In ‘Philosophy as a Humanistic Discipline’, Bernard Williams argued that philosophy has a distinctive relationship with history, and not just the history of philosophy. He had in mind, especially, moral and political thought. For Williams, changes in ethical outlook are typically driven by social and cultural forces distinct from the power of rational argument. When an ethical idea prevails over time, holders of the outlook it supplants often have ‘have [no] reason to recognize the transition as an improvement’ (Williams Reference Williams2000: 486). By their lights, the arguments for change are question-begging. In this respect, Williams believed, the history of ethics is unlike the history of modern science, whose telling typically vindicates later scientists on terms compelling to their predecessors.
I’m not sure Williams was right about the history of science, where vindication (in his sense) may be less prevalent than he claims.Footnote 1 But since he denied that the failure to vindicate an ethical outlook through its history debunks or undermines it – a point on which I agree – we shouldn’t be disturbed.Footnote 2 Instead, William portrays the use of history in moral philosophy as exploratory: it’s about ‘the reflective understanding of our ideas and motivations’, which has to be historical because ‘in many cases the content of our [ethical] concepts is a contingent historical phenomenon’ (Williams Reference Williams2000: 489).
In what follows, I argue that Williams was right to urge that moral philosophers scrutinize the history of the concepts that structure their thinking, but that he was, at least in this well-known essay, too sanguine. There’s a further threat of historical debunking, implicit in Williams’ remark that we are ‘unlikely to be able to make complete sense of our outlook [since it] will be in various ways incoherent’ (Williams Reference Williams2000: 491). At the limit, a putative ethical concept may turn out, in light of its defective history, to be empty or nonsense – a pseudo-concept. This is the case, I will suggest, for one of the pivotal (alleged) ideas of modern moral philosophy: self-interest as a distinctive source or subset of practical reasons.
The topic of self-interest and its place in ethics was a central one for Williams and his work lies in the background of my discussion, returning to the foreground at the end. Hovering throughout is Williams’ sympathy for the ‘young philosopher’, advised to study history, who worries that ‘there is too much that we need to know’ – to which he replied (Williams Reference Williams2000: 495):
I entirely see your, that is to say, our problem. … [But] while it is certainly true that we all need to know more than we can hope to know … it makes a difference what it is that you know you do not know. One may not see very far outside one’s own house, but it can be very important which direction one is looking in.
It’s difficult to prove that an apparent concept is senseless when people claim to understand it; and it’s unreasonable to demand conceptual analysis. Nor can I hope to tell more than a fragment of the history of ‘self-interest’. I will focus on its appearance in Western philosophy and even there, I will be selective. I’ll be glad if I can convince you that the sorts of doubts about morality that are taken seriously by philosophers in light of critiques by Anscombe (Reference Anscombe1958), MacIntyre (Reference MacIntyre1981), and Williams (Reference Williams1985) should be taken seriously for self-interest. But I know how much I do not know, and I’d be almost as glad simply to have raised a worthwhile question and to have pointed in a direction for others to look.
1. ‘Modern Moral Philosophy’
The challenge ‘Why be moral?’ is often traced to Callicles in Plato’s Gorgias or to the myth of Gyges’ ring, which gives its wearer the power to become invisible, in Book II of Plato’s Republic:
Now, no one, it seems, would be so incorruptible that he would stay on the path of justice or stay away from other people’s property, when he could take whatever he wanted from the marketplace with impunity, go into people’s houses and have sex with anyone he wished, kill or release from prison anyone he wished, and do all the other things that would make him like a god among humans. Rather his actions would be in no way different from those of an unjust person, and both would follow the same path. This, some would say, is a great proof that one is never just willingly but only when compelled to be. (Republic 360bc)
In Sacrifice Regained, Roger Crisp contends that there are in fact two questions here. The first is ‘whether there are self-standing or ultimate normative reasons for action grounded in morality itself or in the well-being of others’, the alternative being that all reasons are ‘ultimately non-moral and entirely self-interested’ (Crisp Reference Crisp2019: 3). The second question is whether it’s always in our interest to be moral: ‘Socrates’s own response to Glaucon can plausibly be read as an egoistic defence of morality, according to which the agent’s own self-interest or well-being consists in virtue. This position – moralistic egoism – is characteristic of ancient philosophy in general’ (Crisp Reference Crisp2019: 4).
The idea of a conflict between morality and self-interest, to be settled by declaring a victor – or a dualistic truce – is no doubt characteristic of modern moral philosophy. It’s much less obviously present in ancient Greek thought. Where others have raised historical doubts about the place of morality in ethics, I’ll raise doubts about the place of self-interest. A consequence will be that Crisp is wrong to read ancient philosophers as ‘moralistic egoists’ and to distinguish two questions where there is only one.
My primary inspiration, beside Williams, is Elizabeth Anscombe’s (Reference Anscombe1958) barn-burner, ‘Modern Moral Philosophy’. Anscombe makes two claims of interest here, one negative, one positive.
Anscombe’s Negative Thesis: The moral vocabulary does not correspond to any concept in Aristotle’s ethics.
Aristotle’s primary ethical concepts are eudaimonia, or the most desirable life; virtue or excellence, which makes us good as human beings; concepts that stand for particular virtues; and the concept of eupraxia or acting well. None of these maps neatly on to ‘moral virtue’, ‘moral obligation’, or ‘moral wrongness’. There are things one should do that one is not morally obligated to do – for instance, taking care of one’s health – and things one should not do that would not be morally wrong – for instance, smoking cigarettes. Eupraxia is a matter of acting as one should, not of specifically moral obligation. Likewise, eudaimonia is the life one should choose, or want to live, all told. The concept of virtue is not distinctively moral; nor, despite a common translation, is the concept of ethical virtue. Take courage, for instance: a virtue of character that is neither specifically moral nor non-moral, but relevant to any difficult or frightening endeavour. Aristotle writes about the virtue of justice, but its domain is much narrower than that of morality in the modern sense: it’s concerned with proportionality in punishment, reciprocity, and fair distribution; and there are other virtues of character, like generosity, we would count as ‘moral’.
Anscombe sums up her scepticism with characteristic verve. ‘If someone professes to be expounding Aristotle and talks in a modern fashion about “moral” such-and-such,’ she writes, ‘he must very imperceptive if he does not constantly feel like someone whose jaws have somehow got out of alignment: the teeth don’t come together in a proper bite’ (Reference Anscombe1958: 2). As Williams notes, she was anticipated by R. G. Collingwood, in his autobiography (Reference Collingwood1939: 63–64), complaining that Oxford contemporaries would translate ‘some ancient Greek expression as “moral obligation” and then point out that Aristotle … had an inadequate theory of moral obligation. It was like a nightmare, Collingwood said, in which one met a man who insisted on translating the Greek word for a trireme as “steamship” and then complained that the Greeks had a defective conception of a steamship’ (Williams Reference Williams2000: 478).
According to Anscombe, Aristotelian ethics survives as the framework for Christian ethics in the medieval period, though changes in the conception of human nature produce a changing roster of virtues. God’s commands come to be seen as a source of authoritative reasons, but the account of why they are reasons remains Aristotelian. Respecting them is a condition of virtue, or what it is to be good as a human being, conceived now as a being created by God. (Aquinas is the model here.) ‘Morality’ arises from an attempt to secularize the ethics of divine command. It’s as if we were to fix the meaning of ‘moral obligation’ by saying ‘it stands for reasons that are just like divine commands except that they are not commands and have no relation to God’. This stipulation gives no meaning to the moral vocabulary just as, for Anscombe, it failed to acquire a sense in its historical emergence. On a standard reading, ‘Anscombe thought it obvious … that no practical law of any kind can exist without a lawgiver’ and so ‘concluded that there can be such a thing as morality only if it is legislated by God’ (Darwall Reference Darwall2023: 2). But her view is in fact more radical: that the distinctive moral vocabulary is literally nonsense; there can be no such thing as morality, at all.Footnote 3
Anscombe’s Positive Thesis: The putative content of the moral vocabulary originates in a confused attempt to preserve the lawlike character of divine command in concepts whose sense is wholly secular; the upshot is that this vocabulary has no meaning.
Anscombe’s argument is a model for mine, though her Negative Thesis is plausibly false. Elsewhere, I’ve argued that Aristotle does have a concept corresponding to morality. As well as ‘particular justice’, one virtue among many, he recognizes ‘justice in general’, which he associates with ‘lawfulness’: ‘This form of justice, then, is complete virtue, although not without qualification, but in relation to another’ (NE 1129b26–28). The last clause of this sentence is obscure, but it’s notable that replies to Anscombe’s Positive Thesis, which try to articulate the concept of morality without confusion, can be read as distinct interpretations of ‘virtue … in relation to another’. As these accounts have it, morality is the dimension of living well for which one is socially accountable, or which consists in meeting one’s obligations to other people, or to which others have a right to make one conform.Footnote 4 In effect, morality is justice in general and Aristotle does believe in it, even if its nature is not perfectly clear.Footnote 5
Surprisingly, however, a strong case can be made that Anscombe would be right about the other pole of the modern dualism: self-interest.Footnote 6 There is a substantial literature on Anscombe’s doubts about morality. There is little to nothing on parallel doubts about self-interest. And yet, I will suggest, they are more compelling. Following Anscombe’s model, I will defend, schematically and too briefly, the conjunction of two claims:
The Negative Thesis: ‘Self-interest’ does not correspond to any concept in Aristotle’s ethics.
The Positive Thesis: The putative concept of self-interest in moral philosophy originates in a confused critique of ancient ethics in the early modern period, according to which subjective theories of well-being must be good theories of something; the upshot is that the vocabulary of ‘self-interest’ has no ethical meaning.
I’ll end by exploring the implications, and limitations, of my argument.
2. The Negative Thesis
If the object of self-interest is well-being or a life that is good for the person living it, how can one deny that this concept figures in ancient Greek ethics? According to Stephen Darwall (Reference Darwall2023: 5):
Plato, Aristotle, and other ancient thinkers tend to be eudaimonists, holding that all normative reasons for action must derive from the agent’s own good or happiness (eudaimonia) broadly conceived, even when it comes to virtue and justice. In Socrates’s exchange with Glaucon and Adeimantus, it is simply assumed by all parties that if Socrates cannot establish that it is intrinsically or extrinsically beneficial to the just person to be just, he will not have shown any reason for them to act justly rather than, say, simply to appear to be doing so.
But this interpretation is tendentious. For one thing, it’s misleading to translate ‘eudaimonia’, which stands for the most desirable life, as ‘happiness’ in the contemporary sense of the word: a subjective state of mind. Eudaimonia is a matter of how one actually lives one’s life, not just feeling happy about it. For another, it’s unclear where to locate practical reasons in the Aristotelian view. Is eudaimonia the source of all such reasons, as Darwall suggests, or is it rather that, being the most desirable life, eudaimonia is a function of reasons for desire?Footnote 7 On the first view, the reason to want or choose a certain kind of life is always, in the end, that doing so contributes to or approximates eudaimonia. On the second view, eudaimonia is the life there is most reason to want or choose, whatever the reasons are.
The texts do not readily decide between these readings. Still, either way, eudaimonia corresponds to well-being in one sense: a life that is desirable, all told. Unlike Socrates, Aristotle does not believe that virtue is sufficient for eudaimonia, which also includes the goods of fortune, i.e. living in the circumstance one should want to occupy, not just living as one should in the circumstance one happens to be in.Footnote 8 Since one should want oneself to act as one should act, which for Aristotle means acting virtuously, practical virtue contributes to eudaimonia; but it does not exhaust it.
At the same time, eudaimonia differs sharply from well-being as the object of self-interest in modern moral philosophy. ‘Self-interest’ is supposed to track a particular way in which life can be good, not its being desirable, all told. If there is reason to live or act in a certain way because doing so benefits other people or respects their rights, there is reason to want to live that way; these reasons make that way of life desirable. But it’s meant to be a further question whether acting in accord with them advances one’s self-interest. That was the point of Crisp’s distinction between two objects of inquiry: Are there moral reasons that do not derive from self-interest? And is it in one’s interest to be moral? The stipulation of basic reasons to benefit others, or respect their rights, answers ‘yes’ to the first question but leaves the second question open. By contrast, when there’s reason to live or act in a certain way, it’s not an open question whether doing so contributes to a desirable life, or eudaimonia – though it may be an open question whether it does so more than living or acting in other ways, or whether the circumstance in which it is desirable to act in this way is a circumstance one should want oneself to occupy.Footnote 9 ‘Self-interest’ is meant to demarcate a (perhaps improper, but conceptually distinct) subset of practical reasons; eudaimonia is not. Or, for those who prefer evaluative terms: ‘self-interest’ is meant to denote a way in which life can be good for the person living it that is more specific, and potentially narrower, than its being desirable for them to live that way, all told.
In ‘The Concept of Well-Being’, a representative introduction in a standard survey, The Routledge Handbook of Philosophy of Well-Being, Stephen Campbell distinguishes its topic, ‘a prudentially good life’ from a life that is good simpliciter, a morally good life, and a choiceworthy life, one that is worth choosing or aiming to have (Campbell Reference Campbell and Fletcher2016: 403). Similar distinctions are drawn by L. W. Sumner (Reference Sumner1996: 20–25) and T. M. Scanlon (Reference Scanlon1998: 111–113), among others. These distinctions play a crucial role in defending ‘subjective’ theories of well-being – for instance, the equation of well-being with final desire-satisfaction. Objections to such theories sometimes point to those who indulge base appetites, enjoy immoral lives, or devote themselves to trivial activities like grass-counting, but whose desires are amply satisfied. Surely these are not good lives! The standard reply is to insist that the subjective theory is a theory of well-being only – not virtue, dignity, achievement, or excellence.Footnote 10 These lives may not be virtuous or dignified, but they meet the aspirations of self-interest. Again, this term is meant to track one way in which life can be good, or one kind of practical reason. It is not meant to capture everything that makes a life desirable.
It is this concept Aristotle lacks. He does not demarcate reasons or aspects of a good life in this way, having no concept of a life that is good for a person other than that of a life that is desirable for them, all told. There are two passages that might be cited to the contrary.Footnote 11 One is in Book II of the Nicomachean Ethics, which distinguishes three objects of choice: the noble, the advantageous (sumpheron), and the pleasant (NE 1104b29-35). But we can read the advantageous, here, as the good or desirable life, not an aspect of that life. On this reading, the three objects of choice correspond to three forms of desire: spirited, rational, and appetitive–where rational desire aims at eudaimonia.Footnote 12 By contrast, the noble and the pleasant are distinct ways of being good. It is a substantive and not conceptual truth, for Aristotle, that virtue – which he associates with the noble or kalon – contributes to living well, and that the eudaimon life is pleasant. But to acknowledge this is not to distinguish, within eudaimonia, an element that is advantageous to the subject and a potential element that is not.
The second key passage is in Book VIII, on friendship, which separates what is loved into the good, the pleasant, and the useful (sumpheron) – with their corresponding forms of friendship – and asks, ‘Do men love, then, the good, or what is good for them?’ (NE 1155b22-23) As Aristotle clarifies, however, the issue here is about appearance and reality, what is in fact good and what merely appears to be: the solution to the puzzle is that ‘each man loves not what is good for him but what seems [good for him]’ (NE 1155b25–26).Footnote 13 There’s no suggestion that what is good for me, in fact, is an aspect of eudaimonia that may fall short of it.
Aristotle goes on to distinguish friendships of utility or advantage from friendships of virtue and of pleasure, but the distinction he’s drawing is between instrumental and non-instrumental value, not between self-interest and what is good, all told. He uses ‘sumpheron’ in the same way when he writes, in Book VI, that it is ‘a mark of a man of practical wisdom to be able to deliberate well about what is good and expedient [sumpheron] for himself, not in some particular respect … but about what sorts of things conduce to the good life in general’ (NE 1140a25–28; see also NE 1142b31–33). Despite a surface resemblance, the distinction in Books VI and VIII does not line up neatly with Book II, where ‘advantage’ does not appear to mark instrumental value. This is confirmed by the end of the Book II passage, in which ‘even the noble and the advantageous appear pleasant’, as merely instrumental goods would not (NE 1104b35–1105a1.) In none of these cases is the sumpheron a limited aspect of the good life.
The conclusion stands, then, that Aristotle’s only relevant ethical concept is eudaimonia, the most desirable life. To paraphrase Anscombe: if someone professes to be expounding Aristotle and talks in a modern fashion about ‘self-interest’ or ‘rational egoism’, his intellectual jaws should feel badly misaligned. Aristotle is innocent of the idea that practical reasons can be subdivided, some being self-interested and others not – he does not have the concept of a life that is good for you as distinct from a life that is desirable, all told – and it’s distorting to report this by suggesting that he thereby treats all reasons as egoistic.
There is potential for confusion here, on several fronts. Take the virtue of justice. If eudaimonia is the best or most desirable life, all told, it’s not an open question, when there’s reason to live or act a certain way, whether doing so contributes to eudaimonia, as it is an open question whether it serves one’s self-interest.Footnote 14 To insist that a defence of justice show how it contributes to eudaimonia is therefore not to limit or constrain that defence: it’s merely to insist that it give reasons in light of which a just life is desirable. That doesn’t make this part of Aristotle’s ethics vacuous. For one thing, it makes a difference not to be constrained by rational egoism in our defence of justice; and it remains a substantive claim, denied by the likes of Callicles, that we have reason to be just. For another, Aristotle does seem to accept a form of egoism, in that he thinks we ought to aim explicitly towards eudaimonia: living the most desirable life should be our goal. This is potentially objectionable, portraying the virtuous person as unduly concerned with his or her own virtue.Footnote 15 But we should not conflate claims about the psychology of practical wisdom with claims about the nature of eudaimonia, or the use of ‘self-interest’ for a self-reflexive motive with the use that is our target, on which it purports to mark a subset of practical reasons or circumscribed aspect of the good life. Finally, Aristotle, like most ancient Greek philosophers, accepts a form of naturalism on which eudaimonia, the most desirable life, is at the same time the perfection or realization of human nature. This constrains the defence of justice, not by restricting it to a subset of self-interested reasons, but by requiring us to relate eudaimonia to human flourishing conceived in naturalistic terms. We’ll come back to this connection at the end.
Further confusion stems from interpretive claims that are in the vicinity of, but importantly different from, the Negative Thesis. Thus, Julia Annas notes that, in ancient ethical theories, ‘we do not find the assumption that prudential and moral reasoning are distinct forms of rationality’ but suggests ‘that the missing element seems to be prudential reasoning’ (Annas Reference Annas1995: 342). This sounds like the Negative Thesis. But what she argues instead is that, for some ancient Greek ethicists, prudence or self-interest transcends itself, leading us to recognize the supremacy of virtue: ‘the dominance of the moral point of view, once reached, requires the agent to transform her priorities in accordance with it, and thus to revise considerably the results previously attained by the uncomplicatedly prudential use of reasoning’ (Annas Reference Annas1995: 252). On the Negative Thesis, there is no such use.
A final interpretive claim is concerned with conflict. As we will soon see, modern moral philosophers draw a conceptual distinction between self-interested and moral or other-regarding reasons. At the same time, they are preoccupied with the potential for divergence: between morality and self-interest, and between my interests and yours. It might be argued that such conflict is not a problem for Plato and Aristotle, since they regard the good of one as inextricable from the good of others: the human good is essentially social.
This is not the issue raised by the Negative Thesis. For one thing, arguing for the congruence of my interests with yours is compatible with drawing a conceptual distinction, among our reasons, between those that are self-interested and those that are not. For another, rejecting that conceptual distinction is compatible with allowing for conflict between my good and yours, an observation made by Nicholas White (Reference White1995: 272–273).Footnote 16 He finds in Aristotle’s treatment of friendship a case in which only one of two friends can perform a virtuous action; Aristotle suggests that ‘it may be nobler to become the cause of [one’s] friend acting than to act [oneself]’ and that ‘the good man is seen to assign to himself the greater share of what is noble’ (NE 1169a33–1169b1). If the original act is either more or less virtuous than the act of deferring to one’s friend, there is a conflict between the good of one and the good of the other. What’s more, even if their virtue is equal or on a par, a key distinction remains: the Negative Thesis neither follows from nor entails the impossibility of conflict between the good of individuals.
The point is not that, for Aristotle, our interests inevitably align. It’s that there is no concept of well-being as self-interest, no concept of a life that is a good for the person living it other than that of a life that is desirable, all told.
3. The Positive Thesis: History
According to the Positive Thesis, the putative concept of self-interest in moral philosophy originates in a confused critique of ancient ethics in the early modern period, according to which subjective theories of well-being must be good theories of something.
Darwall (Reference Darwall2023: 30) traces the critique of Aristotelian ethics to Grotius on counsel and command: ‘Once we have the distinction between a mere “counsel,” however well supported by reasons … on the one hand, and an obligating demand, on the other, we are committed to the idea that there must a source of reasons for acting other than the agent’s own good.’ Since the pursuit of one’s own good is the domain of mere advice, Darwall argues, there must be another domain of reasons, corresponding to demands. But Darwall is wrong to think that recognizing obligations or demands requires a fundamental break with Aristotle. There is nothing in Aristotle’s ethics to preclude the existence of reasons that reflect our obligations or the demands that can be made on us.Footnote 17 And if there are such reasons, a life could be desirable, all told, in light of them, partly because it is a life of doing right by others.
More plausibly, the early modern critique of ancient ethics turns on subjective theories of well-being as desire-satisfaction or pleasure over pain, or some conflation of the two. Thomas Hobbes is the canonical figure:
For there is no such Finis ultimus (utmost aim) or Summum Bonum (greatest good) as is spoken of in the books of the old moral philosophers. … Felicity is a continual progress of the desire, from one object to another, the attaining of the former being still but the way to the latter. (Hobbes 1994 [Reference Hobbes and Curley1651]: Part I, Ch. XI)
Where Hobbes emphasizes the insatiability of desire, John Locke stresses the relativity of pleasure and pain:
the various and contrary choices, that men make in this world, do not argue, that they do not all pursue good; but that the same thing is not good to every man alike. … Hence it was, I think, that the philosophers of old did in vain inquire, whether summum bonum consisted in riches, or bodily delights, or virtue, or contemplation: and they might have as reasonably disputed, whether the best relish were to be found in apples, plums, or nuts, and have divided themselves into sects upon it. For … the greatest happiness consists in the having those things which produce the greatest pleasure, and in the absence of those which cause any disturbance, any pain. Now these, to different men, are very different things. … This, I think, may serve to show us the reason, why, though all men’s desires tend to happiness, yet they are not moved by the same object. Men may choose different things, and yet all choose right… (Locke 1975 [Reference Locke and Nidditch1689]: Bk. II, Ch. XXI)
Notably, those who disagree with Hobbes and Locke about human motivation tend to concede that they are right about one aspect of the good. Here are the arch anti-egoists, Francis Hutcheson and Joseph Butler:
Because we shall afterwards frequently use the Words Interest, Advantage, natural Good, it is necessary here to fix their Ideas. The Pleasure in our sensible Perceptions of any kind, gives us our first Idea of natural Good, or Happiness; and then all Objects which are apt to excite this Pleasure are call’d immediately Good. (Hutcheson 2004 [Reference Hutcheson and Leidhold1725]: Treatise II, ‘Introduction’)
Happiness or satisfaction consists only in the enjoyment of those objects, which are by nature suited to our several particular appetites, passions and affections. (Butler 2017 [Reference Butler and McNaughton1729]: Sermon XI, paras. 6, 9)
There’s nothing incoherent in the claim that eudaimonia, the most desirable life, is one of getting what you want or maximizing pleasure over pain. Callicles takes a view like this in Plato’s Gorgias:
[The] man who’ll live correctly ought to allow his own appetites to get as large as possible and not restrain them. And when they are as large as possible, he ought to be competent to devote himself to them by virtue of his bravery and intelligence, and to fill them with whatever he may have an appetite for at the time. (Gorgias, 491e–492a)
This conception of eudaimonia threatens the authority of justice and morality. Their defenders must show that the life of justice or morality best satisfies our desires or is more pleasant and less painful, on balance, than an unjust or amoral life. Alternatively, they must argue, with Plato and Aristotle, that there’s reason to be just or moral that does not derive from our desires, pleasures, or pains, so that well-being does not reduce to pleasant desire-satisfaction.Footnote 18
But Hutcheson and Butler did not follow either path. Believing that there’s reason to be just or moral that does not derive from our own desires, pleasures, or pains, they disagree with Hobbes and Locke about the most desirable life. But they concede the subjectivity of well-being or self-interest, and are thus forced to posit reasons unrelated to it:
If by Obligation we understand a Determination, without regard to our own Interest, to approve Actions, and to perform them; which Determination shall also make us displeas’d with our selves, and uneasy upon having acted contrary to it; in this meaning of the word Obligation, there is naturally an Obligation upon all Men to Benevolence. (Hutcheson 2004 [Reference Hutcheson and Leidhold1725]: Treatise II, Sect. VII.I)
[The] greatest degree of scepticism … will still leave men under the strictest moral obligations, whatever their opinion be, concerning the happiness of virtue. (Butler 2017 [Reference Butler and McNaughton1729]: ‘Preface’, para. 27)
As Darwall writes, ‘Butler’s claim [is] that being contrary to the agent’s good and being ruled out by conscience, that is, being morally wrong in the agent’s own view, are both sufficient reasons not to do something’ (Reference Darwall2023: 168) – though they are not coeval, since conscience has authority over self-love.
The idea of a distinction between self-interested and moral or altruistic reasons took on a life of its own, shared by philosophers from Kant and Mill to the present day, and accepted even by those who deny the subjectivity of self-interest. According to a now-standard taxonomy, conceptions of self-interest divide into hedonic, desire-satisfaction, and ‘objective list’ theories (Parfit Reference Parfit1984: Appendix I). The nature of self-interest, and its significance as a source of reasons, are treated as central, substantive questions in philosophical ethics.
Sidgwick was right about the fundamental shift:
[In] Greek moral philosophy generally, but one regulative and governing faculty is recognised under the name of Reason – however the regulation of Reason may be understood; in the modern ethical view, when it has worked itself clear, there are found to be two, – Universal Reason and Egoistic Reason, or Conscience and Self-love. (Sidgwick Reference Sidgwick1902: 198)
But the modern view is not remotely clear. The confusion, epitomized by Hutcheson and Butler, is to recognize that subjective theories of well-being omit important reasons, to acknowledge these reasons, and yet to insist that the subjective theory must be right about something: the putative object of self-interest. We begin with the concept of eudaimonia, a life that is desirable, all told. We agree that Hobbes and Locke are wrong about that. And then it’s as if we introduce ‘self-interest’ by saying ‘it stands for whatever it is they were giving credible theories of’. It’s fine to stipulate that you will use the word ‘self-interest’ to mean desire-satisfaction, or maximizing pleasure over pain, as shorthand. But you can’t bootstrap a subject-matter for desire-satisfaction theory and hedonism to be contestable theories of by appeal to those very theories. If all you tell me about ‘self-interest’ is that it doesn’t mean a desirable life, all told, and that it’s not unreasonable to equate it with getting what you want, or with maximizing pleasure over pain, though it might not consist in either, you haven’t given meaning to the term.
Here’s another way to make the point. If reasons derive from desires, pleasures, and pains, it makes sense to distinguish not just my reasons from yours, but, among my reasons, those that derive from my desires, pleasures, and pains, and those, if any, that derive from yours. (Hutcheson would accept this picture; Butler would not – but he could draw a similar distinction, adding reasons of conscience to the list of those that have a source outside of us.) What does not make sense is to introduce the term ‘self-interest’ by saying ‘it tracks a distinction in the source of reasons just like one we can make within desire-satisfaction theory and hedonism except that it does not depend on the truth of any theory like them’. (Compare Anscombe on ‘moral obligation’ as purporting to stand for reasons that are like divine commands except that they are not commands and have no relation to God.) It’s only within a certain family of views that we can say, of each reason to which you are subject, that its source lies in a certain aspect of another’s life – as, for instance, their desires, pleasures, or pains – or in a similar aspect of your life, or neither. We can’t abstract a theory-laden distinction in the source of reasons from the theories that give it sense, as the basis for a subject-matter for those theories and their alleged competitors to be about. There is nothing but eudaimonia – a desirable life, all told – for subjective theories to be theories of.
It might be argued, in response, that anyone can distinguish the aspects of eudaimonia that contribute to the good of others from the aspects that don’t. We could introduce a term for the latter, ‘S’. But that would not define the object of self-interest in the modern sense. It’s built into the definition of ‘S’ that what benefits others does not contribute to it, where that is meant to be an open question for self-interest. It’s supposed to be an open question, too, whether there are things I am morally required to do that are not in anyone’s interest, including mine; but it follows from the definition above that, if I am morally required to do something that does not benefit others, it contributes to S.
It’s a good question why critics of Hobbes and Locke were willing to concede the subjectivity of something like well-being. Perhaps they were moved by a nascent liberalism – made explicit in Mill’s (1978 [Reference Mill and Rapaport1859]) On Liberty – according to which no-one can tell us how to live our lives within the bounds of the morally permissible: it’s up to us. To reason from this premise to a subjective theory of well-being would be to conflate what others have a right to make us do with what we have reason to do. The former may be limited to what is morally obligatory; but why the latter? What’s more, if there is reason to act in one’s self-interest, the upshot is incoherent, since both hedonists and desire-satisfaction theorists ‘tell us how to live’ in the second sense: they tell us what there’s reason for us to do within the bounds of the morally permissible, regardless of what we think. Still, that this line of thought equivocates and contradicts itself does not mean that it did not move minds: I’m in no position to reject interpretations of historical philosophers on which they were seriously confused.
What Anscombe said of ‘morality’ is arguably true of ‘self-interest’, then: the putative concept originates in philosophical confusion. It purports to mark a subset of practical reasons, or a specific way in which life can be good. But it does so by illicit postulation – on which subjective theories of well-being are good theories of something even if they are wrong about eudaimonia – or by confusing a distinction in the source of reasons internal to a range of ethical theories with a subject-matter for those theories to be theories of. The upshot is that we can attach no meaning to ‘self-interest’ in modern moral philosophy. To borrow a phrase from Wolfgang Pauli, competing theories of self-interest are not even false.
4. The Positive Thesis: Theory
My argument has been schematic, and there’s no way to approach comprehensiveness here. Consider it a provocation to historians. But I can address at least a few potential concerns.
Suppose I am right that Aristotle had no concept of self-interest, once it’s distinguished from eudaimonia, and that the distinctive notion of self-interest in moral philosophy – on which it is an open question how reasons to live or act in one way or another bear on one’s self-interest – originates in the confusions I have sketched. This would do nothing to undermine contemporary discussions of self-interest if we could explicitly define or clarify their scope. It would be unreasonable to demand conceptual analysis, but perhaps we can elucidate the subject-matter well enough to make it the object of philosophical theory.
I am sceptical. It might be said, for instance, that self-interested reasons are non-moral. But apart from the challenge of explaining what makes a reason ‘moral’, the concept of self-interest is meant to be narrower than this, leaving conceptual room for reasons that are neither moral nor self-interested, as with reasons of love or ones concerned with ‘meaning in life’.Footnote 19
A second approach would emphasize that ‘self-interest’ tracks the aspects of a desirable life that are not merely good but good for the subject of that life. But the phrase ‘good for’ cannot bear the weight it’s given here. On the most credible theory, the basic meaning of ‘good for’ is end-relational, as when we say, ‘This hammer is good for smashing windows and nailing up hooks.’ When we fail to specify an outcome or activity, it is supplied by context. This is what happens when we say that X is good for a living thing, N. The default interpretation is one on which the relevant outcome or activity is N’s well-being or living well. But if this is how we interpret ‘good for N’, the phrase does nothing to clarify the notion of well-being, or to isolate an interpretation on which it is distinct from eudaimonia, a life that is desirable, all told.Footnote 20
It might be said, at this stage, that things must have gone awry, since we evidently mean something when we call a person ‘selfish’ and this meaning must involve self-interest taken to excess. Although I’m wary of appeals to what is evident, given the schematic history above, I do think selfishness is a vice. But its relation to self-interest is obscure. Roughly put, selfishness is relative indifference to others, including family and friends. It’s compatible with selfishness in this sense that one respect one’s moral obligations – one need not be purely ‘self-interested’. It’s compatible with selfishness that one ruin one’s own life: the selfish person may not be moved effectively by ‘self-interest’. And it’s meant to be an open question for theories of ‘self-interest’ whether good relationships contribute to it or not: being selfish may detract from one’s ‘self-interest’. In the last three sentences, the scare-quotes around ‘self-interest’ indicate that no clear meaning is attached to it: I am describing how those who think they understand the phrase would use it. The point is that we cannot leverage the meaning of ‘selfishness’ to give a meaning to ‘self-interest’ that would vindicate this usage.
Finally, it might be argued that, even if it’s true, the history above is inconclusive, since subsequent centuries may have given content to ‘self-interest’, even if that content is hard to clarify or define. How plausible is it, in the end, that everyone who writes or talks about self-interest today is spouting utter nonsense?
A concession is crucial here, implicit in the history so far. Within a given conversational or theoretical context, we can attach a local meaning to ‘my interests’. If we are concerned with desire-satisfaction, or pleasure and pain, it makes sense to draw a contrast between mine and yours and use the language of self-interest accordingly. The same is true when we negotiate a contract, pitting my financial interests against yours or finding avenues for mutual benefit. More generally, economic contexts provide a local sense for claims about distinct, potentially conflicting interests. It is not an accident that ‘self-interest’ emerges as a putative concept in moral philosophy alongside the ideology of the free market, on which the limitless pursuit of our desires, constrained by legal obligation, is no longer demonized as avarice but conceived as the source of social wealth. The notion of private interest makes sense, here, as part of a social-economic theory.Footnote 21 It’s not an implication of my argument that this theory is nonsense. My target is ‘self-interest’ as an ethical idea, like morality, that marks a subset of practical reasons or aspect of the good life.
When social scientists cite self-interest in explanations of behaviour, what they’re doing may be innocent, then, so long as context supplies a workable definition. This isn’t always clear. According to Francis Edgeworth in the delightfully-named Mathematical Psychics, ‘the first principle of Economics is that every agent is actuated only by self-interest’ (Edgeworth Reference Edgeworth1881: 16). Amartya Sen complained, a century later, that the definition of ‘utility’ in terms of preference turns this principle into nonsense: ‘no matter whether you are a single-minded egoist or a raving altruist or a class conscious militant, you will appear to be maximizing your own utility’ (Sen Reference Sen1977: 323).
As this indicates, it’s not just moral philosophers who are confused about ‘self-interest’, and even when they are, much of their work may be salvageable. The recovery will involve explaining away any substantive reliance on ‘self-interest’ as demarcating a subset of reasons or aspect of the good life that has significance outside the remit of a local theoretical role. When moral theorists formulate theories of beneficence, for instance, they should say whether ‘welfare’ means a desirable life, all told, or something narrower, in which case they need to specify what that is without appeal to the language of ‘self-interest’.
5. What is Human Flourishing?
How does the question ‘Why be moral?’ look when it’s liberated from the dualism of morality and self-interest? It is not immediately answered. There is such a thing as well-being, if that means a life that is more or less desirable, all told; and there are corresponding concepts of benefit and harm. It’s a substantive question how justice and morality relate to well-being in this sense: whether justice makes our lives better and injustice makes them worse. As we’ve seen, Callicles and Glaucon raise doubts about the desirability of a just life. But we don’t add anything to the problem they pose, or its solution, if we say that the reasons for a life of justice are, or are not, ‘self-interested’. By dropping that language, we avoid the false impression that there is a further challenge, apart from the reality and force of reasons to be just, about the bearing of those reasons on self-interest. Crisp is wrong to find two questions here, where there is only one. In asking ‘Why be just?’ we are asking about well-being or the good life, broadly understood, but we should not be haunted by the fear that self-interest is the only possible source of reasons to act. This anxiety has no content.
None of this makes the question easy. And disputes about it may look a lot like ones about morality and self-interest. When we turn to pre-Platonic debates about justice, that’s exactly what we find.Footnote 22 The backdrop to these debates is a traditional view on which justice was given to humanity by Zeus, on which Zeus punishes the unjust and rewards the just with wealth and good repute, and on which the just life is therefore desirable, all told.Footnote 23 Confronting doubts about the Olympian Gods, the fifth-century sophist Antiphon recognized the secular limits of this reasoning. If justice is ‘not to transgress the laws of the city of which one is a citizen’, he wrote, ‘a person would make use of justice most beneficially for themselves if they considered the laws as great when in the presence of witnesses and the things of nature [as great] when bereft of witnesses’ (quoted in Anderson Reference Anderson2024: 40, 45). In other words: we do best for ourselves if we have a reputation for justice but pursue wealth and other material goods unjustly when we can. The atomist Democritus took an opposing view, arguing that the good life is one of euthymia or contentment, where this is best achieved by moderating one’s desires, avoiding the envy that motivates injustice and the shame that results from it.Footnote 24 The life of justice is thus preferable by hedonic lights: ‘one doing injustice is more miserable than one suffering it’, Democritus concludes (quoted in Anderson Reference Anderson2024: 92; see also Nill Reference Nill1985: 84).
The dispute is reminiscent of modern moral philosophy. But what is striking about it is less the resemblance than the fact that the ancient Greek antagonists never invoke a conceptual distinction between self-interest and eudaimonia to argue that, while justice may be against one’s interest, it is still better to be just than unjust.
Instead, they take seriously the view that eudaimonia is a matter of material wealth, or honour, or maximizing pleasure over pain, and ask whether a life of justice is desirable in those terms–which is to say desirable, all told. There is conceptual room for that question, just as there is room to ask whether justice is a virtue, and whether it contributes to eudaimonia more directly, as a constitutive not causal means, but there is no further question whether, in doing so, it contributes to self-interest. Far from suggesting that self-interest was conceptualized in ancient Greek thought, as an aspect of the good life, the prehistory lends support to the view that it was not.
It does suggest a puzzle, however. Why would one be tempted by materialism or hedonism as a theory of eudaimonia, the best or most desirable life, all told, unless one were already sceptical of justice as an aspect of living well? Antiphon provides an answer, namely that ‘many of the things just according to the law are hostile to nature’ (quoted in Anderson Reference Anderson2024: 42; see also Nill Reference Nill1985: 54). He is drawing on a tacit premise about eudaimonia, the most desirable life, on which it is anchored in physis. Whatever it consists in, eudaimonia is human flourishing and it must be continuous, somehow, with the flourishing of plants and non-human animals, for whom things can be good or bad, a benefit or a harm. If the flourishing of other living things is a function of their nature, or physis, ours must be, too. And the appeal of hedonism – and similar views – turns on a hard-headed look at what our nature is. Thus, Antiphon writes:
[Life] and death are matters of physis; life is among the things that are advantageous (or life is from whatever is advantageous), death is among the things that are disadvantageous (or death is from whatever is disadvantageous). Advantages laid down by the laws are bonds on physis, but advantages laid down by physis are free. In right reason it is not the case that the things which bring pain benefit physis more than what brings joy, nor is it the case that the things that bring grief are more advantageous than what brings pleasure. (quoted in Nill Reference Nill1985: 59)
It’s human nature to preserve one’s life and fear death, to seek pleasure and flee from pain. It follows, for Antiphon, that a flourishing life is long, pleasant, and pain-free. That’s why the question ‘Why be moral?’ is so hard. Justice may get in the way of longevity, obstruct pleasure, and cause pain; injustice may do just the opposite.
We are in deep waters here. How does well-being in the ethical sense – that is, eudaimonia, the most desirable life – relate to concepts of well-being grounded in nature or physis that apply to other living things? Aristotle denies that non-human animals are capable of eudaimonia: ‘we call neither ox nor horse nor any other of the animals happy’ (NE 1099b33–35). But the sense of ‘well-being’ we apply to them cannot be simply different from the sense we apply to ourselves, as though ‘well-being’ were ambiguous between eudaimonia and a biological notion of flourishing. If it were, the next question would be whether the biological notion applies to us. If so, what role should it play in ethical thought? And if not, why not?
These questions are not the same as those invoked by the dualism of morality and self-interest. But they put pressure on our understanding of eudaimonia in ways that mirror the modern debate.Footnote 25 Williams puts the challenge starkly. ‘There is’, he warns, ‘the figure, rarer perhaps than Callicles supposed, but real, who is horrible enough and not miserable at all but, by any ethological standard of the bright eye and the gleaming coat, dangerously flourishing’ (Williams Reference Williams1985: 46). The thought is not that this figure is doing well according to a subset of practical reasons, or in one aspect of life. It’s that, when we approach him in the spirit with which we study other living things, we’re tempted to say that he’s thriving – and we’re not sure how to square this with the claim, which we would like to uphold, that his is not a desirable life.
The problem is genuine, but it looks more like the one Plato faced in the Gorgias and the Republic than the one we’ve inherited – or seemed to inherit – from modern moral philosophy. What we might call ‘the problem of flourishing’ has been addressed, in recent years, through a notable convergence of neo-Aristotelian and neo-Kantian thought: there are striking commonalities between Philippa Foot’s (Reference Foot2001) Natural Goodness and Christine Korsgaard’s (Reference Korsgaard2018: Ch. 2–3) Fellow Creatures. In outline, they see continuity but not identity between eudaimonia and the well-being of plants and non-human animals. There’s the notion of a plant’s well-functioning in its natural habitat. But when creatures have desires, as many animals do, there’s a second notion: final good as desirability, or the object of well-functioning desire. And when creatures can reflect on and evaluate their own functioning, as we do, there are two more: practical rationality or virtue as the object of well-functioning approval; and, with reasons understood in terms of practical rationality, a life that there is reason to desire.Footnote 26
This thumbnail invites more questions than it answers. What is well-functioning approval? Do its standards derive from the nature of reflection – as in Korsgaard’s ‘Kantian constititivism’ – or from specifically human nature, as Foot believes? Once we distinguish the well-functioning of plants and non-human animals from the psychological functioning that is the object of well-functioning approval, should we conclude that the standard of well-functioning approval is, reflexively, the very standard such approval sets? Does that create a vicious circle? Is it, in effect, the Kantian view? Can we make sense of human nature, as it figures in neo-Aristotelian ethics? If so, is human nature fixed or does it vary over human history?
It’s possible that, in grappling with these questions, we’ll find ourselves forced back towards hedonism or amoralism, or unable to sustain the continuity of plant or animal well-being with eudaimonia. But I believe they are the right questions to ask – and that, despite being relatively marginal in modern moral philosophy, they should supplant the popular, meaningless debate about morality and self-interest.Footnote 27
