¹ Vlastos, Gregory, “The Individual as Object of Love in Plato,” in his Platonic Studies, 2d ed. (Princeton: Princeton University Press, 1981).Google Scholar

² Though I treat the Nicomachean Ethics (NE) as my primary source for Aristotle's ethical views, I also rely on the Eudemian Ethics (EE) and the Magna Moralia (MM). In doing so, I assume that the MM is a useful source of information about Aristotle's ethical theory, even if Aristotle is not its author. Translations of passages from the NE are from Nicomachean Ethics, trans. Irwin, Terence (Indianapolis: Hackett, 1985)Google Scholar; translations of passages from other works by Aristotle are from the Revised Oxford Translation in The Complete Works of Aristotle, ed. Barnes, Jonathan (Princeton: Princeton University Press, 1984).Google Scholar

³ Whereas the previous passage from the NE requires only that A care about B for B's own sake, this passage from the Rhetoric requires that A care about B for B's own sake, not for A's sake. This contrast between intrinsic concern for another and self-concern is a familiar theme in the Rhetoric's reconstruction of common-sense thinking (cf. 1366a36–b6, 1367a4–6). It is not clear that the Rhetoric's conception of selfless altruism can be reconciled with eudaimonism. However, it is significant that Vlastos typically understands the test of adequacy in terms of altruism, rather than selfless altruism, and that Aristotle rejects the popular contrast, articulated in the Rhetoric, between self-love and altruism (NE IX 8).

⁴ My discussion of classical conceptions of love and friendship draws on parts of my discussion in “Self-Love and Altruism” Social Philosophy and Policy, vol. 14, no. 1 (1997), pp. 122–57Google Scholar; indeed, that article and this one have a common ancestry. There, I appeal to classical conceptions of love and friendship to help motivate a systematic conception of the rational authority of morality that attempts to reconcile other-regarding moral demands and self-interest by modeling interpersonal relations and concern on intrapersonal relations and concern. Here, I focus on the classical conceptions of love and friendship in their own right; I try to examine and assess their adequacy, in light of Vlastos's criticisms, in part by exploring their political implications.

⁵ Vlastos, Gregory, Socrates: Ironist and Moral Philosopher (Ithaca: Cornell University Press, 1991), p. 203.CrossRefGoogle Scholar Cf. Irwin, Terence, Plato's Moral Theory (Oxford: Clarendon Press, 1977), esp. pp. 51–54, 249–80Google Scholar; Irwin, , “Aristippus against Happiness,” The Monist, vol. 74 (1991), pp. 55–82CrossRefGoogle Scholar; and Irwin, , Plato's Ethics (New York: Oxford University Press, 1995), sections 36–37, 142.CrossRefGoogle Scholar

⁶ Plato's dialogues provide our most important evidence about the philosophical thought of Socrates. Though Socrates typically appears as the principal character in Plato's dialogues, the dialogues provide the basis for distinguishing between Socratic and Platonic thought. As is common, I divide Plato's dialogues into different periods or groups. I distinguish (1) early or Socratic dialogues, such as the Apology, Crito, Euthyphro, Laches, Charmides, Euthydemus, and Lysis; (2) transitional dialogues, such as the Protagoras and Gorgias; (3) middle Platonic dialogues, such as the Meno, Phaedo, Republic, Symposium, Phaedrus, Parmenides, and Theaetetus; and (4) later Platonic dialogues, such as the Timaeus, Sophist, Statesman, Philebus, and Laws. As is common, I treat Plato's early dialogues as a more or less accurate picture of Socrates' philosophical thought and the middle and late dialogues as representing mature Platonic thought. I regard the transitional dialogues as marking the emergence of Plato's own philosophical voice through reflection on the commitments in the Socratic dialogues. For other views of these matters, see Vlastos, Gregory, “Socrates,” Proceedings of the British Academy, vol. 74 (1988), pp. 89–111Google Scholar; Vlastos, , Socrates: Ironist and Moral PhilosopherGoogle Scholar, chs. 2–3, esp. pp. 45–47; and Irwin, , Plato's EthicsGoogle Scholar, section 6.

⁷ Hence, Vlastos writes: “‘Love’ is the only English word that is robust and versatile enough to cover [philia]” (“The Individual as Object of Love,” p. 4).Google Scholar Whereas Vlastos is surely right that philia includes intimate associations that we might more readily characterize as love than as friendship, it also includes less intimate associations that we might more readily characterize as friendship than as love. Hence, I am not sure that ‘love’ is versatile enough to do the job by itself.

⁸ This translation from the Lysis is by Wright, J. in The Collected Dialogues of Plato, ed. Hamilton, Edith and Cairns, Huntington (Princeton: Princeton University Press, 1961).Google Scholar Here Socrates is making perfectly general claims about the structure of a person's concerns; concern for persons—oneself or others—is a special case of concern for things.

⁹ Vlastos, , “The Individual as Object of Love,” pp. 8–9.Google Scholar

¹⁰ My account has benefited from those of Kraut, Richard, “Egoism, Love, and Political Office in Plato,” Philosophical Review, vol. 82 (1973), pp. 330–44CrossRefGoogle Scholar; Irwin, , Plato's Moral Theory, pp. 241–42, 267–73Google Scholar; Irwin, , Plato's EthicsGoogle Scholar, ch. 18; and Price, A. W., Love and Friendship in Plato and Aristotle (Oxford: Clarendon Press, 1989), chs. 2–3.Google Scholar

¹¹ This translation from the Symposium is adapted from that of Joyce, Michael in The Collected Dialogues of PlatoGoogle Scholar, ed. Hamilton and Cairns (supra note 8).

¹² For a systematic development of these Platonic commitments about the difference between interpersonal and intrapersonal psychological reproduction, see my “Self-Love and Altruism,” pp. 138–43.Google Scholar

¹³ Insofar as this is true, Aristotle can provide further justification for his assumptions that there are posthumous benefits and harms and that the welfare of one's loved ones and the success of one's projects, after one is dead, are part of a complete good (NE 1100a10–31, 1101a23–30).

¹⁴ My own account of these matters has benefited from those of Irwin, Terence, Aristotle's First Principles (Oxford: Clarendon Press, 1988)Google Scholar, esp. ch. 18; and Price, Love and Friendship in Plato and Aristotle, esp. chs. 4–7.

¹⁵ We might wonder whether Aristotle should think similarity is necessary for friendship. Though he sometimes writes as if the relevant relations among friends must involve similar beliefs and values (1159b3–5, 1161b35, 1162a13, 1165M7, 116687, 1167a23–b10, 1170b16), it is not clear that this is or should be an essential feature of his position, at least insofar as he seeks to model interpersonal relations and concern on intrapersonal relations and concern. For within my own life, I exercise deliberative control and establish psychological connections with my future self when I intentionally modify beliefs, desires, or values, as well as when I maintain them unchanged. It may be that my successive selves will typically be fairly similar; perhaps wholesale and instantaneous psychological change is impossible or at least would involve a substantial change, which I would not survive. But intrapersonal psychological dependence is compatible with significant qualitative change. Our own persistence requires only continuous deliberative control, not fixity of character. If so, then it seems that, in the interpersonal case, Aristotle should allow for friends to be psychologically dissimilar provided the mental states and actions of each exert significant influence on those of the other. Indeed, much of the value of having friends depends upon their being no mere clones of me. See my “Self-Love and Altruism,” pp. 144–45, 148–49.Google Scholar

¹⁶ These issues are distinguished in Cooper, John, “Aristotle on Friendship,” in Essays on Aristotle's Ethics, ed. Rorty, A. (Los Angeles: University of California Press, 1980), pp. 318–19.Google Scholar

¹⁷ It is common to contrast a comprehensive conception of eudaimonia that includes external goods and practical virtues or virtues of character, as well as intellectual virtues, with a strict intellectualist conception that identifies eudaimonia exclusively with contemplation. Whereas it appears that the formal criteria of eudaimonia (especially the completeness requirement and the function argument) introduced in NE Book I support a comprehensive conception and that NE II–IX conforms to this conception, NE X 7–8 appears to endorse strict intellecrualism. My discussion of Aristotle's views about the justification of friendship and the political significance of his account of friendship fits most readily with the comprehensive conception, which I take to be Aristotle's considered conception (whether he holds it consistently or not).

¹⁸ This is another reason Aristotle should not treat psychological similarity as a necessary condition of friendship (see supra note 15).

¹⁹ Is friendship, on this interpretation, a necessary part of eudaimonia? The answer may depend on the sort of modality at issue. Friendship is a part of eudaimonia, on this view, insofar as humans are not individually self-sufficient for a complete deliberative good. There may be conceivable circumstances (e.g., in which humans are omniscient) in which humans are individually self-sufficient for a complete deliberative good and in which an individual's deliberations cannot be improved by interaction and discussion with others. Friendship may not be part of eudaimonia in these circumstances. But these circumstances are remote from our own. Friendship is part of eudaimonia in circumstances, such as our own, of individual deliberative limitations.

²⁰ Vlastos, , “The Individual as Object of Love,” p. 31.Google Scholar

²¹ Ibid., p. 33n.

²² Whiting, Jennifer, “Impersonal Friends,” The Monist, vol. 74 (1991), pp. 3–29.CrossRefGoogle Scholar

²³ Ibid., p. 7.

²⁴ Ibid., p. 9.

²⁵ Ibid., p. 11.

²⁶ I think that this distinction also underlies Cooper's attempt to reconcile Aristotle's claim that friendship, as such, involves concern for the other for the other's own sake and his insistence that pleasure-friendship and advantage-friendship involve concern for another only insofar as the other is pleasant or useful. Cooper's suggestion is that I can value another for his own sake, even if my reasons for doing so are the extrinsic benefits I get from him. See Cooper, “Aristotle on Friendship,” pp. 312–13. I suspect, but will not argue here, that my doubts about Whiting's use of this distinction apply also to Cooper's reconciliation strategy.

²⁷ Whiting, , “Impersonal Friends,” pp. 12–13.Google Scholar

²⁸ For related criticisms of Whiting's impersonal conception of friendship, see Jeske, Diane, “Friendship, Virtue, and Impartiality,” Philosophy and Phenomenological Research, vol. 57 (1997), pp. 51–72.CrossRefGoogle Scholar

²⁹ Whiting, , “Impersonal Friends,” pp. 21–22.Google Scholar

³⁰ Sidgwick, Henry, The Methods of Ethics [ME], 7th ed. (London: Macmillan, 1907).Google Scholar

³¹ In a footnote, Whiting characterizes her justification of special concern as “pragmatic” and claims that it is not “purely instrumental” insofar as adopting this justification is a “minimal condition for the sort of agency we take to b e intrinsically valuable” (“Impersonal Friends,” p. 25, note 3). I am not sure how to understand these claims. Even if the impersonal account of special concern can recognize special concern that is psychologically noninstrumental, Whiting's account of its justification of special concern attaches only extrinsic significance to special relations.

³² Cf. Nagel, Thomas, The View from Nowhere (New York: Oxford University Press, 1986), pp. 152–53.Google Scholar

³³ Whiting claims that her account of friendship is in terms of character-relative, rather than agent-neutral, reasons, inasmuch as only the virtuous are party to virtue-friendship and, hence, have such impersonal reasons to care about their friends (“Impersonal Friends,” p. 11).Google Scholar But even if these reasons apply only to the virtuous, they justify concern in proportion to the virtue or potential for virtue of the beneficiary and make no essential reference to the relationship in which the (virtuous) agent stands to the (virtuous) beneficiary. Moreover, insofar as Aristotle does or can claim that the relevant features of the best sort of friendship extend beyond virtue-friendship, it would be inadvisable to build character-relativity into his justification of friendship.

³⁴ Whiting, , “Impersonal Friends,” p. 4.Google Scholar

³⁵ Ibid., p. 7.

³⁶ Though I agree with Vlastos that an impersonal account of love and friendship is defective, my diagnosis of the defect is somewhat different from his. My complaint that the impersonal account of friendship cannot justify concern for the friend qua friend does not, I think, entail that it cannot justify concern for the friend for the friend's own sake. As I shall argue (Section IX infra), the intrinsic value of a whole is compatible with, and indeed requires, the intrinsic value of its constituent parts. If so, an impersonal account of concern can justify intrinsic concern for another. If my friend is virtuous, or is capable of virtue, then an impersonal concern for all virtuous persons, or all those capable of virtue, justifies concern for her as a constituent part of this impersonal concern. If so, the complaint about the impersonal account of friendship is not that it cannot justify intrinsic concern for those who are friends but that it cannot justify intrinsic concern for them as friends.

³⁷ Broad, C. D., “Self and Others,” in Broad's Critical Essays in Moral Philosophy, ed. Cheney, D. (London: George Allen and Unwin, 1971), esp. pp. 279–80.Google Scholar

³⁸ Ibid., p. 280. As I (would like to) understand him, Broad should say that the altruism that common sense approves is of variable weight, rather than of limited scope.

³⁹ Given the apparent incompatibility of impersonal and eudaimonic conceptions of love and friendship, it is somewhat surprising to find Whiting and Vlastos ascribing impersonal conceptions to Plato and Aristotle. Whiting recognizes the way in which the egocentric conception, which she rejects, appeals to eudaimonism (“Impersonal Friends,” p. 9)Google Scholar; she may see her case for ascribing a generic or impersonal conception of friendship to Aristotle as providing reasons for resisting the common interpretation of Aristotle as a eudaimonist. Vlastos, as we have seen, clearly interprets Plato and Aristotle, as well as Socrates, as eudaimonists, yet he apparently does not see this as an obstacle to ascribing to them an impersonal conception of love and friendship. Vlastos believes that Aristotle cannot consistently claim that the friend values another for the other's own sake and that the friend values his friend's virtue; he may think that adding eudaimonism just adds more inconsistency to the mix. However, Vlastos's attitude is complicated. In accusing Aristotle's conception of friendship of failing to distinguish (a) disinterested affection for the beloved—the desire to promote that person's good—from (b) appreciation of the excellences instantiated by that person, Vlastos comments that “(b), of course, need not be disinterested and could be egoistic” (“The Individual as Object of Love,” p. 33n). Vlastos seems to be saying that there could be egoist as well as impersonal versions of (b): (1) A could value some of B's traits insofar as they are good, or (2) A could value some of B's traits insofar as they are beneficial to A. Presumably, Vlastos thinks that both (b1) and (b2) are incompatible with (a). Perhaps (b2) is more easily reconciled with eudaimonism than (b1). But, as I read Vlastos, (b2) is just a logical, not an interpretive, possibility; Aristotle's version of (b), according to Vlastos, is (b1). If so, Vlastos must think that eudaimonism is inconsistent with an impersonal account of friendship.

⁴⁰ Ramsey, Paul, Basic Christian Ethics (New York: Scribner and Sons, 1950), p. 100Google Scholar, quoted in Whiting, , “Impersonal Friends,” p. 16.Google Scholar

⁴¹ Related doubts about the eudaimonist justification of other-regarding concern are expressed in Section IV of Thomas Hurka's “The Three Faces of Flourishing,” elsewhere in this volume.

⁴² Vlastos is prominent among those who have rejected Irwin's suggestion that Socrates assigns virtue only an instrumental role in happiness. See Vlastos, Gregory, “The Virtuous and the Happy,” Times Literary Supplement, 02 24, 1987Google Scholar (and the subsequent exchanges between Irwin and Vlastos of letters to the editor), and Vlastos, , Socrates: Ironist and Moral Philosopher, esp. pp. 6–10.Google Scholar Interestingly, Vlastos does not address an argument for Socratic instrumentalism that parallels this argument in the Lysis, which he does address. As Irwin notes, the argument of the Lysis might be applied to virtue itself (see Irwin, , Plato's EthicsGoogle Scholar, sections 46–51):

(1) Insofar as x is desired for the sake of some other thing y, then x is only instrumentally valuable (219c–220b).

(2) Virtue is chosen for the sake of happiness.

(3) Hence, virtue is only instrumentally valuable.

But perhaps Socrates can avoid (3) while accepting (1) and (2) if he denies that virtue is distinct from happiness. Furthermore, it might well be claimed that he does deny this insofar as his strong claims about the necessity and sufficiency of virtue for happiness (Apology 28b, 29b–30b, 30d, 32b–c, 36c, 41d–e; Crito 48c–d, 49b) are best explained by the identity of virtue and happiness and insofar as he identifies justice and happiness (Crito 48b) and accepts the unity thesis that the various virtues all pick out one and the same state of an agent's soul (Laches 199e). On this reading, though virtue is chosen for the sake of happiness, it is not chosen for the sake of something else. Unlike Irwin, therefore, I think this gives us a way of avoiding this argument for attributing an instrumentalist attitude toward virtue to Socrates. My own view is that there is a tension between instrumentalist and noninstrumentalist attitudes toward virtue in Socrates. Whereas I agree with Vlastos, against Irwin, that it is hard to square instrumentalism with Socrates' view that virtue is necessary and sufficient for happiness and I believe that the Lysis argument does not commit Socrates to an instrumental attitude toward virtue, I do think that an instrumental attitude toward virtue fits with Socrates' craft analogy. Crafts have specific goals that can be specified prior to and independently of craft skills and methods, which these skills and methods are designed to secure causally (e.g., Laches 185b–e). But then it looks as if virtue ought to have some independent aim—happiness—and as if virtuous actions ought to be those that reliably secure this goal. If so, the craft analogy supports an instrumental attitude toward virtue. It is not, I think, until Book II of the Republic that Plato resolves this tension in Socratic thought in a noninstrumental direction.

⁴³ A somewhat different worry about whether my interpretation of Platonic and Aristotelian conceptions of love and friendship can accommodate intrinsic concern for another is this: Does interpersonal psychological interdependence justify associates (a) in being concerned about the other's overall good, though to a degree or strength that reflects the extent of their interdependence, or (b) in being concerned only about that part of the associate's good corresponding to the nature of their interdependence? Consider, for example, my reasons to be concerned about a casual acquaintance with whom I share a hobby. According to (a), our interaction gives me reasons to care about her overall good, though these reasons are weaker than my reasons to be concerned about the overall good of my family members or intimate friends; according to (b), I have reason to care only about those aspects of her welfare that are related to our shared hobby. Whereas (a) would recognize intrinsic concern for another, (b) might underwrite a version of Vlastos's worry that Plato and Aristotle can deliver concern for another's partial attributes but not for the other for the other's own sake. I believe, but cannot properly argue here, that the correct way to model interpersonal relations and concern on intrapersonal relations and concern vindicates (a), rather than (b). See my “Self-Love and Altruism,” esp. Sections VII and XI.

⁴⁴ Irwin draws attention to this passage and expresses concern about the asymmetrical and passive aspects of Platonic love; see Irwin, , Plato's Moral Theory, p. 269.Google Scholar

⁴⁵ Compare Aristotle's similar reasons for claiming that one cannot be friends with wine or other inanimate things (1155b28–31).

⁴⁶ Plato is not always blind to the active and responsible role the student must play in his own moral education. I assume that Plato accepts the Socratic requirement that knowledge is necessary for virtue, even if he disagrees with Socrates about the sort of definitions required for knowledge. The mini-elenchus with the slave-boy in the Meno (82b–85d) shows how A and B might be unequal and yet how A might love and educate B without B assuming a purely passive role. Moreover, the educational system of the Republic clearly assigns students able to become guardians an active role in their own education (531e–541e). The active role that students must play in their own education is also reflected in Socrates' conception of his own role as philosophical midwife (Theaetetus 148e7–151d3).

⁴⁷ In “The Individual as Object of Love,” Vlastos makes important connections between Platonic love and the political theory in the Republic. However, his misdiagnosis of the flaws in Platonic love infect his diagnosis of the flaws in authoritarian rule. Put briefly, we might say that just as the real flaw in Platonic love is not that the lover fails to have concern for the beloved's own sake but that he has an impoverished conception of the beloved's role in his own well-being, so too the real flaw in authoritarian rule is not that the guardians are not concerned with the well-being of auxiliaries and artisans but that the former have an impoverished conception of the role of deliberative opportunities and responsibilities in the latter's well-being.

⁴⁸ Plato's other political writings say more about nonideal theory. In the Statesman, Plato claims that the rule of law makes for imprecise justice (294b–295e); as a result, he thinks that constitutional government is a second-best option, behind case-by-case rule by some person or group that understands the art of good ruling (300c). Plato distinguishes among constitutions both by the number that rule (one, several, or many) and by whether the rule is wise and is in the interest and secures the consent of the governed (276d7–e6, 291d–292a, 293b6–8, 296d5–297b2, 301a–303b). Whereas monarchy is the best ideal constitution, democracy is the best nonideal constitution (302e6–8, 303a3–b8). However, even democratic government apparently excludes slaves and manual laborers from political participation (289d8–e3, 290a), regulates matters of marriage and procreation (310b–e), and imposes severe penalties for criticism of existing laws and institutions (299c, 300c). The laws describes appropriate laws, institutions, and practices for nonideal theory (853b–c). In such circumstances, Plato thinks that the correct constitution should strike some kind of balance between monarchy and democracy (756e–757a). Although he is careful to insist that not every desirable social practice that the state might try to encourage should be mandated by law (772d–773e, 788a, 822d–823a, 885c8–d2), he nonetheless endorses a number of legal restrictions on marriage, procreation, association, expression, worship, a nd consumption (631d–632c, 656c, 662b, 721a–d, 746a, 780a, 783d–785a, 841d–e, 910b–d, 929e–930d, 950d).

⁴⁹ See Popper, Karl, The Open Society and Its Enemies, 4th ed., vol. 1 (Princeton: Princeton University Press, 1963).Google Scholar

⁵⁰ It is somewhat puzzling that the assumption that deliberative capacities are distributed unequally—on which the justification of authoritarian rule rests—should be part of ideal theory. Wouldn't it be more ideal if everyone were a philosopher? Presumably, Plato could have no objection to democratic governance in a world of philosophers. Perhaps he thinks that the appropriate constitution for this possible world is not worth discussing in ideal theory, because he regards that world as too fanciful. Finding a world with a few philosophers, he might think, is improbable enough (Republic 428e, 491a–b, 503b).

⁵¹ Cf. Irwin, Terence, “Socratic Inquiry and Politics” Ethics, vol. 96 (1986), p. 414.Google Scholar Insofar as the Protagoras is a transitional or Socratic dialogue, it would be problematic to appeal to its democratic commitments as evidence of democratic commitments in mature Platonic thought. But we can appeal to these claims in the Protagoras to identify a model, recognized by Plato, in which the operation of moral expertise takes place within a system of democratic governance.

⁵² My account of Aristotle's political theory has benefited from those of Barker, Ernest, The Political Thought of Plato and Aristotle (New York: Dover, 1959)Google Scholar, chs. 5–11; Irwin, Terence, Aristotle's First PrinciplesGoogle Scholar, ch. 19; Irwin, , “The Good of Political Activity,” in Aristoteles' Politik, ed. Patzig, G. (Göttingen: Vandenhoeck and Ruprecht, 1990)Google Scholar; and Miller, Fred D. Jr., Nature, justice, and Rights in Aristotle's “Politics” (Oxford: Clarendon Press, 1995)Google Scholar

⁵³ Although Aristotle allows that the best ideal constitution would assign rule to the ultra-virtuous—whether the ultra-virtuous are many or one (1288a34–36)—his considered claim seems to be that the best ideal constitution is a monarchy. But this invites the question why an ideal world would not be one in which everyone was ultra-virtuous and in which the constitution was, as a result, democratic. It may be that Aristotle's reason for not taking this possibility seriously within ideal theory is that it is just too fanciful; it is much more improbable that everyone would be virtuous than it is that one would be (cf. Rhetoric 1354a33–b1).Google Scholar

⁵⁴ Irwin usefully discusses Aristotle's reliance on this assumption in “The Good of Political Activity,” section 11.