Jump to ContentJump to Main Navigation
Friendship ReconsideredWhat It Means and How It Matters to Politics$

P. E. Digeser

Print publication date: 2016

Print ISBN-13: 9780231174343

Published to Columbia Scholarship Online: January 2017

DOI: 10.7312/columbia/9780231174343.001.0001

Show Summary Details

Motivations, Actions, and the Value of Friendship

Motivations, Actions, and the Value of Friendship

(p.24) Two Motivations, Actions, and the Value of Friendship
Friendship Reconsidered

P.E. Digeser

Columbia University Press

Abstract and Keywords

Establishes that our practices of friendship structure how friends act towards one another as well as the motives that friends must mutually recognize. Different practices of friendship are conditioned by different motives as well as by different adverbial rules governing the actions of the friends. The chapter defends the view that the value of our friendship is intrinsic to the relationship (also called the relationships-view).

Keywords:   actions, Keller, motivations, relationships-view, self-disclosure, Telfer

In arguing for a practice-based view of friendship, chapter 1 claimed that motivations and actions play an important role and that we unavoidably make choices over how to maintain our friendships and possibly whether to initiate them. It also claimed that we should not expect to be able to provide the necessary and sufficient conditions for friendship if it is to be understood as a family resemblance concept. This chapter digs deeper into that idea by discerning its implications for understanding the role of motivations and actions in our practices of friendship. It begins by noting that the philosophical call to provide a set of necessary and sufficient conditions for friendship has not gone unanswered.1 Exploring the difficulties of one such account—offered by Elizabeth Telfer—will help to formulate the character of friendship’s diversity. In many respects, the following discussion is indebted to Telfer’s account even as it backs away from her stronger, essentialist-sounding claims. Responding to her position does not demonstrate the impossibility of setting out the necessary and sufficient conditions for friendship, but it does point to the difficulties of trying to set out those conditions.

After briefly summarizing Telfer’s position, this chapter turns to her account of the motivations of friendship and argues that a plausible case can be made for widening the scope of appropriate motives beyond her discussion of affection and fondness to include desire, respect, appreciative (p.25) love, and care. Instead of friendships being driven by an essential motive that defines the relationship, there exists of repertoire of motives (and actions) that can condition a friendship. Indeed, what is understood to be part of the repertoire of appropriate motives is a function of time and place. The “Essential Motivations of Friendship” section draws on Telfer’s account in another way and claims that friends must recognize the appropriate motives in one another in order to be friends. As we shall see in chapter 3, mutual recognition of motives allows us to distinguish friendships from mere friendly actions as well as to identify false friends from actual ones. The “Mutual Recognition of Ordinarily Appropriate Motivations” section takes up the action conditions of friendship and the “Importance of Acting Friendly” section fills out Oakeshott’s notion of self-disclosure as a way to understand those action conditions. Our practices of friendship are governed less by demands for the performance of specific actions and more by adverbial rules: the question is how friends should act toward one another as opposed to setting out specific actions. Because the term self-disclosure has been employed in earlier understandings of friendship, this section (“Adverbial Conditions of Friendship and Self-Disclosure”) explores the differences in these understandings. The “Explanation, Justification, and the Practice of Friendship” section asks what we have gained in understanding friendship as a family of practices. Have we “explained” friendship? Have we provided a way to “justify” friendship? These questions bring us to the limits of a practice-based account of friendship as well as to issues surrounding the value of friendship. The remaining sections of the chapter defend the view that friendship itself can provide reasons for action by responding to Simon Keller’s argument to the contrary. Friendships are intrinsically valuable, even if their value does not outweigh all others.

The Essential Motivations of Friendship?

In her account of friendship, Telfer argues that friends must share activities. More specifically, she argues that friends need to provide reciprocal services (you shovel my driveway and I take in your mail when you are on vacation), have mutual contact (we talk, go to the movies, or simply spend (p.26) time together), and have joint pursuits (we belong to the same book group, synagogue, community organization). She believes that these shared activities, while necessary conditions, are not sufficient. For, in addition to doing certain things, we must also perform them out of certain desires. In Telfer’s view, these desires are the “passions of friendship.” In other words, “friends must have affection for, or be fond of, each other.” Telfer goes on to define affection “as a desire for another’s welfare and happiness as a particular individual. This desire is thus to be distinguished both from sense of duty and from benevolence” (1991:251).2 She argues that in a friendship, this affection is not connected to character (unlike what Aristotle saw as necessary in the most complete form of friendship). The characteristics of an individual may stimulate affection, but affection, at bottom, is irrational. Consequently, friendships are able to survive changes in character. It is difficult, she argues, to put our finger on why we like someone. It is not simply a matter of adding up a set of characteristics. Rather, it is something about the whole package that draws us to them (253). In addition to having affection for the person, Telfer argues that friendship is defined by a desire to be with our friends, to be in their company and not just “a desire for company as such” (252). Finally, Telfer argues that friends must not only have the passions of friendship, they must also choose to act on those passions (256). But the choice must be mutual. A friendship requires a kind of meeting of the minds. Hence, one cannot “choose to be a friend of just anyone.” As she notes, what is also necessary is that “the existence of the passions of friendship in both parties, and the practice on both sides of acting on them, once established be acknowledged by the parties” (257). Or, in terms that have been used here, friendship requires the mutual recognition of the motives/sentiments of friendship.3

In a broad way, Telfer’s focus on shared activities and the passions of friendship fit nicely into the formal structure of practices that includes both motivations and actions. The question now becomes whether these passions and these kinds of activities are necessary for friendship. If they are, do they offer a plausible account of the necessary and sufficient conditions of friendship? Undoubtedly, in some friendships, the elements discussed by Telfer are indeed essential. In such cases, if these conditions are absent, then the particular friendship would dissolve. Whether they are necessary for all (p.27) friendships is not so clear. For example, Telfer claims that friends must have affection for one another. At one level, this claim seems to be undeniable. What would a friendship mean if there was no affection? Nevertheless, Telfer admits that “to some extent, of course, I am stipulating, rather than reporting, that the presence of the inclinations is a necessary part of friendship” (1991:256). In terms of how other thinkers have conceived of friendship, Telfer’s stipulation would be largely accurate. For example, Preston King notes the importance of affection, but then writes, “At the end, what affection is, and whether it is genuine, and appropriate, may be best left to whoever owns it” (2007:131). Other thinkers place less weight on the importance of affection. For example, it is not at all clear that affection was a necessary condition for all of Aristotle’s conceptions of friendship (Grunebaum 2003:38).4 In his account of friendship between old people, he noted that they may not even “find each other pleasant” (Aristotle 1985:1156a25–30). Nor is it clear that affection is essential to Kant’s conception (1991:216), or perhaps even C. S. Lewis’s conception of friendship (1960:70–72).

If affection does not appear to be necessary to these other conceptions of friendship, then what else motivates individuals to become friends? An alternative motivation that is frequently mentioned is that of desire. But is not desire the same thing as affection? While the two ideas are frequently used interchangeably, it is possible to distinguish a form of desire that is not the same as affection in a way that is useful for understanding different kinds of friendship. Very generally, affection tends to be directed to the particularity of the object of affection. Kim-Chong Chong notes that,

In the case of material objects, the object is valued in such a way as to express its uniqueness and irreplaceability. In the case of persons, one does not have an affection for another simply because he possesses certain qualities—someone may possess the most appalling qualities, and yet one may still have an affection for him. Similar qualities in another might, on the other hand, arouse dislike. Granted that the reason for this could be described, by pointing to more detailed differences, but this does not obviate the fact that not all that goes by the name of “affection” can be so described. (1984:354)

(p.28) In contrast, desire can be focused, not on the particularity of the object of desire, but on the qualities of the object that fit it into a given class. On this view of desire, what matters are those qualities, and if another object or person possesses those attributes, then they can replace the original object or person.

Assuming such a distinction between affection and desire, it is easy to see how affection for another can serve as a motive for friendship. Indeed, for Telfer and, at least initially for Chong, there is a sense in which the particularity associated with affection is the sine qua non of friendship. After all, it is sometimes said that one friend cannot be replaced by another. For Chong, “unless there is or comes to be a care and concern for the other person in himself, there can be no close personal relationship to speak of” (1984:355). It would seem, then, that affection is necessary for the kind of close personal relationship that defines friendship.5

Despite these claims, however, Chong does not go so far as to claim that relationships that are desire-based cannot be friendships. For she also notes, “There are friendships which are not friendships in the full sense of the word, as we have described it. One may have a friend whose company one enjoys, but for whom one cares little or not at all, and some friends are friends whose companionship it would be advantageous and pleasurable for one to have. An egoist is not precluded from friendships of these sorts” (1984:355). What Chong is saying of the ethical egoist can be said of those for whom friendship is based on desire. What is being presumed in this discussion is that while these friendships may not encompass the “full sense of the word,” they could still be understood as friendships. But why do they not qualify as full friendships? For Chong, the claim may be intuitive or merely stipulative, namely that care, affection, and commitment are essential features of friendship and that there is something lesser about “mere” companionship. She may very well be correct that there is something inferior about friendships motivated by desire as opposed to affection, but that does not preclude them from being friendships of a sort.6

To clarify further, it might be helpful to put this in an Aristotelian tripartite framework of friendship. Among friendships of plea sure, utility, and virtue, friendships of plea sure could be understood as friendships that are motivated by desire in Chong’s sense of the term. These are friendships in (p.29) which the parties see one another not as individuals in their particularity, but as persons of a certain sort—for example, they are good skiers, they enjoy partying, they are entertaining to be around, or they can tell a good story. Friendships of this sort are less a function of attending to the unique characteristics of the individual and more focused on the qualities that they happen to possess. In these kinds of friendships, the relationship is subject to the vicissitudes of those qualities. When they change, then the motivations for maintaining the friendship may dis appear. Going fishing with George can be the basis for a friendship for Bill, but if Bill no longer enjoys fishing then the friendship may fade. There is something, as Aristotle noted, short-lived or unstable about friendships of plea sure. For our concerns here, what ever one may think of these friendships, they are still friendships, despite the fact that they are not motivated by affection. Desire, when distinguished from affection, can also serve as an appropriate reason for a friendship.

Even if one is unconvinced by a distinction between affection and desire, there are other motivations for friendship. A third possible motive for friendship is respect. In what Marilyn Friedman calls particularized respect, “someone is admired specifically for her worthwhile qualities, her excellences…. It may involve affection or fond feelings, but it need not” (1993:194). In this sort of friendship, the friends are drawn to one another because they are excellent musicians or excellent tennis players or just because they are smart. Perhaps through one’s association with the other individual, one is challenged and made a better tennis player. This association may be enjoyable, but it may also be difficult and demanding. It does not require the kind of vulnerability in which one’s secrets are brought to light, but rather a vulnerability in which one’s weaknesses are potentially revealed. Perhaps a version of this motivation is implied by an older conception of being drawn to the nobility or virtuosity of another.7 In this regard, it could be expressed in the kind of friendship between Aristotle’s great-souled individuals.8 Obviously, respect can be understood in a variety of ways. In this case, it is not the sort of generalized respect owed to all humankind, but a respect for excellence.

A fourth possible motivation for friendship that could be seen as a shade different from respect is presented in C. S. Lewis’s work. In his essay on friendship, Lewis places great weight on friends acting in a particular way (p.30) and sharing activities. Nevertheless, he is also quite clear that friendship is a form of love. In fact, it is a love that requires a fairly high degree of self-cultivation and restraint. It is not a love that is subject to jealousy (1960:67), and it is “is utterly free from Affection’s need to be needed” (69); it tends to eschew gratitude (70). Rather, the love that motivates friendship is admiration or what he calls “appreciative love” (71). He writes, “In a perfect Friendship, this Appreciative love is, I think, often so great and so firmly based that each member of the circle feels, in his secret heart, humbled before all the rest” (71–72). While Lewis is adamant that friendship must be about something, in the sense that the friends share some truth or purpose, the motivation of appreciative love is distinct from affection, eros, and gratitude. To the degree that it is driven by a sense of humility (perhaps the Christian virtue of humility) it is also distinguishable from the notion of respect just discussed. It may very well be a conception of friendship that is unique to early to mid-twentieth-century Oxford dons, but it is still a conception of friendship.

A fifth motivation for friendship that may be distinguishable from these other motivations is what Blum calls “deep caring and identification with the good of the other.” In many cases, we care for someone because of other feelings that we possess. We care for a friend because we love her or have affection for her. But it may be possible to conceive of care as a free-standing feeling which can be mutually felt and recognized by friends. In Blum’s formulation, “caring means that if trouble arises between … [the friends], they will try to work it through.” On this account, it is because the friends care so deeply for one another that they experience comfort and joy. Moreover, this sense of caring for the other is moderated by one’s sense of separateness from and knowledge of the friend (Blum 1980:70).9 Caring for someone as a motive is distinguishable from simply liking or having affection for them.

There may be other motivations for friendship that have existed, do exist, or will exist that are different from affection, fondness, desire, respect, care, and appreciative love. Kant, for example, thought that need was presupposed in every friendship (1991:213; see also Nixon 2015:128). The need here is not material, but a need for confidence in the other.10 The larger point of this discussion is not simply to multiply the possible motivations, but to (p.31) suggest that affection is not the only motivation for friendship. In fact, there may be no essential passion or passions of friendship. It may be more helpful to argue that in any given culture there is a repertoire of motivations that are understood as necessary elements for interpersonal friendship. Within that repertoire, any single motivation, if it is mutually recognized, can meet the motivational requirements of the relationship. But these motivations can be present in various combinations and to various degrees. One may respect, have genuine affection for, and care for a friend. Alternatively, one may have a friendship in which one simply desires to be with him because he is fun. Moreover, the motives for a friendship can change over time. What began as a friendship based on desire could become one based primarily on affection or appreciative love (or some combination of motivations). As we shall see in part 3’s discussion of friendship among states, even though the mutual recognition of sentiments is not necessary (or even possible), other sorts of reasons will still need to motivate the relationship.

The Mutual Recognition of Ordinarily Appropriate Motivations

The argument so far suggests that friendship is situated in a social context that sets out the repertoire of appropriate motivations. If individuals do not possess the appropriate motivations or fail to recognize them, then the friendship cannot get off the ground or, in the case of an existing friendship, it will have to be rethought or abandoned altogether. Telfer’s view, however, points to an important feature of our current practices of friendship: friends must recognize appropriate motives in one another in order to be friends. Part of our practices of friendship involves conveying and recognizing something from that repertoire of motivations in others. If that capacity to read the sentiments and feelings of others is absent or disabled, as in the case of certain psychopathologies, then the possibilities for friendship diminish and perhaps dis appear. Consequently, friendship requires that we are not entirely opaque to one another.11

How is this motivational requirement met? At some point, two individuals arrive at the belief that their interactions are motivated by the right sort (p.32) of sentiments or reasons. This belief may be bolstered after a mutual declaration of friendship (“You are such a good friend!”), but it is more than likely to be reached by the parties inferring the motives of each other from a series of actions and interactions (“Thanks for going out of your way and picking me up at the airport … again”). The friendship, of course, is genuine if the inferences are correct. It is a misunderstanding or a charade (more on this in chapter 3) if the inferences are incorrect.12

Once the friends believe that their interactions are motivated in the right sort of way, must every interaction be properly motivated to sustain the friendship? Such a high level of self-reflection may make sense in an ideal of friendship (such as the one that will be discussed in chapter 4), but most ongoing friendships are more likely to be sustained by continuing to infer the motives of one’s friend from her actions. Consider an ongoing friendship between Cuddy and Greg. Even if Cuddy, with little thought, invites Greg to her party or happens to be inviting him when she is angry about something he did, her actions will continue to confirm (or disconfirm) their mutual belief that they are friends. If Cuddy does not act in a way that is consonant with their friendship, or simply declares that she does not want to be friends with Greg, then Greg (if he is not completely thickheaded) will conclude that the friendship has come to an end.

The importance of the mutual recognition requirement combined with the role of belief-sustaining actions has a number of implications. The first is that the epistemic requirements for establishing a friendship are not insignificant. Even if the mutual recognition requirement can be eased or evaded by the role of belief-sustaining actions, it points to the difficulties of seeing friendship as a general model for citizenship (discussed in chapter 5) and the problems of distinguishing friendships, friendly actions, and false friendships (discussed in chapter 3). Second, recognizing the appropriate motives for friendship helps to distinguish friendships from other sorts of relationships (lovers, partners, associates, comrades) as well as to identify different practices of friendship. Third, the importance of belief-sustaining actions helps to differentiate the reasons for a friendship (which are traceable back to the mutual recognition of appropriate motivations) from reasons for acting within a friendship (which are understood to be consistent with those motives because they accord with a mutually acceptable practice (p.33) of friendship). Cuddy does not always have to feel affection for Greg when she interacts with him. If she wants to remain a friend, however, she will need to act as a friend would act (enough of the time). Finally, the importance of belief-sustaining actions does not diminish the importance of the appropriate sentiments for the relationship itself. However friendly Cuddy may be to Greg, the relationship fails as a friendship if Cuddy does not recognize the appropriate motives in herself or Greg comes to disbelieve that Cuddy actually possesses those motives.

The Importance of Acting Friendly

The role of belief-sustaining actions brings us to the importance of action in friendship. The mere recognition of a set of appropriate sentiments is not sufficient. For example, an announcement by Sue and Kathy that they respected one another would not be sufficient for friendship. As Telfer notes, friendship is not merely about the mutual recognition of appropriate motives, it is also about doing the sorts of things that friends do. In her account of friendship, the doings involve reciprocal services, mutual contact, and joint pursuits. While important to some understandings of friendship, these activities are not necessary for all friendships. Friendship does require mutual contact, but that is a condition for the existence of many types of relationships and is not a particularly helpful thread. Must friends engage in joint activities? In Kant’s understanding of friendship in the “absolute sense,” individuals open themselves up to one another (1991:214). This kind of relationship is perfectly intelligible as a form of friendship, but it is not one that requires joint activity. Must friends engage in reciprocal services? Once again, it is not evident that either Kant or Montaigne thought this was necessary. In addition, when friends are unequal, it may be very difficult for one friend to reciprocate, which may encourage the friends to avoid being useful to one another; i.e., placing the unequal friend in the embarrassing situation of having to reciprocate when she lacks the means to do so.

Nevertheless, it is impossible to think of friendship without friends interacting. Could one distinguish friendship from other relationships based on certain categories of activities? For example, what do we do with friends that we do not do with strangers, acquaintances, relatives, teachers, or (p.34) psychoanalysts? Moreover, can we distinguish different types of friendship based on different types of activities? In the former case, perhaps friendships create an expectation that one will share one’s secrets with one’s friends, but not with a mere acquaintance.13 In the latter case, perhaps some friends are for doing fun things, some are for doing useful things, and others challenge us to be the best sort of person that we can be. Does it make sense to focus on the sorts of things that are done with friends?

To some degree, focusing on specific activities to identify friendships is a plausible way to differentiate them from other sorts of relationships and to distinguish among different practices of friendship.14 While plausible, this approach is not particularly parsimonious. The problem, of course, is that we can do many different things with the same friends without necessarily changing the character or quality of a friendship. Miles may be friends with Jack because they were college roommates years ago and because they occasionally go golfing together now. Miles’s decision to take Jack wine tasting to teach him about the virtues of pinot noir need not transform their friendship. Alternatively, the activities that may define a friendship can also be done with nonfriends. I may share the deepest secrets of my life with my friend, but I may do the same with an analyst, a priest, or a prosecutor. We may seek to be friends with those who are useful to us, but not every one who is useful is a friend. The virtue and zeal of St. Joan may be admirable, but being her friend might be unbearable. Basing friendships on activities points to something important, but differentiating practices in this way will either result in an unwieldy number of categories or be untrue to the flexible character of our friendships.

The Adverbial Conditions of Friendship and Self-Disclosure

As we saw in Oakeshott’s discussion of practice, it is possible to differentiate practices not on the basis of what is specifically done, but on the basis of how actors should interact. The adverbial conventions of friendship establish a kind of protocol to which the friends subscribe when they are acting as friends and distinguish their friendship from how they go about interacting with those who are not friends.15 For example, how Tony interacts (p.35) with his therapist may be quite different from how he interacts with his friend Paulie. With his therapist, his interactions are governed by norms of professionalism. He may approach her deferentially (she is trained), expeditiously (she is being paid), and inattentively to her welfare (he has no other interaction with her). In contrast, when he tells Paulie about a secret in his life, he speaks comfortably and easily and is attentive to how the secret may affect his friend.

A set of adverbial conditions may not only distinguish friendship from other relationships, but also distinguish different types of friendships. On this account, a specific practice of friendship does not establish whether one goes to a movie or reveals one’s secrets to another but how one does those things. For some friendships, interacting intimately, honestly, and authentically is prized while in other friendships interacting usefully, consistently, and reliably is seen as the norm. I may go to a movie with an old friend from college or with a friend from work, but how I (and she) will interact in that experience may be (assuming they are different sorts of friendships) quite different. In both cases, while we may share a cab, buy tickets, and talk about the plot, how we go to a movie can be performed differently. With the friend from work we may act reservedly, speak carefully, and eat popcorn moderately. In the case of an old chum, our actions and words may flow more freely, wholeheartedly, and unconditionally. Alternatively, how we tell a friend about a personal trauma may differ depending on whether the friend is someone from a reading group or someone we have known for years.

It is also possible that different practices of friendship are dependent on gender, class, sexual orientation, or ethnicity.16 Perhaps the expectation of how fervently friends will converse is a function of gender. Friedman notes that while friends generally talk to one another, this is not always the case. The friends, she writes, “might be men, and might have been raised to avoid emotional expressiveness and intimate self-disclosure even in close personal relationships. Such patterns of noncommunicativeness in close relationship seem not to be part of the cultural conception or idealization of friendship but to derive instead from other contingently related social conditions, such as the practices of masculinity” (1993:226). In contrast, I am suggesting it may be possible to describe the norms of masculinity as having been incorporated into the cultural practices of friendship. The question of the number (p.36) of different practices of friendship and whether those differences are correlated with such things as gender is an empirical question.17

The adverbial conditions associated with friendship are not, of course, always enabling. Sometimes they disable and preclude forms of friendship. Class distinctions, for example, may limit the interactions between individuals. Gendered distinctions may be governed by practices that are incompatible with the expectations for friendship or they may limit the ways in which individuals can be friends. Alternatively, how one interacts within a friendship may be conditioned by sectarian divisions. Crossing borders may entail grave risks for friends. Derrida, for example, notes that the dominant ethico-politico-philosophical discourse on friendship excludes friendships between women and between men and women. “This double exclusion,” he writes, “of the feminine in the philosophical paradigm of friendship would thus confer on it the essential and essentially sublime figure of virile homosexuality” (1988:642).18 If this is the theoretical world, the practice may be different. For example, Friedman writes that “friendship among women has been the cement not only of the various historical waves of the feminist movement but also of numerous communities of women throughout history who defied the local conventions for their gender and lived lives of creative disorder” (1993:248–49).

The idea, then, is that there are conventions for how we proceed with our friends that are different from how we proceed with acquaintances, strangers, teachers, police officers, and doctors and allow us to distinguish different practices of friendship. Our friends come to know us in terms of how we are friends; that is, how we go about living up to the adverbial conditions that define friendship (or a particular sort of friendship). It is, then, not inappropriate to use Oakeshott’s description of this sort of activity as a form of self-disclosure.19 Self-disclosure does not necessitate the opening of one’s heart to another, the confession of one’s inner-most secrets, the revelation of one’s authentic self, or the finding of an individual who replicates oneself (although friendship could entail any of those actions). Rather, self-disclosure is much more prosaic. It is to be found in the choices made while acknowledging and subscribing to a set of adverbial rules that hang together in a manner that composes a social practice. Self-disclosure is achieved (p.37) through the actions that we perform when living up to the conventions of friendship. It is judged in terms of quality of those performances and is frequently marked by such terms as whether one has been a good, bad, or merely fair-weather friend.

The idea of self-disclosure has been associated with friendship by a number of different thinkers. For example, Laurence Thomas argues that self-disclosure is the “predominant way in which companion friends can and do contribute to one another’s flourishing, where the emphasis here is upon improvement of character and personality” (1987:227). By self-disclosure, Thomas means the reciprocal confiding of private information about our lives.20 In contrast to Oakeshott’s usage, Thomas’s understanding of self-disclosure is a disclosure of secrets. By engaging in self-disclosure, we make ourselves vulnerable to one another and create the bonds of mutual trust that convey a sense of the special regard that deep friends have for one another (223).21 It also signals to those to whom we have confided intimate information that we are willing to accept advice and counsel from them because they know us in a particular way. In contrast, Thomas argues that individuals who are public about every thing in their lives—who are not discrete in their self-disclosures—have signaled to the world that they will accept advice from pretty much every one and that they have neither special regard for nor trust in anyone in particular. Consequently, Thomas believes that they are incapable of deep friendships (224).

It is important to note that Thomas’s account of friendship (and his conception of self-disclosure) is tied to an ideal of friendship and not friendship as such. Consequently, the differences in the meaning of self-disclosure can easily be understood as connected to differences in use. Perhaps what is more relevant for adopting Oakeshott’s use of self-disclosure is a particular criticism that has been raised of Thomas by Cocking and Kennett. They argue that the problem with Thomas’s account of self-disclosure as a condition for friendship is that it presupposes a very static and discrete conception of the self. As such, it precludes, what they take to be, the most significant facet of a deep friendship, namely that the self is itself a product of the friendship.22 They argue that our most important friendships rest on what they call a “drawing account of the self” (1998:505), in which we (p.38) are receptive to our friends’ interests in a way that can change our own character and we are susceptible to our friends’ interpretations of ourselves which can alter our self-conceptions.

Assuming that Cocking and Kennett are correct about the limits of Thomas’s conception of self-disclosure and assuming that their alternative deep conception of friendship is in fact a valued and valuable conception of friendship, does Oakeshott’s conception of self-disclosure foreclose the kind of ideal of friendship that they are suggesting? If it does, then Oakeshott’s notion of self-disclosure may be unnecessarily preempting particularly important understandings of friendship.

Let us begin by assuming that Cocking and Kennett are carving out a plausible ideal practice of friendship. In that practice, there are certain expectations for both what the friends want and how they are going to get it. First and foremost, what they must want is the kind of friendship that Cocking and Kennett outline. In other words, they want a “deep” friendship in which their self-understandings will be susceptible to the interests of their friends (and not just anyone) and they take seriously the interpretations of themselves that their friends offer. If deep friendship is both a goal and a practice in an Oakeshottian sense, then its pursuit will be governed by a set of motivations and adverbial conditions. Consequently, it is not surprising to see that deep friendship requires acting responsively, openly, attentively, willingly, endlessly, flexibly, and acceptingly in addition to being motivated by affection and the desire for shared experiences (Cocking and Kennett 1998:524). From an Oakeshottian perspective, what distinguishes deep friendship from other forms of friendship are both the possible motivations that the friends mutually recognize (e.g., other practices of friendship could be motivated by appreciative love, respect, or desire) and the list of adverbial conditions that govern the relationship.

Still, Cocking and Kennett may argue that the adverbial conditions set out above do not define deep friendship as they understand it. Deep friendship is not a function of how things are done. It is not a matter of protocols and procedures. Rather, it is a function of doing something quite specific with another human: it is the establishment of a relationship in which there is the “emergence and acceptance of a degree of direction and interpretation (p.39) of each by the other” (1998: 508). Perhaps this is true, but, as they also note, the problem with this condition is that it can be met by other sorts of relationships that are not friendships. For example, they point out that two psychoanalysts may also engage in interpretation and direction of each other. At this point, Cocking and Kennett argue that the difference between deep friendship and the relationship between the two analysts is that in the latter case their relationship will terminate once they achieve psychological health. In contrast, deep friendship is an endless engagement. When they begin to use this language, then it becomes clearer that the easiest way to distinguish deep friendship from other practices is to turn to either the motivations or the adverbial conditions that govern the relationship. In the case at hand, it is because the activity is endlessly engaged in. In addition, one could note that the analysts are motivated by psychological health and not affection. In any case, Oakeshott’s formal notion of practice is flexible enough to accommodate a malleable, nonstatic conception of the self.

Explanation, Justification, and the Practice of Friendship

The last sections of this chapter step back to ask what this view of friendship achieves. If the idea of friendship as a family of practices is conditioned by the mutual recognition of ordinarily appropriate motives and adverbial conditions is at all plausible, what does it provide us? Is it an explanation of friendship? Is it a justification of friendship? Is it a description of friendship? Focusing on the conditions of the practice is not the same as providing a causal explanation for why the practice exists in general, or why it has the form that it does in any particular culture or period. Understanding the conditions of friendship should help inform how friends are friends, but it cannot establish that Denny and Alan’s friendship, for example, was caused by the presence of these conditions or that these conditions caused Denny and Alan to maintain their friendship. Moreover, that certain kinds of motivations must be present or that certain forms of actions must be performed in order to have a friendship is not the same as providing the sort of justification (p.40) that says that friendship is a good thing or that a particular friendship is desirable.

Nevertheless, even though elucidating the conditions of friendship is not a project of causal explanation or justification, explanation and justification are not far off (and therefore this exposition of friendship is not merely a description). Because friendship requires the presence and recognition of certain motivations, referring to those motivations (say, for example, a friendship based on mutual affection and respect) can provide something of an explanation for why Denny and Alan became and are friends: they have affection and respect for one another. Moreover, setting out the adverbial forms of self-disclosure also explains something about how they are friends: e.g., they interact honestly, openly, intimately, caringly, considerately, and affectionately.23 Pointing to the motivations of a friendship may give us appropriate, perhaps even necessary reasons for a friendship, but those reasons may not be sufficient. Although they have affection and respect for one another, perhaps Denny is fighting on the other side of an armed conflict. Perhaps Alan has lied repeatedly to Denny. Perhaps Denny committed some horrible crime. Elucidating the character of our practices moves us only so far in sorting out those sorts of questions of justifying a particular friendship. The project of setting forth conditions of the practice of friendship will point to possible explanations of particular friendships, the plausibility of which can only be assessed empirically.

A practice-based account of friendship may also address features of the justification of friendship. If friendships draw on a certain repertoire of appropriate motivations (e.g., affection, fondness, desire, respect, appreciative love, care, need) and it is a prima facie good thing for human beings to act upon their desires and feelings, then it is a good thing for friendships to exist. These relationships allow us to cultivate and foster human desires and feelings. This is not to say that all desires and feelings are good or that all desires and feelings should be acted upon, but in the absence of countervailing reasons, I will presume that humans should not be frustrated or forestalled in doing what they want to do. It is one of the great joys of friendship that even if such relationships are not voluntary through and through, they do express a voluntary engagement with others that should not be disparaged or devalued. Friendship, then, is a prima facie good thing.

(p.41) Because our practices of friendship also entail subscribing to certain adverbial conditions, the justification for entering into and maintaining a friendship may be connected to the value we attribute to acting in particular ways with other human beings. To be a friend to another may mean that we have to pursue our interests and desires in ways that we would not other wise have to do or do in the same way. To strangers, acquaintances, colleagues, students, and fellow citizens, we have certain duties—for example, acting considerately, respectfully, courteously, peacefully, or perhaps professionally, helpfully, and so on. These are all worthy ways of acting and speaking with others and many of these ways of being are also shared with our friends or with certain sorts of friendships. That friendship adds to or intensifies these adverbial conditions by adding playfully, affectionately, caringly, critically, intimately, altruistically, etc., and that these are also valuable ways of interaction suggests that the practice of friendship opens up opportunities for other admirable forms of being in the world. Our friendships place demands on us that, while not always endorsable, do raise the bar. The fact that these conditions are not always met and that sometimes we fail to live up to the expectations of our friends can be seen as a tribute to the demands of the practices of friendship. In other words, these practices require that our actions live up to higher expectations of what we ordinarily understand to be worthy modes of action. This line of argument is further explored in chapter 4 where friendship is linked to the idea of individuality.

The Relationships View

In many ways, the position under consideration is consistent with the view that relationships of friendship themselves are seen as valuable. Seeing friendship as a family of practices reinforces the notion that human beings are creatures who value things in the world and value relationships between one another. Among those relationships is that of friendship. Of course, this does not mean that all friendships are valued equally or that friendship overrides all other values. Some friendships are closer and more significant. Other friendships may be judged to be harmful to the friends themselves or to other parties. Nevertheless, practices of friendship are important enough to us, that they seemed to be capable of generating what Samuel Scheffler calls (p.42) “relationship-dependent” reasons for action (2010:104). It is precisely because we are in a friendship that we have reasons to act in particular ways.24 Depending on the friendship, these reasons carry more or less weight vis-à-vis competing reasons for action. This relationships view is consonant with our commonsense view of morality; namely, what ever morality may entail, it should be able to incorporate reasons for acting out of friendship. Friendships are relationships that have intrinsic value which, in turn, can generate reasons (of variable magnitude) to act in a manner that favors our friends over others. Even if morality is deeply connected to impartiality, it should allow for reasons of partiality associated with friendship. This position raises the broader and more difficult problem of how the sort of impartiality prized in morality can accommodate reasons of friendship. This book begs off that meta-ethical question.25 Chapter 7 takes on the problem of impartiality, but only as it relates to impartial legal and political institutions that pressure our practices of friendship.

Our practices of friendship, however, are complicated by the fact that they are rarely just intrinsically valuable. From friendships also come the pleasures of doing things together, the advancement of the interests and goods of ourselves and others, the comforts of shared experiences, the expectations of assistance and protection, the blessed release of secrets held, and many other goods that could be characterized as external or extrinsic. As I have suggested, certain practices of friendship may be deeply connected to such goods. For example, both historically and culturally, various practices of friendship have been and are valued primarily because they yield pleasure for or are useful to the friends. In addressing the question of the value of these various practices of friendship, I am left with a rather ragged set of assertions. In some practices of friendships, reasons for action are derived primarily from the relationship itself (i.e., the intrinsic value of the friendship). In other practices, reasons for action are tied to the instrumental character of the relationship. In many practices of friendship, however, the value of the relationship is both intrinsic and extrinsic.

One way to clean up this conclusion is to deny the possibility that friendships are extrinsically valuable. This possibility strikes me as deeply implausible. Such a view would mean that the external goods of friendship (e.g., that friendship may be useful for or bring plea sure to the friends) tell us nothing (p.43) about the value of friendship. It is true, as I shall argue in chapter 3, that the role of those external goods in providing reasons for action is itself complicated by the fact that some of them have a friendship-undermining character. Nevertheless, it would be a mistake to claim that the external goods of friendship have no role whatsoever in the value we attribute to this relationship. The difficulty, tackled in chapter 3, will be to make sense of the nature of that role.

It is also possible to clean up the position by rejecting the possibility that friendships have any intrinsic value. Unlike the prior possibility, it is important to spend a bit more time sorting out this argument, particularly in light of a recent book by Simon Keller (2013) that powerfully argues just this case. Although what Keller has to say against the intrinsic value of friendship is framed in terms of a general argument about all partial relationships, I will focus on his arguments as they apply to friendship.

The heart of Keller’s criticism is that we value friendship because of its importance for our needs, interests, and flourishing and not because of the relationship itself (2013:76). Keller writes, “If you ask why you have reason to do something special for your friend, then the answer, I say, is that your act would advance your friend’s good, in some respect or other; it would further his best interests, manifest concern or respect for his autonomy, or help him to become a more excellent human being” (113). To grab onto the relationship itself (as writers such as Scheffler does) and its purported intrinsic value is to grab onto the wrong end of the stick. The individuals within the relationship are what matter, not the relationship as such. From Keller’s perspective, to say that you have reason to do something special for an individual because she is your friend is both misleading and inadequate. It is misleading because the relationship itself does not itself carry any weight and inadequate because the real reasons have to be grounded on the value we attribute to the individuals concerned and what ever it is that they may merit. In the end, Keller argues that there is something “misanthropic” about the relationship view. More strongly, it “drags human relationships from their human context” (77).

Many of the arguments that Keller presents in support of his position lead to a reductio which strips away from a relationship all of the good consequences for the individuals involved in order to reveal the empty, inhumane (p.44) nature of the friendship relation itself. In response, I do not argue that all practices of friendship are based on the intrinsic value of the relationship, but the intrinsic value of a friendship can certainly be a source of partial reasons for many practices of friendship. My defense of the relationships view will entail variations on the theme that Keller’s reductio should be resisted. In other words, we cannot understand a relationship of friendship without understanding its motivational and action-related conditions as they apply to the specific individuals involved in the relationship. By stripping the notion of friendship to what Keller sees as its “bare structure,” one is left with something of a “straw friendship.” I would imagine Keller’s responses to this move would, in part, be to claim that I have simply brought into the idea of friendship extrinsic matters of context and effects that are not necessarily part of the “bare relationship.” In other words, my response has stacked the deck in favor of the possible intrinsic value of friendship. At this point, it is difficult to proceed without having a better sense of what the bare relationship of friendship actually entails. In the following section, I will explore Keller’s arguments against the view that friendships have intrinsic value.

Keller’s Case Against the Relationships View

The first argument against the relationships view is what Keller calls the “At Least It’s a Relationship” argument. The basic idea is that if a friendship was utterly dysfunctional and entirely lacking of any external value, then, according to Keller, we “should not say, ‘I can see that this friendship is not doing any good for anyone, but at least it is a friendship. At least it has that going for it. So there is something good about the relationship, even if there are bad things about it too.’” In other words, we should not say these things because in the absence of any external goods, there would really be nothing “to recommend the relationship” (2013:57).

Keller, of course, is correct. But, on the other hand, we should not say the opposite either. For example, we should not say, “I see this friendship between Greg and Wilson as abusive and harmful, but at least it’s good for the individuals involved.” We should not and would not say such a thing largely because it is not clear if their relationship would even qualify as a (p.45) friendship. If interpersonal friendships presume that the friends hold certain beliefs about the motives of one another and that they mutually subscribe to a friendly manner of acting toward one another, then “friends” who believed and acted other wise (for example, acting in an abusive manner) would be “friends in name only.” Moreover, this response would accord with the idea that friendship must contain some element of intrinsic importance. Keller’s judgment in the original case is correct, not because the friendship has no intrinsic value, but because either it is not a friendship (given that the ordinarily appropriate motives and action-conditions are not met) or because the intrinsic value of the relationship is utterly overwhelmed by the situation he describes. For example, suppose that Pretty Boy and Jesse are both real friends, but inept bank robbers. We would not recommend their friendship precisely because it has bad consequences for us and perhaps because it has bad consequences for them. They may eventually come to the conclusion that they need to end their friendship (they are constantly getting into trouble and thrown into jail), but that may be because those external effects have overridden the intrinsic value that they attribute to their friendship. From the perspective of the friends, it does not seem wildly implausible to think that they could still believe the relationship has some value to them (they are admired, liked, respected by their friend, and treated in a friendly manner) even though that value is overwhelmed by the unfortunate externalities suffered by others and themselves.

Keller’s second argument against the intrinsic value of friendship is counterfactual. Imagine a people who lack the bundle of practices of friendship that have intrinsic value. In such a society, there would still exist friendships, but only of a certain sort. Perhaps people act benevolently toward one another because they enjoy it or because they believe that such benevolence will lead to their flourishing or to eternal salvation. In this imagined case, Keller argues that we should not regret the absence of the intrinsically valued friendships, nor judge such as society worse off than our own. Consequently, we would have no reason to rearrange such a society to create friendships that have intrinsic value. Keller writes, “Whether the society’s arrangements are good or bad, and whether there is reason to change them, are matters of how the people within the society fare, not of how the bare structure of the relationships in which they participate” (2013:59).

(p.46) But what is the bare structure of intrinsically valued friendships? I have suggested that in such relationships two individuals mutually recognize the appropriate motivations of friendship, subscribe to a set of adverbial conditions to guide their action and find something of intrinsic worth in the satisfaction of those conditions. In contrast to Keller, I believe that society is richer if that sort of relationship is possible. Does that belief commit me to the further claim that I should have reason to transform a society that lacked that sort of friendship? Perhaps. If achieving that possibility entails clicking my heels three times, that is one thing. If, as some sociologists argue, that possibility could only arise with the emergence of a free-wheeling capitalist system and that system results in the impoverishment of millions (perhaps it does or perhaps it does not), that is another thing. As a historical hypothesis, it is possible that in the West, seeing friendship as an intrinsically valuable relationship is something that emerged only in early modernity. Nevertheless, it is plausible to think that the world is a better place with the possibility of intrinsically valued practices of friendship, ceteris paribus.

A third argument concerns how we judge whether something has intrinsic value. Keller argues that we perform this task by separating what is claimed to have intrinsic importance from other values and “ask whether it holds value regardless of its consequences and context” (2013:61). Whether friendships can have intrinsic value and serve as the basis for special reasons depends on the possibility of being able to separate friendships from their context and other values. In Keller’s view, this makes no sense. Whether a friendship is valuable cannot be separated from its context and effects. We simply do not set aside the external effects of the relationship in judging its worth.

In response, I think that Keller is correct with regard to our judgments about many friendships. We look to the presence or absence of external goods to determine whether our friends are behaving as they should. Consequences matter. If Bert becomes a stick in the mud and is simply no fun, then a friendship with Bert motivated by the plea sure of dancing in the park may evaporate. The situation is complicated by the fact that we also look to context and consequences to assess our beliefs about motivations of (p.47) our friends. So, we may come to believe that what was a friendship was merely an exploitative relationship and hence no friendship at all. Alternatively, the idea of the intrinsic value of friendship makes more sense of the possibility that individuals may sometimes stick together through thick and thin. For some, the true test of a friendship may have been found in a context of pain and deprivation (such as during war or fighting wildfires) in which the friends continued to act as friends and are motivated by the appropriate sentiments of friendship. In such circumstances, the irrelevance of external goods to indicate the value of the friendship seems to point to its intrinsic importance.

In a fourth argument, Keller notes that sometimes we have reasons to end a special relationship: “Though I love my friend and he loves me, we may realize that we are incompatible, always interacting in ways that leave each of us bitter and depressed” (2013:61). As I understand this argument, Keller is claiming that this possible reason for ending the relationship cannot be produced by the friendship itself. You may have reasons to sacrifice a relationship for the sake of the individuals, but “you do not have reasons to sacrifice individuals so that the relationship, or its putative intrinsic value, can survive” (62).

Once again, Keller is correct in noting that there may be powerful reasons based on the functioning and consequences of the relationship for ending a friendship. It is also important to note that individuals may remain in relationships that have unfortunate consequences for bad reasons—inertia, maintaining a public image, weakness of will. On the other hand, could it not be the case that one of the reasons individuals remain in a relationship whose effects are troubling is because of the intrinsic value of the relationship? Overall, they may be mistaken in thinking that this is a friendship that should be preserved, but they may not be mistaken in thinking that the relationship is itself important. Does not the intrinsic value of friendship better account for why friends sometime stick together beyond the point that they should?

Keller’s fifth argument is that while friends sometimes act for the sake of a friendship, we more often act for the sake of the individual with whom we are friends. More strongly, “A friend who is always thinking of improving (p.48) your friendship, a colleague whose main concern is with the value of collegiality, a parent who thinks mainly of how important it is to have a good relationship with his child—all of these characters are annoying to have around, and all of them seem to be missing what really matters in their relationships. In a relationship with such a person, you may feel that he cares less for you than for his relationship with you” (2013:63). Keller’s singular insight here is worth the price of his book. He goes on to argue that the relationship itself usually does not motivate us to act (although Keller admits that sometimes it does, but this is exceptional) and if it is motivating us, it will prove to be an irritant to our friends. In contrast, what usually motivates us or what usually should motivate us are the actual interests and needs of our friends. For example, if you hear that the town in which your friend lives has been flooded, you do not call her up saying that your friendship depends on it. Rather, you call because you want to be sure that your friend is okay. It is not the value of the relationship that motivates (or should motivate) us to act.

I understand Keller’s argument to be the sort of objection that could be made to someone who loves being in love as opposed to someone who loves another person. His argument is that the relationships view is like the former stance. Attributing intrinsic value to friendship is nothing more than loving to be in love. If this is true, then once again we can wonder if there is a relationship of friendship. If Wilson cares more about the “relationship” (where the word relationship does not include an implied reference to Greg) than he does about Greg, then it is unlikely that Greg would believe that Wilson is a particularly good friend. Absent is a shared belief that the friends are motivated by the appropriate sentiments of friendship. Absent a belief that those motives are present, there is no friendship. In this situation, we should not say that the friendship with Greg is motivating Wilson because it is unlikely that he really has a friendship with Greg. Similarly, we should not say about the person who is in love with being in love that she is necessarily in love with her beloved.

Keller could argue that this is precisely why the relationships view does not work or why, if the relationships view works, it must separate the true reasons for action (the intrinsic value of friendship) from the motives for the action. In contrast, I am calling into question whether there is a relationship (p.49) unless the description of the relationship specifies the individuals involved. To put this point another way: The reason why we must specify the individuals involved is because they exist in a special relationship to us. It is the intrinsic value of this friendship (and not any friendship in existence), meaning this relationship with this person (and not a stranger for instance), that explains why, when Cuddy hears that the town in which her friend Wilson lives has been flooded, Cuddy can explain that she wants to call to make sure that Wilson is okay. Wilson is not anyone, but her friend. To those who know Cuddy and Wilson, this friendship is already bound up with the person “Wilson.” To those who do not know Wilson, she would say that he is her friend. It is the relationship with Wilson that motivates her action.

Perhaps Keller would be unconvinced. For Cuddy to say Wilson is her friend is not necessarily to say that Cuddy has an intrinsically valuable friendship with him. She could also be saying that she is especially concerned with Wilson’s value “as an individual in her own right,” which is Keller’s preferred account, which he calls the individuals view (2013:74). On this view, Cuddy’s motives for acting are “fully shaped” only because she knows that it is Wilson. To make this point clearer, one could imagine a situation (an example that Keller does not use, but I hope gets to the spirit of his argument) in which Cuddy learns that a town has been flooded and the list of the casualties have been posted in the local hospital. Cuddy has no reason to check the list because she does not know anyone from the town. An acquaintance of hers, however, tells her that, “I think a friend of yours was visiting that town on the day of the flood.” When asked who it was, the acquaintance says, “I don’t remember his name.”

In this case, Keller would argue that on the relationships view, knowing that it is “a friend” will not be enough to fully shape Cuddy’s motive. While Cuddy now has a reason to check the list, she will be “groping for more information” and feeling that she does not “know why exactly, this is such an important act to perform.” Because she does not know who that friend specifically is, she would not fully know what her reason is for checking the list of casualties. However, if Cuddy’s acquaintance suddenly turns around and shouts, “Oh, I remember, it’s your friend Wilson,” then Keller argues that her “motive will change in quality” (2013:96). Knowing that a relationship exists (“I think a friend of yours was visiting the town”) will not focus (p.50) and fully shape Cuddy’s motive. At a phenomenological level, that focus happens largely because it is Wilson and not merely “a friend.”

Because I have suggested that our relationships of friendship include reference to specific individuals, there is not much to distinguish my version of the relationships view from Keller’s in the example of checking the list of casualties. On my formulation, once Cuddy finds out that it is Wilson, then her motive will be more focused and shaped, but it will be focused and shaped because she has a specific sort of friendship with Wilson. In large part, because of the variety of ways in which we can be friends, Cuddy’s motives for looking at the list will be shaped by the sort of friendship she has. Not knowing the specific friend means that the quality of her motive is still open. Suppose Cuddy’s acquaintance, who is bad with names, says, “Your best friend—I don’t remember his name—was visiting the flooded town,” then, it seems, that would be enough to fully form her motive. Perhaps one can read this possibility as either an argument in favor of the relationships view because it is “your best friend,” or in support of Keller’s individuals view because “your best friend” clearly identifies a specific individual, namely Wilson. In either case, the relationships view seems to be able to accommodate what Keller is trying to capture with his individuals view.

Keller offers one final argument against the relationships view: The relationships view cannot explain why my valuable friendships are of special significance to me, but Cuddy’s friendship with Wilson is not equally valuable to me. If the relationship of friendship is the thing that counts and I have friendships and Cuddy has friendships, then why should I not care for Cuddy’s friendships as much as my own (Keller 2013:118–19)? As Diana Jeske notes, we can ask not only why we form friendships in general and why we form this or that friendship, but also what reasons we have to care for a person with whom we have an intimate relationship that go beyond the ways in which we care about people with whom we are not intimate (2008:46). Why should we owe more to our own friends than to other friendships? The relationships view simply asserts that friendships provide agent-relative reasons in support of their special character. At best, it argues that we value our own friendships over the friendships of others because we subjectively value our own friendships. In short, it fails to provide (p.51) an answer other than that our own friendships just are more important to us (Keller 2013:119).

The answer to Keller’s query may be found in the structure of friendships and perhaps could be called the practices view. I have claimed that contemporary practices of interpersonal friendship share a structure in which the existence of the relationship depends on the following conditions:

  1. 1. The friends entertain the appropriate sentiments of friendship toward one another.

  2. 2. The friends believe that the others’ actions are motivated by those sentiments of friendship.

  3. 3. The friends subscribe to a shared set of adverbial conditions that structure how they act toward one another—i.e., they interact in a friendly manner.

Keller asks why you should respond differently to the individuals “with whom you share special relationships, even though those individuals are no more inherently valuable than other individuals” (2013:124). Why should Cuddy’s friendship to Wilson matter more for Cuddy than what ever friendships I may have? The practices view suggests the following answer: Cuddy cannot have a friendship with Wilson unless she accords some sense of appropriate priority to her own relationship with Wilson contingent upon Wilson responding in a similar way to Cuddy. That is what it means to participate in a practice of friendship. In part, it is a matter of understanding how the relationship is itself constituted by the manner in which individuals are expected to act. Cuddy should give special significance to Wilson’s interests because that is part of the practice to which they subscribe. On the practices account, Cuddy would have a very difficult time if she were to attempt to have a friendship without subscribing to the conditions under which it is possible to have a friendship.

From a practices view of friendship, the question is not whether the relationship should generate special reasons for action. That is what the relationship is all about. To separate the possibility of special reasons from a practice of friendship is to misunderstand what it means to be in a friendship. To make this separation would be like saying, “I want to make a promise to be here tomorrow at 9:00, but can I get rid of all that baggage (p.52) associated with creating an obligation?” From the practices view, the question about the partiality of reasons of friendship becomes a much broader question of why we should have a set of practices that create reasons of partiality. Asking this question, however, brings us to the limitations of a practices approach to friendship: namely, it will not yield an explanation or justification for the family of practices of friendship nor the shape of particular practices in a particular culture or time period. What a practices approach will stress is the idea that there are many different kinds of friendship. Not all friendships are intimate relationships. Not all friendships turn on the sentiment of affection. Not all friendships are based on the same level of concern (e.g., friendships motivated by desire versus those motivated by affection). The range of relationships that we call friendship, the kind of reasons or the weight of those reasons that support those relationships may differ quite a bit from relationship to relationship and over time within a relationship. Instead of focusing on whether there are reasons to give greater attention to friends over others or whether there exist special duties, I assume that such reasons and duties exist but that their weight varies. Assuming such reasons and duties, I argue in chapter 3 that how they function depends on to whom they are addressed.

A friendship, then, is a rather complicated relationship—historically contingent, dependent on a range of mutually recognized motivations, and realized in a variety of adverbial forms of self-disclosure, all of which can define distinctive and multiple practices of friendship. These practices are not united by any one sentiment or reason, by any given substantive actions, or by certain essential adverbial conditions. Rather, the practices of friendship condition friendships (as opposed to some other relationship) because they bear a family resemblance to one another. They share similarities that crop up and dis appear. They do share the feature of being practices, but that feature does not distinguish them from other sorts of relationships. Seeing friendship as a family of social practices may help alleviate the temptation to close down or disparage different ways in which we can be and have been friends as well as open us to the meaningful use of friendship in other cultures and times. What distinguishes friendship from other practices is the repertoire of what are taken to be the appropriate motives and norms of (p.53) self-disclosure that we recognize and enact as we more or less successfully make, lose, strengthen, or weaken our friendships. Further attention, however, needs to be devoted to the broader reasons for friendship and the place of self-interest and obligation as motivations for friendship. These issues are considered in chapter 3.


(1.) This chapter’s focus is on the work of Elizabeth Telfer, but Marilyn Friedman provides another such example when she writes, “The relationship of friendship is motivated by mutual affection and positive regard. Without such a mutual interest, a relationship would simply not constitute a friendship” (1993:224). For an ancient and concise definition of friendship, one may turn to Cicero’s statement: “Now friendship is just this and nothing else: complete sympathy in all matters of importance, plus goodwill and affection and I am inclined to think that with the exception of wisdom, the gods have given nothing finer to men than this” (1991:80). Fernando Santos-Granero suggests that Montaigne’s view of (p.290) friendship has become enshrined as the ideal Western conception of friendship (2007:8). In response to that ideal, Santos-Granero argues every one of the aspects that Montaigne sees as central to friendship (that it is voluntary, unselfish, intimate, informal, between equals) “has been attacked by anthropologists on the basis of cross-cultural analyses” (9).

(2.) For Marilyn Friedman, “Affection encompasses the fond and tender feelings of liking and love with which we respond to (some) other persons. Affection need not involve any judgmental or evaluative component” (1993:193).

(3.) In the “Lysis,” Socrates explicitly rejects the idea that mutual liking (and hence mutual recognition of the sentiments of friendship) is necessary. As Julia Annas, notes, however, this is an artifact of the notion of philia in that its “nonmutual” senses could apply to quail, wine, and philosophy—all things that could be liked without liking in return (1977:533). In contrast, she argues that Aristotle takes those nonmutual senses of philia as secondary. Aristotle writes, “Love for a soulless thing is not called friendship, since there is no mutual loving, and you do not wish good to it” (1985:1155b25–30). The mutuality condition means that for Aristotle friendship is not the same as simple benevolence or wishing someone well. Aquinas also talks of the necessity of mutual recognition when he writes, “neither does well-wishing suffice for friendship, for a certain mutual love is requisite, since friendship is between friend and friend: and this well-wishing is founded on some kind of communication” (quoted in Schwartz 2007:172). The political significance of the mutual recognition of the appropriate sentiments of friendship will be discussed in chapter 4.

(4.) For Aristotle, friendship “is said to be reciprocated goodwill” (1985:1155b34). Toward one’s friend, one must “wish good for his own sake” (1155b30), and this attitude needs to be recognized by both individuals. According to Grunebaum, this attitude is encompassed by the notion of caring for another, which is not necessarily the same thing as fondness or affection (Smith also notes that “Aristotle does not focus on this sharing of an emotional life” [2011a:73]). Feelings are not as important to Aristotle as the motivation of goodwill. Despite this absence, Grunebaum notes that “it is almost impossible to imagine friends, as Aristotle understands them, wanting to spend all of their lives together engaging in joint virtuous activities and failing to like each other or without growing in affection for each other” (2003:39). In this case, affection is not a necessary condition for friendship but a likely result of it. In contrast, Konstan notes that philia “means affection in Aristotle and in Greek generally” (1997:73).

(6.) This sort of move in which a thinker sets out what is understood as a complete or true or more worthy conception of friendship is frequently accompanied by an addendum or sidebar in which there are other incomplete, degraded, or merely common understandings of friendship. Nevertheless, it is rare for a thinker who makes this sort of a claim to argue that the latter do not qualify as (p.291) friendships. Jeske comes close to doing so when she suggests that for Aristotle “friendships for utility or plea sure are friendships in something like the sense that rubber ducks are ducks” (2008:44). For a contrasting view, see Annas (1977:546).

(7.) In his development of an Arendtian conception of friendship, Jon Nixon argues that Arendt saw her friendships as based on mutuality of respect. He then goes on to conclude that it was therefore based on equality (2015:8). It is not clear, however, whether equality necessarily follows from that motivation. If, as Nixon suggests, Arendt and Heidegger negotiated a friendship in the postwar period and if one of the motives of that friendship was respect, it is not clear that Heidegger’s respect for his former pupil was at the same level as Arendt’s respect for him.

(8.) Lynch notes that “modern friends do not necessarily develop friendships with those whose goodness they admire or esteem. Such admiration and esteem are not sufficient on their own to ensure the development of a friendship” (2005:31). Lynch is correct if she means that admiration or esteem by themselves need to be supplemented by mutual recognition and certain actions.

(9.) On Blum’s account, “not every one does have friends in the same way” (1980:71). For Blum, however, the difference is in how they care or in the level of caring, where care entails a willingness to work through trouble. Can one have a friend and not care for her (in this sense of care)? A friendship, say, between selfish persons may be largely based on their enjoyment of one another’s company—they are fun to be with. It may not be a very valuable or deep friendship, but it can be a friendship nonetheless. In contrast, the idea here is that in addition to levels or kinds of care, friendship can be motivated by a range of other reasons and feelings.

(10.) For Kant, the roles of need and confidence place friends in a bind. He argues that a true friend is one in whom I have confidence that she will help me at the drop of a hat. But if I am a true friend of hers, then “I ought not” expect such help from her or place her “in any quandary.” Kant notes, that “I must have confidence only; rather than make demands, I ought to bear my own troubles” (1991:213). Similarly, my friend should feel the same toward me. In this form of friendship, there exists a benevolent disposition to aid combined with a disposition to abstain from calling on such aid: “The finest sweets of friendship are its dispositions of good-will; and on these we must avoid encroaching.” This bind, however, may be the result of seeing friendship as incompatible with notions of self-interest and utility—a view that will be challenged in chapter 3.

(11.) This understanding of our practices of friendship illuminates the character of online friendships. In cases where the individuals do not mutually recognize the motivations for their interactions, they may acknowledge one another as “friends” and even “unfriend” one another, but more is needed for a friendship. In these cases, the “friend” is more like a “befriend” in which what exists is a (p.292) friendly relationship and not a friendship. On the other hand, online friends may quickly form the belief that they enjoy one another’s interactions and simply desire to share their posts, pictures, etc. When this mutual recognition of motivation happens and is accompanied by a mutual subscription to a set of adverbial conditions for how they should interact, then an online friendship has begun. Although it is arrived at in a somewhat different way, this conclusion is no different than Ethan Leib’s conclusion that some online friendships will qualify as friendships and some will not (2011, 25). The difference is that for Leib, “Affection is what matters so much in friendship, and online ‘friendships’ do not easily translate or transmit warmth” (28), whereas affection (and the account offered here) is merely one motivation in a larger repertoire that can drive friendship.

(12.) In her biography of Hannah Arendt, Elizabeth Young-Bruehl notes that Arendt described Rahel Varnhagen as “my closest friend, though she has been dead for some one hundred years” (1982:56). Varnhagen, the subject of Arendt’s first book, of course, could not say the same about Arendt. Given the impossibility of mutuality, it would seem strange or metaphorical for a third party to discuss the Arendt-Varnhagen friendship.

(13.) A “secrets” view versus a “mirror” view of friendship is found in a discussion by Dean Cocking and Jeanette Kennett.

(14.) At least in quantitative analyses of friendship, Pahl notes “The fact that there are different kinds of friends is rarely considered” (2000:7).

(15.) While he does not say anything about the nature of the considerations associated with friendship in On Human Conduct, Oakeshott himself notes that the intercourse of friends is itself qualified by such adverbial conditions (1975:57).

(16.) Andrew Sullivan writes,

(17.) For an alternative perspective that sees gendered distinctions as connected to other variables, see Pahl (2000). The questions surrounding how friends interact are also cultural and historical. Employing Alan Bray’s work on medieval friendship, Vernon notes that the physical aspects of how to be friends changed significantly in Eu rope. The practices of medieval friendship included a bodily intimacy that was ultimately confined to a heterosexual marital space in the West (2010:175).

(18.) In the ancient world, “The term amica or ‘girlfriend’ had pejorative connotations and husband and wife were no more likely to be described as friends in Latin than in Greek” (Konstan 1997:146).

(19.) Within psychology, much has been written about the relationship between self-disclosure and friendship. See Cozby (1973) and Schubertwalker and Wright (1976).

(20.) As Konstan notes, this conception of friendship is very old. St. Ambrose wrote, “Therefore a friend hides nothing, if he is true: he pours forth his mind, just as Lord Jesus poured forth the mysteries of his father” (quoted in Konstan 1997:150).

(21.) Cocking and Kennett call this the “secrets view” of friendship (1998:514). Although she does not use the language of self-disclosure, Lynch also sees friends as having or aspiring to have “an intimate knowledge of one another” (2005:4). She argues that “the possibility of friendship rests on our acceptance of a fiction—or what Derrida argues is an illusion—of connection, despite the impossibility of any complete or sustained connection between friends” (95).

(22.) This difficulty also plagues what they call the mirror conception of friendship in which one’s closest friends reflects one’s own traits (Cocking and Kennett 1998:505). For a psychoanalytic account of mirroring and its relationship to friendship see Lynch (2005:171–75).

(23.) The possibility that there are no essential motivations or adverbial conditions for friendship and that the relationship is both interpretative and willful may go some distance in accounting for the difficulties that social scientific explanations of friendship have had. Pahl notes, “Friendships, more perhaps than any other aspect of our social lives, have eluded the attempts of social scientists to be classified and codified” (2000:142). This problem comes again in chapter 6 when considering whether the law should attempt to foster and protect interpersonal friendships.

(p.294) (24.) In Scheffler’s view, valuing relationships noninstrumentally also yields special responsibilities or special duties (2001:97). Agreeing with Keller, I do not believe that such relationships necessarily yield duties (Keller 2013:51, 65). The relationship between duties and friendship is considered in chapter 3.

(25.) Blum argues that friendship tends to raise difficulties for any ethical theory that stresses impartiality, such as Kantianism or utilitarianism. In contrast, virtue-theorists are less puzzled by the weight accorded to friendship insofar as friendship is understood as an exhibition of human excellence. However these ethical theories deal with conflicts between our responsibilities toward our friends versus the responsibilities to others, a discussion of the practices of friendship is unlikely to provide a comprehensive solution to those conflicts. More generally, I am inclined toward Keller’s view on this matter that it may be impossible to deduce the moral value of a partial relationship such as friendship from moral standards of impartiality, such as those found in consequentialism (Keller 2013:150). For me, this appears to indicate a deep form of pluralism within our morality.