Here’s a draft of the first section of a work-in-progress, tentatively titled “Friendship as a Model for the Other Virtues,” which I will be contributing to Jon Webber and Alberto Masala‘s collection, The Architecture of Personality and Ethical Virtue.
Normally, I throw in a thematic picture or two. In this case, 15 minutes of searching on google images failed to turn up anything that was maudlin to the point of being nauseating, so it’s just text.
As always, questions, comments, criticisms, etc. are most welcome.
Friendship as a model for the other virtues
Draft, 25 March 2013, please do not circulate or quote without permission
Friendship might seem like a bizarre virtue – or not a virtue at all. In Aristotle’s discussion of the various moral virtues in the Nicomachean Ethics, we see courage, temperance, generosity, magnificence, magnanimity, pride, wit, and justice. These would all seem to be, in the first instance and primarily, monadic properties of individual agents. To be courageous is to be disposed to think, feel, desire, deliberate, act, and react in characteristic ways. Even if no one else is courageous, it would still be possible – though extremely difficult – for you to be courageous. Of course, if there are no threats to be opposed, you may never have a chance to express your courage. Furthermore, it would surely be easier to develop courage in the company of others who either are or strive to be courageous. And it may also be easier to develop or sustain courage when others think of you as courageous and signal those thoughts to you. But, one might think, even if none of these enabling facts obtains, it would still be possible, conceptually speaking, to be courageous. Or consider generosity. To be generous is to be disposed to think, feel, desire, deliberate, act, and react in characteristic ways. Even if no one else is generous, it would still be possible – though extremely difficult – for you to be generous. Of course, if there were no other people who needed or wanted or would appreciate what you have, or if you were so down on your luck that you had no resources to offer, you may never have a chance to express your generosity. Furthermore, it would surely be easier to develop generosity in the company of others who are or strive to be generous (and, for that matter, grateful). And it may also be easier to develop or sustain generosity when others think of you as generous and signal those thoughts to you. But, one might think, even if none of these enabling facts obtains, it would still be possible, conceptually speaking, to be generous.
Friendship appears to be different. It seems to be, in the first instance and primarily, a dyadic relation between two people. To be a friend is to be disposed to think, feel, desire, deliberate, act, and react in characteristic ways towards a particular person, who is likewise disposed to think, feel, desire, deliberate, act, and react in those same characteristic ways towards you. If no one else is a friend, then it is conceptually impossible – not just difficult – for you to be a friend as well (NE VIII:2, 1155a). It is not just easier to develop friendship in the company of others who are doing so as well; it is in fact impossible to become a friend without there being someone else who also becomes a friend, namely your friend.
Being a friend isn’t just a matter of your first-order cognitive, affective, evaluative and behavioral dispositions; to be a friend means to have particular de re attitudes towards another person (your friend), and for that person to have congruent de re attitudes towards you (NE VIII:2, 1156a). That is, for you to be my friend, you need to think of me as your friend, to wish me well for my own sake, to wish me well in virtue of my good character (or, in other types of friendship, in virtue of my contributing to your utility or pleasure), and so on. Likewise, I need to think of you as my friend, to wish you well for your own sake, to wish you well in virtue of your good character (or in virtue of your contributing to my utility or pleasure), and so on.
But even that is not enough: not only must both you and your friend have these attitudes, but the existence of these attitudes must be mutual or perhaps even common knowledge between you (NE IX:5, 1166b). If I wish you well for your own sake and in virtue of your good character, and you wish me well for my own sake and in virtue of my good character, but neither of us knows how the other feels, we are not friends. Instead, we merely harbor mutual but unrecognized good will towards one another. To be your friend, I need know that you wish me well for my own sake and in virtue of my good character, and you need to know that I wish you well for your own sake and in virtue of your good character (NE VIII:3, 1156b).
In fact, even that is not enough. We could satisfy this description and yet still not be friends. If we each think of each other in this way, and each find out through reliable testimony that the other does as well, it would still seem strange to say that we are friends. I also need to know that you know that I wish you well for your own sake and in virtue of your good character, and you need to know that I know that you wish me well for my own sake and in virtue of my good character. It’s arguable that even this is not enough, and that what needs to hold is that we share common knowledge of our attitudes: I know that you wish me well for my own sake and in virtue of my good character, and that you know that I know that you wish me well for my own sake and in virtue of my good character, and that I know that you know that I know that you wish me well for my own sake and in virtue of my good character, and so on. I will not press this point here, for even if all that’s required is two orders of mutual knowledge (I know that you know, and you know that I know), my point still holds that friendship is an interesting virtue because it requires reciprocated de re attitudes and some kind of mutual recognition of the existence of this reciprocation. Friendship is not just causally but constitutively dependent on there being another person who has the same virtue. It is also not just causally but constitutively dependent on there being another person towards whom you harbor certain de re attitudes, and who reciprocates them. It is not just casually but constitutively dependent on your thinking of yourself as someone’s friend. Finally, it is also not just causally but constitutively dependent on there being between you and your friend at least two orders of mutual knowledge of these attitudes.
One might worry that these arguments press too hard on the relational aspects of the virtue of friendship. Surely, one might think, I can be a friendly person even if everyone else is an asshole and either snubs or betrays my attempts at friendship. There is an important sense in which, even in such an unlucky social environment, I can still be a friendly person. This is a fair point, and one which should lead us to distinguish between the disposition or trait of friendliness or agreeableness, which is arguably a monadic property of an individual, and the virtue of friendship, which clearly is not. One test that seems to do a good job of drawing this distinction is to ask whether the person in question is friendly or a friend. There is a double dissociation between these. Someone might be friendly but still not have any friends. Conversely, it’s possible for someone to be dispositionally grumpy or unagreeable but nevertheless to be a friend.
In this paper, I explore the prospects for using the interesting features of friendship identified above as a model for all moral virtues. This exploration is motivated on three independent grounds. The first is historical. Of the ten books of the Nicomachean Ethics, fully two are devoted to discussing the virtue of friendship. This is twice as much attention as justice receives, and as much as all of the other moral virtues combined. Yet contemporary neo-Aristotelian treatments of virtue rarely address friendship, and give it short shrift when they do. Annas (1993, pp. 249-260) devotes twelve of the five-hundred or so pages of her book to friendship, and mentions it only twice in her (2011, pp. 76, 151) book. Geach (1977, p. 80) mentions it once. Hurka (2001, pp. 35-6, 200) mentions it in only a couple passages. Hursthouse (1999, p. 11) calls friendship an “awkward exception” because it is relational. MacIntyre (1981, pp. 123, 156-8), the granddaddy of the virtue ethics revival, mentions friendship only twice. Russell (2009) barely engages friendship in his massive tome. Slote (2001) only discusses friendship in the context of broader discussions of love, community, and achievement. Snow (2008), despite the fact that her book is titled Virtue as Social Intelligence, never once uses the word ‘friendship’. Adams (2006, pp. 25-7, 69-92) is the exception that makes the rule. The main topic of discussion in the contemporary literature on friendship is the extent to which various moral theories induce “moral schizophrenia” by calling on us to be motivated by abstract principles – such as maximizing good outcomes or acting from duty – that seem incompatible with the warmth and intimacy of friendship (Stocker 1976). It would be surprising if this bias in the scholarship did not distort our understanding of character and the other virtues.
Second, as I have argued elsewhere (Alfano 2013, forthcoming), there are empirical grounds for doubting whether virtue as conceived in currently-dominant neo-Aristotelian theories is an achievable ideal. As John Doris (1998, 2002), Gilbert Harman (1999), and I (2013) have argued, most people’s behavior does not seem to be structured by robust, global dispositions such as honesty – at least when they are tested in a decontextualized laboratory context. Seemingly trivial and normatively irrelevant situational influences, such as mood modulators and ambient sensory stimuli, predict and explain people’s cognitive, affective, evaluative, and behavioral responses as well as – and sometimes better than – personality variables. This is not the place to delve deeply the dialectic between philosophical situationists and defenders of neo-Aristotelian ethics. Instead, I merely want to point out that there is suggestive empirical evidence – much of which I canvass in my (2013) book – for the phenomenon of factitious virtue. A factitious virtue simulates its neo-Aristotelian counterpart through the stabilizing influences of self-concept and social expectations. Someone may not be disposed to think, feel, and act as a generous person would think, feel, and act except insofar as she both thinks of herself as generous (self-concept) and knows both that others think of her as generous and that they know that she knows that they think of her as generous. When this happens, she does not have the trait of generosity construed on neo-Aristotelian grounds, but she does have factitious generosity.
Factitious generosity thus mirrors several of the more striking structural features of friendship. You cannot be a friend unless another person thinks of you as a friend, and you know that they do, and you know that they know that they do. You cannot be factitiously generous unless another person thinks of you as generous, and you know that they do, and you know that they know that they do. You cannot be a friend if you don’t think of yourself as a friend. You can’t be factitiously generous if you don’t think of yourself as generous. It may be possible to satisfy both the historical motive and the empirical motive by reconstructing all moral virtues on the model of friendship – as essentially and constitutively social.
The third motive for exploring the friendship model of virtue is the compelling evidence that has begun to pile up for the idea that many seemingly individual psychological phenomena are better understood as extending beyond the limits of the skin of the person to whom those phenomena are attributed. In the 1970s, Kripke (1972) and Putnam (1975) popularized the idea that mental content is external, that the meaning and reference of some words is not determined solely by what’s in the heads of people who use those words. In the 1980s, Nozick (1981) and Dretske (1981) introduced the notion that one’s justification for a given belief might not be determined solely by what’s in one’s head. In the 1990s, Clark and Chalmers (1998) suggested that the mind itself might extend beyond the limits of the skin. Pritchard argues in his (forthcoming) article that cognitive abilities – which we might think of as intellectual virtues or parts thereof – may extend. I have argued (Alfano forthcoming) that the phenomenon of stereotype threat is evidence that cognitive processes extend. The current proposal is that we should explore the extent to which this research program can be applied to ethics, that is, the extent to which it makes sense to say that some psychological dispositions that we might reasonably call moral virtues also extend beyond the limits of the skin of their possessor.
 It is worth noting in this connection that in Spencer’s Faerie Queene, in which each book is devoted to a different virtue as exemplified by its protagonist, only one book has a pair of protagonists. Redcross exemplifies holiness, Guyon temperance, Britomart chastity, Artegall justice, and Calidore courtesy. Only in book four do we become acquainted with Cambel and Telemond, who together embody friendship.
 What about intellectual virtues? I think that similar arguments work for them, at least for the “responsibilist” or motivational intellectual virtues, such as creativity, curiosity, open-mindedness, intellectual courage, and so on – though probably not for purely (or at least primarily) cognitive intellectual virtues, such as intelligence. I explore these issues in my (2013) book and in my (forthcoming) article on stereotype threat and intellectual virtue.
 As factitiously generous? I’m not sure.
 All? Perhaps not humility and modesty, which seem to involve a paradox of self-reference insofar as it’s hard, though maybe not impossible, to be humble and modest and to think of yourself as humble and modest. I explore this paradox of self-reference with some of my colleagues in Alfano et al. (forthcoming).