Is Aristotle’s account of friendship too egoistic? Would this be a problem for Aristotle?

Essay

No, Aristotle’s account of friendship (philia) in the Nicomachean Ethics is not egoistic, although it is, in a weak sense, self-serving. Both the substance and the language of Aristotle’s account makes clear that friendship requires both individuals to care about the other for the other’s own sake, in stark contrast to the egoist’s instrumentalisation of friendship. If his account really were egoistic, that would create problems for the coherence of his arguments about ethical virtue (arete ethike) and the fine (kalon) – but a conception of friendship stressing the connection between friendship and self-love (philautia) fits neatly with his views about human flourishing (eudaimonia) consisting in the exercise of virtue. In this essay, I first briefly clarify what I take the terms “egoistic” and “self-serving” to mean, before presenting Aristotle’s account of friendship as outlined in Books VIII and IX of the Nicomachean Ethics. I then explain why one might take this account to be objectionably egoistic, and argue that such a reading is not only at odds with Aristotle’s emphasis on the importance of other-concern and active care in friendships, but also based on an incorrect understanding of the relationship between egoism and lack of self-sacrifice. Finally, I conclude by suggesting that suspicions about Aristotle’s account of friendship being too egoistic may flow from scepticism about his conception of eudaimonia, rather than any shortcomings in the treatment of friendship per se.

When trying to characterise an egoist, we might initially define them as somebody who acts only (or excessively) in line with their personal interests. But this is too expansive – as Kraut (2025) observes, altruism need not involve self-sacrifice, and conversely a lack of self-sacrifice does not imply egoism. What matters more, intuitively, is an individual’s motivations and decision-making process: if someone has internalised concern and care for others’ interests, then we would not describe them as egoistic, even if all the actions they take do benefit themself; if instead the thought that others’ interests might be of non-derivative importance never occurs in their deliberative process, that person is egoistic. Self-servingness, on the other hand, is when an agent’s actions always promote their prudential good – and there is nothing particularly objectionable about this a priori. Since it is in principle possible for my self-interest to coincide with what is broadly ethically valuable, rejecting self-servingness without an empirical argument that these two goods do in fact conflict seems to amount to a strange fetishisation of self-sacrifice for its own sake.¹ So, it is important to distinguish between objections from egoism and objections from self-servingness when criticising Aristotle’s account of friendship.

Let us now turn to examining what that account comprises. Aristotle devotes two books of the Nicomachean Ethics to the topic, starting with a taxonomy of the different species of friendship and the conditions for such a relationship, before discussing various reasons for conflicts in friendship, and finally arguing for why friendship is a crucial component of eudaimonia. According to Aristotle, there are three types of friendship, predicated on utility, pleasure, or character (VIII.3, 1156a6-14). These correspond with expediency, pleasure, and fineness, the three objects of choice he described in Book II (Whiting 2006, p279), so it is unsurprising that, for Aristotle, friendships of character are the most choiceworthy and complete of the three. Aristotle describes friendships of utility and pleasure as diminished subtypes of friendships of character (VIII.4, 1157a30-32), sharing some of its features but lacking in its durability and mutual-improvement, because they are centred around incidental, changing characteristics of friends rather than the stable state of virtue that friendships of character are grounded in (VIII.4, 1157b1-5).

• Aristotle makes clear that he sees the friendships of utility and pleasure as being highly contingent and somewhat transactional (VIII.3, 1156a18-23). So you might first argue that his account of friendships of pleasure & utility is overly egoistic.

  • Aristotle is being descriptive, not prescriptive, though. He doesn’t think that these are the sorts of friendships that you should seek out, and in fact explicitly says that the virtuous person has no need for them (IX.9 1169b24-28). So the objection must be that his description is inaccurate and overly egoistic – that is, imperfectly virtuous people have friendships of pleasure & utility where they really do care about the other person.

  • These can be rebutted: Cooper (p304-5) notes how Aristotle’s initial description of eunoia includes mutual goodwill (eunoia) between the parties, and knowledge of this eunoia (VIII.2, 1155b31-56a5). Aristotle says this before distinguishing between the types of friendship, so it’s unlikely that he thought that utility & pleasure friendship lacked euonoia completely. Moreover, there is no reason to think that Aristotle’s use of dia is claiming that the friend of utility/pleasure cares about the person instrumentally to obtain utility/pleasure – the more natural translation is in a causal sense, whereby the production of utility/pleasure is what gives rise to the non-instrumental concern between friends (Cooper, p309-10). Aristotle does not look on this as highly as friendship of character or think that the other-concern goes as deep, because it’s about changeable features of the friend, but he doesn’t assert that friends of pleasure/utility lack other-concern entirely.

  • Whiting (p281) offers a convincing interpretation of what Aristotle means in when he suggests that friendship and utility don’t produce goodwill (VIII.5, 1155b31-34): they lack it in the sense of caring about each other in respect of their character. But I can certainly still care about someone else’s interests for their own sake even without that being based in their character. So this aspect is not overly egoistic.

• A more interesting and challenging objection is that Aristotle’s account of friendships of virtue is overly egoistic. It’s more interesting & challenging because (a) that is Aristotle’s ideal of friendship, and so gets at what he is advocating for as opposed to merely describing, and (b) gets at questions about whether virtue ethics as a whole is objectionably egoistic.

  • He talks at length in IX.9 about how having friendships is good for you, and puts forward a view on which the friend is an extension of yourself in IX.4. It might seem to follow that Aristotle ascribes only self-regarding motivations to the seeking out (or in) these friendships.

• But this view can be rebutted robustly too.

  • First, just from the text, Aristotle is clear that you need wish goods to the friend for their own sake, and want them to live for their own sake (IX.4, 1166a10-13). Moreover, he repeatedly stresses that giving is more important than receiving (e.g., VIII.8, 1159a34-35; IX.7, 1167b33-36), lending further weight to the view that he sees genuinely altruistic other-concern as crucial to friendships of character. There’s a clear link with the discussion of decency in V.10, which again is about giving without the expectation of receiving back.

  • Maybe all this shows is that the text is inconsistent & self-contradictory. Critics might say that Aristotle makes the right sounds about friendship and altruism, while really advocating for the virtuous agent to selfishly pursue nothing except their own eudaimonia.

  • However, there’s really no conflict here. Virtue is a state (hexis) that generates an appropriate situation-specific response and decision (prohairesis) grounded in the right principle (arche). So it’s not that the friendships of character are motivated by selfish considerations; a virtuous agent would not have those in mind. What moves the friendship, for the virtuous agent, is a deeply-held motivation to benefit the other person for their own sake.

• When people have the reaction that Aristotle’s account of friendship is too egoistic, they’re really objecting to his account of prudential wellbeing.

  • If Aristotle thought that the virtuous person was motivated by concern for their own happiness when seeking friendship, then that would sit uncomfortably with the idea that they act in a noble, fine, and praiseworthy way. But, as we have just discussed, there is no reason to ascribe this view to him. Cooper (note 6) helpfully highlights how one can act for someone else’s sake even if there are supplementary (or even stronger) reasons of self-interest which also motivate that action.

  • What Aristotle does think is that if one has sufficiently good reasoning ability and follows one’s true self-interest (Kraut 2022), one will always act rightly. This is at the heart of the case he makes in Books I and II of Nicomachean Ethics that eudaimonia consists in performing the human function well, which amounts to a life of excellence, including arete ethike.

  • People have the strong intuition that friendship requires self-sacrifice – for instance, you have to miss a trip to the opera in order to visit your friend in hospital. Aristotle does not deny that friendship might demand this of you, but he rejects the idea that there is any real sacrifice occurring. The virtuous person will happily make the hospital trip because they care about their friend in a non-instrumental way, but they also know that in doing the fine thing they are attaining eudaimonia through the exercise of their virtue.

  • This is why it is accurate to describe Aristotle’s account as self-serving: given his view on eudaimonia, friendship does not involve any sacrifice. Our suspicion about his seemingly-egoistic account of friendship can be attributed to this account of wellbeing – and that should give us some cause to doubt it.

References

Annas, Julia. 1977. ‘Plato and Aristotle on Friendship and Altruism’, Mind 86(344), 532-554

Cooper, John M. 1980. ‘Aristotle on Friendship’, in Rorty (ed.), Essays on Aristotle’s Ethics, University of California Press.

Kraut, Richard. 2022. 'Aristotle's Ethics', in Zalta & Nodelman (eds.), The Stanford Encyclopedia of Philosophy (Fall 2022 Edition). https://plato.stanford.edu/archives/fall2022/entries/aristotle-ethics/.

Kraut, Richard. 2025. 'Altruism', in Zalta & Nodelman (eds.), The Stanford Encyclopedia of Philosophy (Spring 2025 Edition, forthcoming). https://plato.stanford.edu/archives/spr2025/entries/altruism/.

Whiting, Jennifer. 2006. ‘The Nicomachean Account of Philia’, in Kraut (ed.), The Blackwell Guide to the Nicomachean Ethics.

Tutor’s feedback

This is a very nicely written essay with a clear structure and a systematically developed argument, well done. I think that you aptly recognise the importance of giving an account of how to understand ’egoism’ and of how this understanding can form the basis of a charge against the Aristotelian conception of friendship. Having laid out these terms clearly at the beginning of the essay serves you well throughout your argument. The contextualisation of the theme is done well, and it is in a way further developed when you relate the complaint against Aristotle to the idea of this actually expressing a broader dissatisfaction with the eudaimonistic picture as such.

In addition to a few points where clarifying the usage of certain terms would be desirable and a few places that lack clarity in expression, the main point of improvement that I would suggest is to encourage you again to demonstrate your ability to deal with and unpack passages of primary text. The conceptual framework and arguments that you offer are very good, but work in the history of philosophy requires you to engage with the text that you are concerned with – this means to do more than offering references to particular parts of NE but rather trying to discuss certain passages in more detail. (Admittedly, the questions are designed to prompt answers that you could also give in the exam, where obviously you are not expected to quote primary text.) Nonetheless, the argument is very well presented and should give us ample of material for discussion in our tutorial session.