Jump to ContentJump to Main Navigation
Wounded HeroesVulnerability as a Virtue in Ancient Greek Literature and Philosophy$

Marina Berzins McCoy

Print publication date: 2013

Print ISBN-13: 9780199672783

Published to Oxford Scholarship Online: January 2014

DOI: 10.1093/acprof:oso/9780199672783.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. date: 10 August 2020

Friendship and Moral Failure in Aristotle’s Ethics

Friendship and Moral Failure in Aristotle’s Ethics

(p.140) (p.141) 6 Friendship and Moral Failure in Aristotle’s Ethics
Wounded Heroes

Marina Berzins McCoy

Oxford University Press

Abstract and Keywords

Chapter six carefully analyses Aristotle’s discussion of the loss of a friendship of virtue in which one’s friend has become bad. While Aristotle argues that a virtuous man will only love what is truly good—and so not a friend turned bad—his discussion of the loss of such a friend displays how friendship itself makes us vulnerable. While we need friends to practise the virtues and to reflect ourselves back to us, a friend is also valued in himself, as ‘another self’. The loss of friendship entails a concomitant loss of an aspect of one’s own self. Aristotle suggests that such friendships can be retained to some extent in memory, but even memory retains the pain suffered in loss as well as the goods of the friendship. Aristotle also posits the relative limits of not only friends, but also family, civic institutions, and laws in rehabilitating those who have fallen away from virtue.

Keywords:   Aristotle, Nicomachean Ethics, friendship, virtue, vice, loss, self, otherness, self-sufficiency, love, memory, soul, narrative, family, law, community, politics

Aristotle’s account of friendship (φιλία‎) extends over nearly two books of the Nicomachean Ethics, granting friendship a central place in his account of an ethical life. Aristotle writes, ‘No one would choose to live without friends even if he had all the other goods’ (N. Ethics 1155a).1 Whatever value Aristotle grants to the necessity of friendship for exercise of the moral and political virtues, his account of friendship is not limited to such necessity. Instead, we find Aristotle speaking of a true friend as a refuge in times of trouble; as a person whom we can trust; as one who knows one’s self more deeply than anyone else; and in the best cases, even as ‘another self’. Aristotle opens his account of φιλία‎ in Book VIII with a brief statement of its essentiality. He does not require that extensive argument be given for friendship’s centrality to human nature, noting only that φιλία‎ is so basic to both our earliest bonds to our mothers and our most civilized and highly developed political institutions that its natural origin cannot be denied. Aristotle grants friendship the status as the noblest of human relationships, a relationship of not only necessity, but also honour.

Working out the details as to why Aristotle grants such centrality to friendship has been something of a puzzle for commentators, for Aristotle clearly claims that the virtuous man is by definition self-sufficient. Commentators have wondered about the relationship (p.142) between Aristotle’s claim to the self-sufficiency of virtue, and his remarks about the centrality of friendship to a life well lived, given the contingencies involved in friendship. Cooper, for example, asserts that if the virtuous man does not need others in order to be virtuous, we might have little reason yet to understand why the virtuous man would even choose to make friendships of virtue. But Aristotle clearly makes such friendships essential to the flourishing of the human person. Cooper finds the answer in Aristotle’s claim that friends help to increase our self-knowledge, for they show us a reflection of our own lives so that we might be able to study (θεωρεῖν‎) it in another (Magna Moralia 1213a). Such self-knowledge is part of human flourishing.2 Nussbaum persuasively argues that relational goods such as friendship are both essential to human flourishing and yet inherently vulnerable to a variety of contingent circumstances that can diminish such relationships and the happiness that they bring. Friendship is central to human life, both because we are political beings by nature whose humanity is bound up with relationality, and because human virtues such as justice and generosity are by nature connected to human neediness.3 Generosity exists only as a virtue because we have need of one another’s generosity as human beings who cannot flourish alone. Yet, a wide variety of contingencies make the development of such friendships difficult, such as the luck involved in finding another of high character and aspiration, or the ability to trust in others. Friendships can be lost to departures, distrust, and inevitably to the death of the other friend.4

Relatively little attention has been paid to Aristotle’s discussion of how to respond to the moral failure of a friend. To the extent that this topic has been addressed, commentators have mostly focused on his discussion as an example of Aristotle’s over-reliance on good character as central to true friendship. Some, for example, object that Aristotle requires far too much of friendships by eliminating the possibility of friendship with those who lack virtue. He even argues for the abandonment of a friend who no longer possesses virtue. However much it may be the case that Aristotle is not generous enough in his responsiveness to the good of loving the other in his imperfection, I would (p.143) argue that Aristotle expresses a remarkable care and sensitivity for the genuine loss of such friendships. Aristotle’s sensitive discussion of moral failures among friends deepens our understanding of his openness to the difficulties of human vulnerability in relationships. Indeed, his understanding of friendship accounts for a certain degree of contingency and even love of other individuals in their particularity that is not often emphasized in commentaries.5

I shall focus on Aristotle’s account of how one properly reacts to the loss of a friend who has turned bad, in order to explore more deeply Aristotle’s sense in which friendship extends mutual dependence and even increases our vulnerability to one another.6 Such vulnerability moderates somewhat his claims that the virtuous man is self-sufficient. While the good man is self-sufficient in virtue, his friendships not only are necessary for his full flourishing as a good person, but also leave him with genuine loss when such friendships come to an end. Aristotle’s care for the self-sufficiency of virtue is also balanced by a realistic and sensitive acknowledgement of the virtuous man’s vulnerability to loss in engaging in intimate relationships. Such loss includes not only the loss of another with whom to spend time and to exercise the virtues, but also the loss of the particular and irreplaceable nature of the friend as an individual.

I next argue that Aristotle’s recognition of the relative limits of close friendships, family relationships, and even civic laws suggests his acknowledgement of human limits and the need to care for one another in the midst of human finitude. Precisely because no single set of relationships—including friendships of virtue—can provide for all human needs all the time, Aristotle emphasizes the need for multiple sources of care in community. We find in his discussion of friendship, then, a readiness to acknowledge human limits in guaranteeing εὐδαιμονία‎. Political communities must develop multiple forms of care if they are to be successful in developing good citizens.

(p.144) I

Friendship plays an important, though limited, role in ongoing moral development. Aristotle grounds all forms of friendship in our most basic human dependencies: ‘[T]he young need it to keep from error. The old need it to care for them and support the actions that fail because of weakness. And those in their prime need it to do fine actions; for when “two go together … ” they are more capable of understanding and acting’ (N. Ethics 1115a). While Aristotle goes on to argue that the best forms of φιλία‎ occur between two virtuous individuals, its origin lies in human weakness. Power, honour, and prosperity are all goods that are subject to risk, and the more that one possesses such goods, the greater the need when they are taken away. Childhood and old age, both times of weakness, are balanced by the strength of friendship. Even in the best part of life, with the fewest weaknesses, friendship is required: not only to have someone with whom to practise the virtues, but also someone with whom to discover how to live them.

Aristotle’s above use of the proverbial ‘two go together’ refers to the events of Homer’s Iliad in which Odysseus and Diomedes plan to spy on the Trojans to see how far the Trojans have advanced into Greek territory. Their mission takes place during a time of great Greek military risk, and so is especially fearful. Diomedes asks for a companion who might accompany them with the idea that two people are better suited to discover the truth, since an individual mind does not reach as far as two together. Socrates, in the Protagoras, quotes the same passage: ‘Protagoras, please don’t think that I have any other desire in our discussion than to examine the things that are difficulties for me each time. For I think that there is a great deal in what Homer says: “When two go together, and one sees it before the other.” For somehow we human beings are all more resourceful (εὐπορώτεροι‎) this way, in each deed, word, or thought; but if someone observes something alone, he has to go around searching until he discovers somebody to whom he can show it and who will confirm it’ (Prot. 348c–d).

Aristotle takes up the theme of ‘resourcefulness’ and extends it to the realm of friendship. In friendship, an individual who lacks something finds the means he needs to accomplish good action with the presence of another alongside of him. A friend provides not only concrete resources to accomplish ends (means for the elderly to (p.145) survive, or good advice for the young); friendship also seems to enhance one’s own internal resources. When Aristotle suggests that friends aid in both ‘thinking and acting’, the implication is that in the presence of a friend, one’s own abilities are enhanced through the shared experience of pursuing a goal alongside another. For example, the presence of a courageous friend in the heat of battle not only protects another from bodily harm, but ideally also inspires the other to persevere and to continue to draw upon his own resources for courage in the face of a threat. Friendship provides for human need in a quite different way from material resources. While food satiates hunger and weapons preserve a threatened man’s welfare, friendship does not merely fill a need. Instead, friendship cultivates the development of one’s capacities. Virtues such as generosity, courage, and wittiness require others as models for learning how to cultivate and live such virtues in particular situations. But even once such virtues are firmly rooted in character, we need others with whom to be generous, courageous, or witty. The ongoing existence of such virtues as active parts of our lives requires others; nor can courage be a meaningful part of life unless my courage is exercised on behalf of others. To this extent, one’s own happiness is fundamentally intertwined with the presence and shared activity of other friends, whose existence allows my own excellence (and also their excellence) to flourish.7

Nussbaum has argued for the centrality of luck or τύχη‎ in Aristotle’s ethical thought. While the life of virtue is stable and has some resistance to the experience of unlucky circumstances, this resistance is not endless. For example, one might say that a person in an irreversible coma as a result of a terrible accident is unlucky and unhappy. This is not to say that he or she lacks value, only that we can say there is something fundamentally missing from an unconscious life that makes it ‘unhappy’, or at least significantly unhappier than a life of vibrant activity (ἐνέργεια‎). However, many causes of such (p.146) unhappiness are beyond our control. Aristotle, in particular, was concerned with large upheavals in the community, such as those found in enslavement after lost wars, or political turmoil within one’s own community.8

Friendship, too, is subject to luck, insofar as many factors can limit or end friendships. Aristotle emphasizes that friendships of virtue are less likely to come to an end than friendships of utility or pleasure, which remain only as long as the usefulness or the pleasure of the friend is of benefit. If two people are friends as a result of utility, for example, work together as study partners for a university class, and then the class terminates, the friendship itself must terminate, or else the reason for its continued existence must transform. The friends in question might find that they also share a common love of sports, and so continue in their friendship now as friends of pleasure. Or they may be in another class and so remain friends of utility whose primary relationship stems from the human dependence upon others to learn and to grow intellectually. Friendships of pleasure and of utility are less stable and lasting than other forms of friendship because the raison d’être for such a friendship is likely to change. To this extent, they are incomplete forms of friendship.9

Aristotle understands friendships of virtue to be significantly less likely to change or to end, insofar as the reason for their existence—the pursuit of a life of moral excellence—does not terminate so long as both persons live. Moral virtue, by its definition, is a stable and abiding character, a characteristic of soul that takes much time to develop, but once established is unlikely to be lost. Nonetheless, Aristotle grapples with the possibility of a friendship in which one friend is faced with the moral failure of the other friend, a moral failure that tests the very resilience of friendship. Aristotle’s explanation of the difficulties with deciding how philosophically to resolve this problem exhibits the complexity and nuance of his understanding of human dependency. Indeed, despite his claim that the best form of friendship displays ‘complete’ virtue, Aristotle exhibits a tremendous sensitivity to human moral weakness in his account of friendship.

Aristotle’s notion of friendship is grounded in the premiss that human beings love only what is lovable, that is, that which is worthy (p.147) of love or appears to be so: ‘For, it seems, not everything is loved but [only] what is lovable, and this is either good or pleasant or useful’ (N. Ethics 1155b). The familiar typology of three types of friendship categorizes friendship according to what is loved about the friend, that is, whether he is useful to one’s self, gives pleasure in being a friend, or is a good human being. While the first two forms of friendship are forms of love in which the object of friendship is loved for the sake of what is good for the one who loves; a friend who makes another laugh is pleasant and so is loved for how the character of wittiness produces a feeling of pleasure in oneself. In the case of perfect friendship, however, each friend loves the other because he is good—not only good for one’s self, but also good in himself: ‘Now those who wish goods to their friend for the friend’s own sake are friends most of all; for they have this attitude because of the friend himself, not coincidentally. Hence, these people’s friendship lasts as long as they are good; and virtue is enduring’ (N. Ethics 1156b). Here Aristotle asserts an identity between the good of the friend for one’s self and for the sake of the other. The good of the friend is not simply ‘added on’ to the benefit that one receives in friendship. Rather, his goodness is essential both to him and to why he is loved and lovable, namely, for himself. Although Aristotle emphasizes that virtuous friends are both pleasant and beneficial, what is loved in the virtuous friend is his selfhood, which in a deep way is constitutive of the human person.10

Unlike later models of ethical behaviour, such as Kant’s ethical theory, which locates the good of the self in a single faculty such as the will, Aristotle’s understanding of what it means to be good permeates the whole of human nature. Book I, particularly Chapter 7, lays out (p.148) the Aristotelian model of excellence as grounded in the fulfilment of human nature. Just as excellence in a plant lies in the good functioning of its nutrition and growth, and animal excellence lies in sense perception, to be a person of virtue requires that one exercise one’s work, or ἔργον‎, one’s activity, in a way that is fulfilling of a distinctively human existence. To this extent, perfect friendship possesses a complete harmony of love of the other for his virtue, but also for himself, for the two are identical for Aristotle: to be good is to be most fully and authentically human. Given that the friend who loves is also a person of virtue, he loves what is good by his own nature. As Annas has argued, Aristotle argues that to be φίλαυτος‎ (self-loving) is to be virtuous, insofar as the man who is becoming virtuous comes to love his true (i.e. virtuous) self more and more, and to love less virtuous aspects of himself less and less.11 Thus, in friendships of virtue, there is a tight unity between the good; the friend’s good; and the fulfilment of one’s own desires.

For this reason, wicked men cannot be friends for long, since their love lacks both steadfastness and completion. Aristotle attributes the inconstancy of friendship between wicked persons to an internal lack of steadfastness of desire, ‘since they do not even remain similar to what they were’, but are changeable in what sorts of objects of desire are pursued (N. Ethics 1159b). In such cases, there can be no unity of the fulfilment of one’s own aims and objectives and the true love of another for himself, for the wicked man by definition does not love that which is truly good, but only that which is apparently good in his own eyes. Even if two friends share a common aim, there are constant threats to the persistence of such friendship. Take the example of two wicked men who work together in order to rob a bank. Those who lack virtue are likely to betray one another, for they do not possess constancy in courage, generosity, pursuit of honour, or the other virtues. In addition, the friendship exists for the sake of the achievement of some other good external to the friends, not the friend’s own good. Once the money has been stolen, one easily imagines that one thief may wish to ‘take the money and run’ now that the friend does not continue to be useful. Even those who delight in one another’s wickedness are not likely to remain friends for long, for such delight will be short-lived. Fundamentally, Aristotle’s conception of the virtue is that it alone provides for lasting and complete pleasure. While (p.149) superficially other kinds of goods might provide short-term pleasure, only ‘real’ goods allow for the lasting pleasure that the human being seeks. While Aristotle does not reduce pleasure to the good, he does regard pleasure as that which ‘completes’ the good.12 Thus, the vicious who cannot participate in the highest forms of friendship also lack the most significant and lasting pleasures of life.

However, a deeper set of issues arise when only one friend in a pair of apparently virtuous friends turns out not to be good. Aristotle asks what he takes to be a difficult, but common, question: ‘But if we accept a friend as a good person, and then he becomes vicious, and seems so, should we still love him?’ (N. Ethics 1165b). Aristotle’s answer is complex and exhibits especially well the difficulty with the problem of moral education, given human vulnerability to harm from those who lack virtue. Aristotle’s most basic answer is that ‘what is bad is not lovable’, but only that which is good (N. Ethics 1165b). Philosophically, Aristotle understands the human person to be capable only of loving what he understands or opines to be good. What is bad is not lovable by us, unless we mistakenly believe it to be good. Therefore, it is not possible to love an evil man, if one is good, for one will love the good and not the evil, according to one’s own character. Nonetheless, in his additional remarks, Aristotle expresses hesitancy as to whether such careful and neat philosophical categories can capture the full extent of the experience of choosing whether to continue in friendship with a friend who repeatedly fails morally.

One consideration is whether one might have an obligation to love someone who is not good. Aristotle argues that if a friend has become bad, but not incurably so, we have even more of a duty to assist him than before his failure: ‘If someone can be set right, we should try harder to rescue his character than his property, in so far as character is both better and more proper to friendship’ (N. Ethics 1165b). Indeed, Aristotle suggests that the true friend will only abandon the relationship when it becomes clear to him that there is no possibility of its restoration, in which case it is understandable if he gives it up. However, such a loss is understood to be difficult; even in cases where the virtue of one person leaves behind the other, Aristotle says that in (p.150) memory, a kind of φιλία‎ continues, as a friend will keep a ‘memory of the familiarity they had’, as a sign of what the friendship once meant to each person (N. Ethics 1165b). Aristotle’s statement here is quite important, for it shows that the intimacy of the friendship is itself a good, and not only the qualities that the friend possesses or once possessed. Memories of past friends continue to have significance for us because of the intimate connection enjoyed, as well as for the particular qualities that our friends had.

Aristotle here points to the interdependency of happiness in the context of friendships of virtue. Friendships of trust and shared living, in which another’s self is the primary object of love, affect the development of the virtue of the one who loves (and who is loved). Positively, this means that two friends of virtue can support one another in living well and reinforcing their mutual love of what is virtuous, and their ongoing virtuous action as exhibited in moral and intellectual virtues.13 A wicked person can also influence another’s character, though perhaps is less likely to do so in the case of a person of fully formed virtue, who will not be inclined to alter since he is (by definition) of a firm and unchanging character.

Aristotle also denies the possibility of loving a bad man in his badness. If one means that one friend ought to love an evil friend, even in his evil, no such duty can be possible, for it is not a duty to love evil. Indeed, there is not a simple absence of such an obligation, but in fact a positive danger that might arise were one to love even the wickedness in a wicked man. Aristotle asserts that ‘we ought neither to love what is bad nor to become similar to a bad person, and we have said that similar is friend to similar (N. Ethics 1165b).’ That is, insofar as we love what is evil, the evil object of love does not only reflect an ongoing distortion of desire. To love an evil object can also seemingly become a cause of further distortion of desire in oneself. If a man loves a wicked person, Aristotle implies, the love of a wicked object makes the lover like that which he loves, that is, potentially more akin to the wicked man whom he loves. Thus, it might be necessary for the sake of one’s own flourishing to end a relationship with another who threatens it—even a former friend.

And yet, Aristotle also takes pains to emphasize the genuine experience of loss in the end of such a friendship. In lesser forms (p.151) of friendship, because less of the self has been offered to the friendship, the loss at its termination is accordingly less deeply felt.14 In the case of a friendship of virtue that dissolves when one of the friends is seen to lack virtue, one has lost not only someone of pleasure or of use, but ‘another self’.15 Here, the philosophical tensions inherent in the term ‘another self’ become especially apparent. As Stern-Gillett has argued, the idea of a friend as another self raises difficulties with the notion of selfhood.16 How to characterize the idea of the friend as also one’s self is notoriously difficult. In the Magna Moralia, the friend is presented as like a mirror, which reflects back to us something about one’s self, since one cannot see the self without it (1213a). However, if the ‘other self’ is only a mirror that reflects one’s own characteristics back to one’s self, then the other seems only to be loved for selfish reasons, and not ‘seen’ at all in his otherness and independence, as the seat of his own moral actions and choices. If the other friend is truly ‘another’, then there must be a sense of separateness of each friend. Distinctness of some sort is required for relations to exist, for without two independent persons, no relation between them can come to be, only the use (or misuse) of one by another.

One way to resolve this difficulty is to focus on the friends’ shared activity and shared goals as constitutive of their bond. As two people walking on a path to a common destination are unified by their common goal and method of attaining that goal, friends of virtue participate in shared activity (ἐνέργεια‎) as part of the pursuit of a good life.17 Still, solely attending to their shared activity also misses something of the intimacy that Aristotle wishes to communicate through the phrase ‘other self’: a friend who is somehow both ‘other’ and yet also so intimately loved that the boundaries between care for self, and care for other, are not fixed and firm. Thus, one friend wishes to assist another, Aristotle says, not only for one’s own sake, but rather for his sake (Rhetoric 1380b–1381a).18 Moreover, the friends’ love for one another is not simply a reflection of their love for the good. Aristotle emphasizes that no one would wish for his friend to become (p.152) a god, but only to be as happy and virtuous a human being as a human being can be (N. Ethics 1159a).19 Thus, Aristotle suggests that one loves a friend not as an embodiment of perfection, but as a human being, even when one takes into account the limited nature of another’s humanity.

Annas has argued that one strong motivation for the love of another in his goodness is that it allows us to contemplate (θεορεῖν‎) the good in another, since to do so in another is easier than to do so in oneself (N. Ethics 1168b2–4).20 One difficulty with such a view is that it seems to make the beloved lovable only insofar as he ‘shows’ the good to us. This would seem, then, to make the good, but not the friend himself the real object of love. Annas suggests limitedness in Aristotle’s approach to friendship on these grounds. She writes, ‘Aristotle is wrongly insisting that friendship involves approval of and respect for the friend’s character, and ignoring the irrational element in friendship, which can lead us to like and love people of whom we strongly disapprove. And to the extent that he runs these ideas together, he could be said not to have fully attained the notion of loving someone truly as an individual, rather than as a bearer of desired qualities.’21 However, Aristotle’s discussion of the loss of a friend who has become bad shows that there is something much more in the loss than the grief associated with, for example, fewer opportunities to contemplate the good for oneself. For Aristotle’s claim that one ought to come to the assistance of a friend more than a non-friend because of the intimacy of the friendship conveys the significance of intimacy as a good of friendship. The friend is not loved for his qualities alone, but also as one with whom one has shared past activity and with whom one has shared oneself.

Stern-Gillett argues that the ‘other self’ loved in a perfect friendship is limited to the rational self. Her argument relies upon the development of the notion that practical virtue is the ‘hub’ around which the rest of the self is formed; to the extent that practical wisdom is responsible for the ordering and arrangement of the person’s self, the Aristotelian ‘self’ is such a rational faculty.22 She writes, ‘Since human beings are to be identified with their nous, those who fail to be rational will correspondingly fail to be fully themselves. To that (p.153) extent, they cannot be loved in themselves.’23 And yet, we find this claim difficult to maintain in the light of Aristotle’s description of the centrality of memory in cases of friendships lost to moral failure. As I will argue below, Aristotle’s account of memory allows for a kind of retention of the friendship. The presence of a friend in memory and the development of a narrative about the meaning of the lost friendship preserves the care present in the friendship, as well as retaining elements of the vulnerability therein.


Aristotle’s description of the centrality of memory as a mechanism for maintaining a connection to a lost friend, even after the friendship has been dissolved, suggests a different object of love from the friend’s virtue alone. Indeed, that Aristotle finds a need for a coping mechanism at all suggests that the love of a friend who was once good, but now is not, involves much more than a mere intellectual assent to the friend’s good, or a regret about the loss of an abstract good (for example, regretting the mere loss of having around oneself a good friend). Memory becomes an important faculty for dealing with such losses because such losses are not limited to the loss of an abstract good, nor even to a disinterested regret on behalf of the friend (e.g. sorrow that he no longer has the good for himself). Instead, in the loss of ‘another self’ one finds the loss of some portion of one’s own selfhood.

What is beloved in the memory of the lost friendship is not only the recollection of how the friend once exhibited virtue, though this might well be part of such a memory. Nor is there simple regret of the contrast between the past good and the present lack of such a good in the friend, though again this sense of loss might be part of such a memory. Instead, Aristotle claims that the memory of the past experiences of the self and one’s lost friend are to form a part of one’s present love for the friend. Why might such care for memory be of such importance?

One reason might be that the memory of the friend’s past goodness continues to inform the present attitude toward the friend as a greater whole than his current character and lived action seem to embody. (p.154) While others might regard the friend’s current activity—for example, bad behaviour in the midst of an addiction—to be the whole of the friend’s selfhood, the Aristotelian emphasis on memory serves to connect his current deficiencies to a larger picture of the friend in his wholeness. While those who lack deep knowledge of the friend might regard his current activity as indicative of his entire nature, the former friend of virtue can understand a larger context and may even be able to contextualize the friend’s shift of character in terms of a larger narrative that is lacking for those who do not know the friend well. For example, a friend who behaves poorly as a result of a reaction to a traumatic experience (as in a wartime trauma), but who has been known to be brave or generous in other circumstances, is not reducible to his current behaviour. His life as a whole communicates a different reality from what his current activity or character seems to do. Memory allows for the possibility of an overarching narrative about the friend that incorporates specific moments in the person’s life into a larger meaningful whole.

Likewise, the indeterminate future makes an evaluation of the goodness of a person’s life difficult to evaluate once and for all, although a virtuous man’s character is more likely to be stable and consistent (N. Ethics 1101a–b). Just as past exemplary actions of a friend’s once-good character make current judgements of his worth incomplete, so too does the friend’s current bad behaviour not limit his future behaviour entirely. Memory of the friend provides a greater context for a more accurate, and more generous, understanding of the lost friend’s worth and the possibility of his future self’s virtue.

The presence of memory as significant in Aristotle’s account of lost friendship also suggests that the particularity of the friend is crucial in the care experienced after loss. Memory itself is not a faculty of nous, Aristotle argues in ‘On Memory and Reminiscence’; instead, memory essentially belongs to the faculty of sense perception. The person who, in a sense, continues in friendship with his friend through the faculty of memory relies on his prior sense perceptions, and not only upon an intellectual judgement that his friend was good or could be good again. If memory focuses upon sense perception, then one would expect Aristotle to include among such memories of friendship particular contingent experiences of the friend—experiences that were specific to the pair, and not only memories of general characteristics that the friend once possessed. For example, there is a significant difference between remembering abstractly that my friend once used (p.155) to be ‘generous’, but now has become rather closed in upon himself, and recollecting particular moments of generosity that I witnessed him exhibit, such as when he offered to buy a meal for an individual who lacked the means. An old shared joke takes on a distinct and particular value among friends that exceeds the way in which the joke exemplifies the virtue of wittiness. Such shared experiences are ‘excessive’; that is, their value exceeds the way in which the event exhibits an abstract judgement about the good. Instead, such experiences are constitutive of the goods shared in friendship, which are concrete and lived experiences known through sense perception and recollected through memory’s revitalizations of such experiences.

Indeed, memory and narrative together tend to reshape and to reformulate the value of such experiences: a joke told many times among friends, or word play understood only among intimates who understand the reference, has a deeper value than a similar joke outside of the context of its presentation in friendship. The memory of such witty moments becomes a reminder of the past bond of friendship and of the particular moments shared in friendship. Thus, even if one’s current love for a friend diminishes when a friendship ceases, nonetheless, not only his past virtue as an objective attribute of him, but also, particular past, contingent moments in a virtuous friendship remain alive in memory and in narrative retellings of the experience. Thus, Aristotle emphasizes in his discussion of memory that it is not merely a case of contemplating my own experience. I also possess an awareness of the object of contemplation as a presentation of the other (real) object:

Just in the same way we have to conceive that the mnemonic presentation within us is something which by itself is merely an object of contemplation, while, in relation to something else, it is also a presentation of that other thing. In so far as it is regarded in itself, it is only an object of contemplation, or a presentation; but when considered as relative to something else, e.g. as its likeness, it is also a mnemonic token. (On Memory and Reminiscence 450b)24

In other words, memories are not Cartesian inner experiences, nor are they simply reproductions or copies of past experiences. Ordinarily, memories are understood by the subject to be tokens of the original (p.156) object to which the memory has a likeness, if not identity. In the case of friendship, a memory of my friend is not merely a re-experiencing of my own affection or knowledge of her, or what I knew through her, but rather the memory about her, in her very being. Thus, Aristotle’s emphasis on memory in the case of lost friendship indicates the importance of faculties other than νοῦς‎ in the love extended in friendship; sense experience, memory, and other faculties not only work in the present moments of friendship, but also continue to play a role in the meaning of the friendship long after the friendship itself may have ceased.

Indeed, because the activity recalled through memory is often shared activity, whenever I remember a past incident with my friend in which I also shared, my own selfhood is also the object of reference. Memories of friendship are both mnemonic tokens of another and of one’s self in relation to the other. Practical activity, exercise of my human faculties, takes place not only alongside other rational beings, but also with and in relation to them in a strong sense. Memories of those activities constitute a part of my selfhood. To some extent, it can even be the dissonance between one’s past memories of a beloved friend and one’s self, and a current lack of friendship that inevitably cause the grief, pain, and especially sense of loss of one’s self within the current grief. In such experiences of loss, one beholds not only a lost friend, but also an aspect of one’s self that has been lost. For, the present moment cannot match past memories of shared experiences of φιλία‎, and yet the latter are still present as ‘reminders’ or tokens of what has been lost. Indeed, to the extent that memory memorializes the lost friend, it also keeps present the vulnerability experienced in the loss itself. However, memory also can assist in positively reconstituting the meaning of the loss, as in the case of developing a narrative about my friend’s larger experiences, or in using memory as grounds for hope that he will eventually live better.

Rather than characterizing the notion of the ‘other self’ as a logical paradox, the phrase might best be understood as communicating a truth about human interdependence: namely, that those closest to us in friendship both dissolve some of our sense of independence and separateness as we come to care for them, and yet always maintain a degree of alterity that shows us to be distinct beings whose characters, choices, and preferences can veer away from our own.25 In the case of (p.157) friends of virtue who find that one friend has repeatedly failed morally, such alterity suddenly comes to light, replacing the sensibility of connectedness and shared ‘selfhood’ that had been experienced until the discovery. In friendships of virtue that are going well, the harmony between oneself and the other is a source of ongoing pleasure. While the otherness of a friend can delight by its complementarity, such alterity is held in balance with a sense of abiding and shared values. The otherness of the other friend is not a threat, but a welcome addition to one’s own range of experiences and interests. However, when major differences of value appear, and indeed, these values concern not mere preferences (for example, for sports or musical interests), but affect deeply held conceptions of the good life, the friend’s alterity suddenly comes into relief. Indeed, in the case of a friend of apparent virtue who now reveals himself to be capable of major, repeated moral failure, the possibility that one’s true self will no longer be fully received in friendship also becomes apparent.

The decision to leave a friendship when one’s friend has repeatedly failed morally, when his character seems to be beyond repair, is not only a practical decision for a friend. Such a decision involves genuine loss for him, even if the decision is right. It is not simply that the person with whom one wishes to share activity, to develop one’s virtues, and to enjoy for himself is no longer there. One’s sense of trust is also diminished. Aristotle suggests that older people can so often suffer such losses that in time, many older individuals are shaped by a pervading sense of distrust for the imperfections of human nature (N. Ethics 1157b–1158a). Such distrust diminishes their capacity for happiness, but Aristotle does not suggest that their judgements are thereby mistaken. Rather, along with the great possibilities for happiness that intimate friendship opens up, so too is the possibility of diminished happiness brought on by repeated loss or betrayal.

Contemporary theorists have conceptually distinguished trust from reliance: while reliance implies only a belief that another person will follow through on a particular activity—for example, providing money for a trip home, or showing up to teach a class—trust implies a great set of interpersonal attitudes that guide one’s very orientation toward being in relationship.26 Trust can involve, for example, a belief (p.158) in the other person’s good will and integrity, a disposition to interpret the other’s actions favourably when they are in doubt, and an ‘entrusting’ of one’s vulnerabilities to another with the expectation that they will be reliably received.27 Aristotle’s conception of friendship requires that trust be built up in order for perfect friendship to develop, precisely because the self is being entrusted to another. It is not only the promise of shared activity, or shared conversation, that takes place in the promise of friendship, but also the choice to give over oneself to another and to receive the other as ‘another self’ that requires such trust. Violations of trust, then, are also felt to be violations of the self, or at least of belief that one is valued as ‘another self’ to another, and can reliably be entrusted to the other, in the context of friendship. Such violations are not escapable, but instead part of the human condition. While Aristotle advises that friends must share many meals together before trust is given, his advice is descriptive and not prescriptive. The development of trust is dependent upon many factors, including one’s own past experience, and can only arise when the right conditions are met. To this extent, Aristotle values a balance between self-sufficiency and dependency that takes account of ongoing human vulnerability in friendship.


The interdependency of human beings in moral development poses another kind of paradox for Aristotle, one with social and political consequences. For, if the goodness of one person can profoundly reinforce and even assist those still growing in virtue, then there would seem to be a duty to assist another person, even a wicked person, in growing to become a better person. As we have seen, such assistance potentially poses dangers for the friend who takes on this task, if it detracts from the good of his own virtuous and lived activity or simply seems impossible to accomplish. But because the development of virtue depends on others for its very possibility, there ought to exist some social resources that would allow a man who has morally (p.159) failed to regain his character. Virtue is far more important than property. And yet, the presence of such a vicious or incontinent person in public life also produces problems for the flourishing of the community, and especially for those with whom he is in closest relation. A former criminal, for example, who especially needs social resources that might lead him to live better, is also a potential threat to that community; and yet, to exclude him from participation within society and its goods is likely to make him grow worse, not better.

Here Aristotle recognizes a deep problematic in the nature of the development of virtue: to become good is partly a matter of personal responsibility, but also partly conditioned by those with whom we surround ourselves. In the case of a man who already lacks virtue, the only real possibility of his reform lies in exposure to and education by those who possess virtue. This is especially true in the case of the vicious, who believe what is bad to be good, and so lack an understanding of the good, not only an ability to follow through on it in action. However, those who possess virtue, seemingly, will not love those who lack it, for by their character they only love what is good and wish to pursue the good along with other good people. At most, they will only love the other for who he or she might someday become, or who they once were, but not for their currently bad state. Aristotle thus encounters the problem of how and whether friendship or other analogous relationships can sufficiently address the need for the community to engage with those who lack virtue.

At the purely theoretical level, Aristotle’s account of friendship is consistent. To understand the end of what had seemed to be a friendship of virtue, on account of a friend’s repeated moral failures, one friend need only see that the other was not the virtuous person he was believed to be. The virtuous friend who had loved, discovers that the beloved friend is, in fact, not even the person he had at first thought him to be. When the formerly good man is now seen to be bad, that person’s self is no longer identical to the person who was once the object of φιλία‎. Indeed, for Aristotle, one can not love what one does not understand to be good. Even in the case of a friend hesitant as to whether to continue a friendship with one who has failed morally, the remaining love for the friend concerns what is still good or lovable about him.28 For example, a drug addict working to (p.160) turn around his life and live cleanly is loved not for his drug addiction, but rather for his potential for living better. To this extent, Aristotle is not inconsistent in his account of virtuous friendship.

However, the notion of moral failure in friendship raises an additional practical and political problem, namely, how societies are to care for those who fail at a moral level when rehabilitation of such character depends upon engagement and the care of others who surround them. In other words, if we are vulnerable to harm from those who lack virtue, and yet also ought to care for the cultivation of virtue in others more than any other good, then vulnerability to harm becomes a political and ethical problem. For Aristotle, to be self-sufficient (αὔταρκες) also means to live a life among family, friends, and fellow citizens (N. Ethics 1097b7–11).29 However, no one of these types of relationships alone suffices for the expression or development of a virtuous and happy life. Particularly when it comes to the question of moral development or rehabilitation, we find in Aristotle a continual need for dialectical interplay between the role of the family and of the law for the possibility of establishing virtue in its citizens. Neither resource is adequate alone, and each one is subject to the possibility of misfortune. For Aristotle, the role of parents is significant, but limited, insofar as moral development continues long past the period of childhood or formal education (N. Ethics 1103b). The role of law, while commendable, is also limited in the extent to which it can shape virtue.

In Book X of the Ethics, Aristotle emphasizes the importance of laws that can reinforce the nurture and the earlier habits of temperance provided by parents (N. Ethics 1180a). Such laws are crucial to education itself, since the city’s flourishing requires that all of its citizens have a similar orientation to the civic good, rather than the happenstance education that any given parent might choose on his own for a child (Politics 1337a). Without some shared idea of the goal of education for children, individual moral upbringing cannot be harmonized into a civic morality.30 Such goals need to be continually lived out in the course of adult relationships as well, so that the law continues to assist its citizens in their moral development by setting (p.161) the proper conditions for the cultivation of good habits (N. Ethics 1180a).

The law, however, is imperfect as an instrument to shape moral growth, insofar as it can chastise those who are bad, but cannot provide the bonds of natural affection and individual attention to a specific person’s needs that take place within the context of a family. Just as a doctor can best address an individual’s illness by attending to the particulars of his illness, a person with moral deficiencies requires attention to his particular circumstances, which the law in its generality cannot legislate (N. Ethics 1180a–b). To this extent, the law is a second-order kind of institutional ‘insurance’ for the moral education of its citizens. The law can legislate how to educate children, but does not itself act as the primary educator. The law can punish those it deems to need fear as a motivator against bad action, but cannot single-handedly cultivate good internal motivation to seek the good and to reform one’s habits. But the sorts of resources that can provide such motivations—family, friends, political peers, and smaller communities within the πόλις‎—are even more subject to τύχη‎ precisely because of their more intimate nature. The family and class to which one belongs, the education one receives, and the stability of the political parties that might support such education are subject to chance and prone to instability. And yet, it is precisely the interdependent nature of the human being that both leads to such vulnerability to luck, and also vulnerability that makes φιλία‎ and other love relationships possible in the first place.

Moreover, certain moral states seem to arise not only from the influence of education, but also from luck. For Aristotle, the brutish state, one step below vice, in which there is hardly even a moral sensibility, seems to be a result of bad luck. He writes, ‘some bestial features also result from diseases and deformities’ (N. Ethics 1145a), such as the cases of a slave who ate the liver of another fellow human being (N. Ethics 1148b), indicating that the brute is not brutish or amoral by virtue of his own choice, but as a result of a defect. We can therefore say that the case of the brute is less evil, but more alarming, precisely because no choice is involved (N. Ethics 1150a). Even the incontinent man who is incontinent by virtue of strong passion is overcome by his bodily state, and akin to a person who is asleep, mad, or drunk. Aristotle explains the possibility of incontinence in a state of knowledge as acting in the presence of perceptual knowledge (such as ‘this is sweet’ and ‘sweet things are pleasant to eat’), rather than in (p.162) the presence of universal knowledge (‘Sweet things are unhealthy’). But although Aristotle mostly emphasizes that incontinence is the result of many habitual incontinent acts, he also allows for the possibility that some forms of incontinence may be present from birth:

Incontinents through habituation are more easily cured than the natural incontinents; for habit is easier than nature to change. Indeed the reason why habit is also difficult to change is that it is like nature, as Euenus says: ‘Habit, I say, is longtime training, my friend, and in the end training is nature for human beings.’ (N. Ethics 1152a)

Aristotle thus suggests that there are at least two obstacles to moral freedom, two different types of vulnerability to moral weakness: the presence of incontinence from birth, or from a state that, after enough time, so resembles the inborn state that it is immutable, though it may not have once been so.

For Aristotle, the family is natural, and the procreative bonds and relations between couples, parents, and children are found even in the animal world (N. Ethics 1162a17–19). Parents seem to love their children with an intensity that is unmatched in other forms of relationship, while children owe their parents a debt that cannot be repaid (N. Ethics 1161b–1163b). However, the family alone is limited in the moral and social goods that it can provide, while the πόλις‎ provides for the possibility of not just living, but living well (Politics 1252b12–13).31 The family, while attentive to the needs that individuals have of one another in certain respects, is itself limited and partial in its possibilities. The πόλις‎ is not merely a collection of families and associations, but rather a greater whole that offers more than the sum of its parts. The πόλις‎ offer its participants not only the advantages of laws and of extra-familial connections, but also opportunities for magnanimity, shared religious-political experiences, and the intimate connections between others found in friendships of virtue and even friendships of pleasure.

Moreover, the care that any given family provides is limited by the virtues and abilities of the parents, who may themselves be limited in the moral structure and nurture that they can provide for their children. As the contemporary ethical theorist Gheaus has suggested, we might consider such care to be a primary good that limits the possibility of political egalitarianism, as advocated by political (p.163) theorists such as Rawls.32 While Rawls argues that the distribution of basic goods in society must be provided, Gheaus argues that ‘care’, broadly construed, is subject to unequal distribution in society. Under the category of care, Gheaus includes: the care required to come into existence; to be sustained through food and physical care through childhood; care as an attitude that teaches self-care, for example, receiving good nurture during illness or encouragement in studies that lead the grown individual to provide good care for herself as an adult who has experienced such modelling of care; networks of care, such as social networks, like those who assist one’s immediate family with job connections, or other forms of favouritism; and deeper social connections such as friendship or extended family. All of these forms of care significantly contribute to the development of character and to the ability to exercise virtue, and yet are subject to a great deal of moral luck, and not moral choice. Children do not choose their families of origin, the communities into which they are born, and whether they are strong or weak social networks. While some social institutions can help to mitigate bad moral luck, institutional care cannot replace personal, one-on-one care that stems from deep, personal relationships of family and friendship, so Gheaus suggests a certain degree of caution in thinking that institutional care can be a sufficient response to moral luck. Her conclusion is to state that egalitarian political theories can never adequately accommodate the ethics of care, that is, no political system can really prevent the contingencies of moral luck that affect care relationships.

Aristotle’s strength is to acknowledge the significant role of such luck in ethical formation. Aristotle admits to a certain kind of weakness in those who are incontinent as a result of birth or as a result of how they are educated. Care or nurture (τροφή‎) can enhance or diminish the possibility of becoming a virtuous person, and yet, Aristotle is well aware that they are not equally distributed. Aristotle does not discount the possibility that some forms of human unhappiness are beyond our control, insofar as being born prone to incontinence, or being poorly raised by inadequate parents or inadequate laws, is partly a matter beyond individual choice. To this extent, Nussbaum is right to identify in Aristotle a sensitivity to the tragic (p.164) element in human life; this tragic element exists not only for those who are like Priam (virtuous, but then hit hard by loss), but also those who are unable through circumstance to get to Priam’s level of virtue in the first place. Especially if we concede that care is fundamental to the development of virtue, and yet unequally distributed, Aristotle’s theory is attentive to the role of chance in moral development.

Despite this acknowledgement of unequal beginnings, Aristotle ultimately places most of the responsibility for particular moral choices on the individual. Even in the case of vice, a moral actor must make those choices that can reshape his own character in order to become more virtuous. Book III’s explanation of the voluntary and involuntary begins with the basic claim that virtue and vice are in our power. Virtue and vice alike develop in our characters as a result of repeated, chosen actions. Habitually acting generously makes us generous; habitually acting courageously makes us courageous. While it might be possible for a person to choose to live poorly so often that eventually he lacks the freedom to choose otherwise, Aristotle still places the burden on the moral actor: like someone who has cast a stone away, and cannot now recover it, a moral actor acted in ways that once were in his own power, and so he is morally responsible for who he has become: ‘The same is true of weakness or maiming; for everyone would pity, not reproach someone if he were blind by nature or because of a disease or a wound, but would censure him if his heavy drinking or some other form of intemperance made him blind’ (N. Ethics 1114a).

However, the case of the person who was raised poorly raises certain difficulties. If Aristotle thinks that the incontinent person is incontinent and unable to control his impulses because of past education and upbringing, how do we expect him to act virtuously now? If those who lack good experiences of early care and nurture can act in ways that only increase vice as a result of early experiences, how do we get them to act in a more virtuous manner now? Put succinctly, if one needs virtue in order to act virtuously, then it seems problematic to ask that the person who lacks virtue as a result of moral luck take responsibility for his moral failure, if its possession is what is required for a proper response. There seems to be a problem of circularity: one would need virtue to act virtuously, and need to act virtuously, in order to be virtuous; yet, this is precisely what has been harmed through moral luck.

(p.165) Aristotle’s strength is to recognize the ongoing reality of this paradox, and not to seek to escape it through theoretical machinations. Aristotle’s understanding of the difficulty of moral education includes sensitivity to the ways in which human beings are mutually dependent upon one another for their moral care, as well as physical and emotional care. The circular connection between being virtuous in character, and acting virtuously out of that state, is not necessarily a vicious circle. We do not find in Aristotle a logical problem with the mutual dependence of virtuous action on character, but instead a practical one. In some cases of incontinence (ἀκρασία‎), even the practical limitations might be overcome. While it is true that the generous person will find pleasure in generous acts, and these acts continually reinforce his generosity, it is also the case that a non-generous person can choose to become more generous through repeated, deliberate practice of generosity. Even if now it is uncomfortable, eventually through repetition, he will find pleasure in it when he does become a generous person. Indeed, he might already find a certain kind of pleasure in the anticipation of becoming such a person. This is how we raise children to become generous, through small steps, where we reinforce the child’s acts of sharing—for example, praising a young child with even the smallest act of sharing a toy with another. The same might be said in the case of an adult who experienced a lack of care, or a Philoctetes who feels damaged by a long abandonment: a certain degree of choice to perform the virtuous act over the non-virtuous act can be fostered and nourished over a longer period of time. Not only individuals like Neoptolemus, but also entire societies can take on the role of encouraging and supporting such choices in others.

The deeper difficulty arises with cases of vice, in which the moral actor not only cannot act virtuously, but even understands the bad to be good. In such cases, there will be no motivation to act in accordance with virtue, for the ends toward which the moral actor is oriented are mistaken. Here Aristotle argues that the law must take over where family and friends are unable to develop proper sensibilities about moral virtue, by at least punishing those who act without care for the good either of themselves or of society. However, the law and institutional support of such moral choice can never substitute for the individual responsibility to participate in the process of developing virtue.

(p.166) Political bonds between those who are neither intimate friends, nor merely formalized legal relationships, are also a necessary presence for the moral education of the citizens. Notably, Aristotle does not rely on the phrase πολίτικη‎ φιλία‎, or ‘civic friendship’, in developing his Politics.33 Neither is concord primarily a bond based in virtue; instead, concord concerns shared interests of a utilitarian nature.34 His avoidance of φιλία‎ as the grounding of political relationships rests in part on his recognition that the kind of care that takes place in an interpersonal relationship of φιλία‎ is qualitatively distinct from the care that legislators can offer to the citizens of the πόλις‎. For Aristotle, φιλία‎ provides a sort of intimacy of care that is generally lacking in institutionalized forms of care. Such a human need for φιλία‎ suggests that there are limitations to political systems, such as that in Plato’s Republic, which seek to make the city and public life the primary source of care and ‘brotherly love’. While the law’s primary motivator is punishment, the motivation for friendship consists of genuine love of the other in a one-on-one relationship.

Nonetheless, the grounding of the community’s bonds with one another exists as the result of social bonds that are analogous to friendship. Aristotle asserts that while the city initially comes into being for the sake of life, its ultimate purpose is for the sake of living well (Politics 1252b29–30). Civic relationships, as well as laws, are part of the means by which the city assists its members in flourishing, and not simply surviving. As opposed to contemporary rights based on liberal systems of government, Aristotle’s political system requires a strongly experienced feeling of community.35 The bonds between citizens are not understood as respect for rights, equally given to all citizens. Instead, the political community is grounded in active participation in a variety of associations, both formal and informal. The bonds between various forms of association are variable in both intensity and intimacy, and do not always manifest equality between all citizens. Book VIII describes all political associations as aiming at a particular advantage for their members. Such associations simultaneously bind together their members, while also separating those who do not belong from the particular form of friendship and care engendered by such association. Inevitably, different groups of associations will cause conflicting loyalties, and can even threaten (p.167) political cohesion if taken too far. Competition for honours, too, can cause division.36 However, precisely because Aristotle’s πόλις‎ is grounded in more intensely felt relationships and interdependencies than those found in formal legal bonds, it also deeply engenders an experience of belonging. Aristotle’s description of concord, for example, includes the experience of genuine good will for others, beyond the simple utility that agreement brings for one’s self interest.37 One cares not only for the common good in abstraction, but also for those whom one knows to be a part of the same political community, in their concreteness, as those who share similar opinions and goals. And while competition for honour can be factious, honour is interpersonal by nature. Honour binds together those who participate in merit-based competitions and in the giving and receiving of honour.

Political friendships and intimate friendships are distinct, but powerful, motivators for the moral development of human beings, because of the need that human beings have of one another, and the limits of any single form of relationship for all forms of moral and political growth. Aristotle keeps alive the tensions inherent in our interpersonal relationships: for example, between the desire for friends to care for one another, even in the midst of moral failure, and the importance of recognizing the other as other, even as sometimes beyond our ability to help. The community, too, must play a role in the care of those still growing in virtue, or who suffer in lacking it and the accompanying εὐδαιμονία‎ that virtue brings. Indeed, it is precisely at points where individual relations of friendship or family fail that the larger community must step in to assist its citizens. At the same time, law and the larger community can never replace such intimacies, as contingent, imperfect, or powerless as they can be. For Aristotle, to be human is to require the vulnerability of being in relationship with family and friends, imperfect though they may be.


(1) Translations used for the Ethics in this chapter are from Aristotle, Nicomachean Ethics, trans. Terence Irwin (Cambridge, Mass.: Hackett, 1985). Greek consulted in Aristotle, Nicomachean Ethics, ed. J. Bywater (Oxford: Clarendon Press, 1894), accessed online at the Perseus Digital Library, 〈http://www.perseus.tufts.edu/hopper/text?doc=Perseus%3atext%3a1999.01.0053〉, 2010–12.

(2) John Cooper, Reason and Emotion (Princeton: Princeton University Press, 1999), 341–3.

(3) Nussbaum, Fragility, 350, 357.

(4) Nussbaum, Fragility, 359–61.

(5) Here I am particularly thinking of Stern-Gillett’s argument that Aristotle’s notion of friendship centres on loving the rationality of another, thus excluding his particularity and contingent goods from what is loved in the friend. See Stern-Gillett, Aristotle’s Philosophy of Friendship, especially chapter 3.

(6) Cooper also sees vulnerability, particularly psychological vulnerability, in such relationships. See Cooper, Reason and Emotion, 351.

(7) Annas argues that the primary good of a friendship of virtue is to contemplate the goodness of one’s friend, on the grounds that the self-sufficient man has enough in his own virtue that he is not in need of others there. See Julia Annas, ‘Plato and Aristotle on Friendship and Altruism’, Mind 86 (1977), 532–54. However, since happiness includes the activity associated with the virtues, and not only their possession, to have others with whom to share this activity is central to our ability to do it more easily and more often. See also Julia Annas, The Morality of Happiness (Oxford: Oxford University Press, 1993), chapter 12. See also Cooper, Reason and Emotion, 345–51.

(8) Nussbaum, Fragility, 345.

(9) Lorraine Smith Pangle, Aristotle and the Philosophy of Friendship (Cambridge: Cambridge University Press, 2003), 39.

(10) In contrast, Dale Jacquette argues that even friendships of virtue are good primarily because of the personal benefit that results in having them, for example, through encouraging the growth of one’s own virtue. Peter Drum instead argues that Aristotle understands φιλία‎ in friendships of virtue to serve ‘the good’ and sees good friendships as part of loving the good. But for Aristotle, the very point of virtuous friendship is that love of the good and love of the friend coalesce: in loving a person who is truly an exemplar of what it means to be human, I love the good in him. To suggest that loving him is a means of loving ‘the good’ introduces an element of Platonism foreign to Aristotle’s understanding of the good as inhering in particular types of beings, and even in relations between these beings. See Peter Drum, ‘What is the Motivation of Morality for Friendship in Aristotle?’, Journal of Value Inquiry 37 (2003), 97–9 and Dale Jacquette, ‘Aristotle on the Value of Friendship as a Moral Motivation’, Journal of Value Inquiry 35 (2001), 387.

(11) Annas, Morality of Happiness, 261.

(12) David Bostock, ‘Pleasure and Activity in Aristotle’s Ethics’, Phronesis 333 (1998), 251–72; Peter Hadreas, ‘The Functions of Pleasure in Nicomachean Ethics x 4–5’, Ancient Philosophy 24 (2004), 155–67; and Gary Gurtler, SJ, ‘The Activity of Happiness in Aristotle’s Ethics’, Review of Metaphysics 56 (June 2003), 801–34.

(13) Annas, Morality of Happiness, 251–2.

(14) Pangle, Friendship, 50.

(15) Thus, Annas says, when a friend is lost upon death or leaving town, we naturally feel exposed. Annas, Morality of Happiness, 252.

(16) Stern-Gillett, Aristotle’s Philosophy of Friendship, chapter 1.

(17) Pangle, Friendship, 146.

(18) Stern-Gillett, Aristotle’s Philosophy of Friendship, 59.

(19) Pangle, Friendship, 147.

(20) Annas, ‘Plato and Aristotle’, 532–4.

(21) Annas, Morality of Friendship, 548–9.

(22) Stern-Gillett, Aristotle’s Philosophy of Friendship, 26–7.

(23) Stern-Gillett, Aristotle’s Philosophy of Friendship, 41.

(24) Translation of On Memory and Reminiscence from The Basic Works of Aristotle, ed. Richard McKeon (New York: Random House, 2001), translations by Rhys Roberts and Ingram Bywater.

(25) Nussbaum, Fragility, 351.

(26) Margaret Walker, Moral Repair (Cambridge: Cambridge University Press, 2006), 74–5.

(27) Walker, Moral Repair, 74–9.

(28) Cooper argues that friends love one another for moral properties of character; in his view, such motivation does not make friendship utilitarian, for a person’s character is essential to his being in a way that utilitarian goods are not. See John Cooper, ‘Friendship and the Good in Aristotle’, Philosophical Review 86 (1977), 290–315.

(29) Nussbaum, Fragility, 344–5.

(30) Nussbaum, Fragility, 346.

(31) Pangle, Friendship, 86–7.

(32) Anca Gheaus, ‘The Challenge of Care to Idealizing Theories of Distributive Justice’, in Lisa Tessman (ed.), Feminist Ethics and Social and Political Philosophy: Theorizing the Non-Ideal (Springer), 105–19.

(33) Stern-Gillett, Aristotle’s Philosophy of Friendship, 149.

(34) Pangle, Friendship, 158.

(35) Pangle, Friendship, 81.

(36) Pangle, Friendship, 84–5.

(37) Pangle, Friendship, 156–7.