be. He concludes that neutral desires, such as desire for food or drink or even erotic desire, will remain after all evils cease (Socrates silently drops the good; 221b3–6). Erotic desire may be believed to entail friendship (221b7–8); hence, “friends” will continue to exist in the absence of evils.

I argue that, somewhat contrary to appearances, Socrates does briefly suggest that all beings, including especially friends, are best thought of as evils (a position I call ontological malism). To begin with, his premise that erotic desire entails friendship is suspect because it advances Socrates’ seduction rhetoric that Hippothales is Lysis’ friend. Overlooking distinctions between types of love fits poorly with the larger question Socrates is pursuing, whether goods are really lesser evils. If friends and lovers are both lesser evils, it matters which one is less evil relative to the other. For example, it is important for Lysis to find out whether he should flee from a lover like Hippothales into the arms of a friend like Socrates. In addition, Socrates has omitted the possibility that being neither all good nor all bad could mean a mixture of both good and bad elements. Instead, he envisions a pure neutrality in which the bad has no “concern” with the neither good nor bad (221b4–5). As we will see, this pure neutrality goes together with the presocratic ontology, in which antagonism becomes epistemologically necessary. Nevertheless, the meaning of the cessation of evils passage is contested. Penner and Rowe (Reference Penner and Rowe2005) clarify the issue at stake, noting that Socrates’ argument appears to “deprive the good of any content but the absence of bad.” “On such a picture of the good, the bad (i.e., evil) is the only real thing here. . . . The good can be dropped from the ontology.”Footnote ¹⁵ Can Socrates entertain such a perverse ontology? Socrates’ commonsense argument—after he has first thrown up his hands—relies on observation of the world as we know it: hungry people can be both harmed and benefited by different things they eat. The usual conclusion would be that some things are good, some bad, and others neither (cf. 221a7–b3). The familiar ancient view of Book 1 of Aristotle's Nicomachean Ethics (cf. Symposium 204c7–5a3) would then go on to distinguish between instrumental goods and those that are good in themselves as ends. Desires such as a carnivore's desire for meat are not evils or defects, but rather needs that define a carnivore. The human desire for friends could be similar (cf. Pangle Reference Pangle2003, 29–31). This argument assumes we know the needs of natural kinds, including at least humankind. But the Socrates of the Lysis entertains the presocratic ontology in which nature does not form a cosmos with natural species in it. He is willing to think apart from natural kinds in the Lysis. And as we saw previously, it is the absence of a highest end or good that makes all goods instrumental, at the same time as it exposes them to the possibility that they are merely lesser harms. Socrates shows a greater openness or more radical awareness of ignorance here than some other places in the Platonic corpus. Socrates does not have evidence that no highest good exists; what evidence could refute so foundational a thesis? The question arises on what grounds Socrates could base his preference for one ontology over another, if ontologies are themselves the ground-level assumptions.

In view of the fact that Socrates entertains a presocratic “scientific” ontology, could he not have recourse to a neutral view of being in order to avoid the conclusion that all beings are evil? The neutral ontology of presocratic science will enable us to think further about Socrates’ true reasons for accepting or rejecting an ontology. On the neutralist view, things “in themselves” are “neither good nor bad.” Things just are. But as we have seen, the undifferentiated heap of neutral beings envisioned by presocratic science was a short slide away from the inability to perceive distinct beings at all. Neutralism sharpens, rather than answers, the epistemological question, even leading us to privilege antagonism as the fundamental relation, as a way of creating beings at all. That is, ontological neutralism tilts toward ontological malism. Neutralism's inability to hold its own is not a decisive refutation, however, because the truly natural orderings could indeed reside at a more elemental level than most of the beings we perceive. The strongest objection to neutralism is practical and is found at the moral level of the choiceworthiness or unchoiceworthiness of Socrates’ philosophic search. There is no reason for philosophic searching if all being—including being ignorant—is known in advance to be “neither good nor bad.” The philosopher has no incentive to remedy that ignorance if, by acquiring wisdom, he merely moves from one neutral state to a second neutral state. Philosophy would be no more choiceworthy than any other pursuit. The dominant theme in the dialog's final section, the love of one's own, would indeed seem to make all objects of love neutral. What we love are not goods, they are just “ours.” But the love of one's own is inimical to the philosophic quest.

Thus, if Socrates rejects ontological neutralism, the tail of the philosophic life wags the dog of philosophic content. Socrates does not somehow have evidentiary confirmation that being is bad (or good). Rather, because an embrace of neutralism would render unchoiceworthy any continued inquiry into the truth (of neutralism or anything else), Socrates takes it for untrue. Such a practical decision has a decidedly second-tier status: it is very different from theoretical insight into the untruth of neutralism. Socrates’ rejection of neutralism would be an intensely provisional assumption. Apparently, then, Socrates’ litmus test for an ontology would be the same as his litmus test for a friendship: is it supportive of the philosophic life? In seeking an ontology that would clarify the relative standing of friendship and philosophy, neutralism would be interesting if it favored one over the other. But, in fact, it entails no seeking for either friends or wisdom. It might seem that Socratic zeteticism would insist on withholding judgment: keep ontological questions in abeyance until theoretical answers become available, if ever. But can Socrates be neutral toward neutralism if his practical choice of inquiring into it commits him to deny it? At the very least, the worth of the philosophic life must become thematic in Socratic philosophizing if the latter is to explore assumptions that remove the grounds of its own choiceworthiness. The philosophic way of life inevitably becomes philosophy's highest theme.

We recall that Socrates has left out an earlier possibility for the neither good nor bad that would suggest a fifth ontological possibility, namely, that only mixed entities, partly good and partly bad, are real. Goods in the absence of a perfect, highest good might be phantoms, and that objection may be enough to prevent Socrates from considering this fifth possibility here. But Socrates’ conflation of this type of neither good nor bad with mere neutrality calls attention to the omission. In its absence, the only remaining possibility is that only evils are real. The perversity of this doctrine serves notice, in the dialog, that practice rather than theory must decide the orienting questions of ontology. Thus, crucially, Socrates’ transformation of friendship into a negative association is normative rather than descriptive. He does not present evidence that all beings are evil and prove that greater evils drive us into the arms of lesser evils, which we call “friends.” Rather, he implies that we ought to consider all beings evil and that we ought to consider our friends to be lesser evils. For only then would we philosophize with the required urgency. To illustrate phantom friends and the first friend, Socrates adduces an example of a father whose son has drunk hemlock. Ostensibly, the example illustrates the simple hierarchy of means to ends: the father “loves” a cup because it contains the wine he believes is the antidote that can cure his son's poisoning. Yet, hemlock points the reader to Socrates’ own judicial execution as the subtext. The language is strikingly odd: the father literally “makes much of the wine for the sake of believing that his son is worth everything” (219d7–e2). Rather than illustrating means and ends, the father's bizarre friendship for the instrument that might save his son actually wrenches normal hierarchies into unrecognizability. The urgency of the son's approaching demise leads the father into irrationalities, but Socrates’ young interlocutors fail to register that emergency conditions are being foisted on them as the norm (Benardete Reference Benardete, Burger and Davis2000, 222–25). In another emergency context, however, that of Socrates’ death cell, such language makes perfect sense. Socrates sacrifices his own life in order to show the importance he attaches to philosophy. He “makes much of” the hemlock and its curative properties just as the father makes much of the wine.Footnote ¹⁶ Socrates establishes belief in those around him that he holds philosophy to be “worth everything.”

But in what sense does Socrates’ everyday philosophizing have the same urgency as this emergency situation? Socrates may believe that normal conditions anaesthetize us to where loved ones actually stand in our hierarchies. In the emergency situation, the question arises whether the son is the father's “first friend,” and whether the father would sacrifice everything for his son—including, if that would help, his own life. Making much of something for the sake of establishing belief seems to imply theatricality on the father's part. Perhaps the theatricality is also internal, and he wants to convince himself along with everyone else that he is bending every effort on behalf of his son (“for the sake of believing that the son is worth everything”). The father wants to reassure himself and others that he would indeed make the supreme sacrifice. He makes much of the cup in order to shore up his faith and quiet his doubts. The father might suffer from a lie in the soul (Republic 2.382a–b). On this assumption, Socrates lives his life urgently, seeing the real hierarchy all the time: he is always awake where most of us are asleep. Socrates wants to live as though emergency conditions always prevailed while yet keeping his head, which the poisoned boy's father fails to do.

Ontological malism thus makes the best case for philosophy, at least as philosophy relates to friendship. The philosophic quest gains in urgency if our ignorance is not merely neutral but a dire sickness for which philosophy is the painful remedy. The gain in wisdom that might have seemed Socrates’ first and ultimate friend becomes a merely provisional gain—a replacement of one ignorance with a lesser ignorance, and the unfinished character of the philosophic quest comes into better focus. Ontological malism entails a discontent and dissatisfaction with oneself (the enemy for whom we do everything) that is more in keeping with the restless Socratic quest than are the self-images entailed by the other ontologies considered. Socrates would philosophize even if he did not know philosophy was the natural good for humankind (indeed, it is hard to see how this could be known in advance of philosophizing). Socrates would philosophize because he was driven to it by the “evils” or ignorance manifest in the alternative ways of life available (cf. Apology 21c–3b). Crucially, if we believed that friends were lesser evils, we would be less likely to fall into the trap of the love of one's own (oikeion)—the very definition of ordinary friendships that hold us back from philosophizing, such as Lysis’ initial complacency toward his own good looks, his own wealth, his own family, and his own friend Menexenus. Yet, if ontological evil enhances our desire to philosophize, it detracts from our desire to have friends. We do not want to believe that our friendships are negative associations. We could not give our hearts so readily to friends if we believed that they merely hated us less than their nonfriends. What gets isolated by Socrates’ ontology is the quantum of sentimentality in friendship, a feeling that does not need to be present in philosophy to make us philosophize. Socrates’ practical choices thus provide a roundabout way of doing ontology. Socrates has not conclusively refuted the other possible ontologies, nor has he conclusively established that only evils are real. What argument could finish off foundational enquiry? By illustrating how friendship is unsuited to ontological malism while philosophy is admirably suited to it, however, Socrates has opened up a distance between friendship and philosophy. His ontological analysis is geared toward illuminating the social and political phenomena of friendship.