The Decline of Egoism

Abstract

Sidgwick saw egoism as important and undefeated. Not long afterward, egoism is largely ignored. Immediately after Sidgwick, many arguments were given against egoism – most poor – but one argument deserves attention as both influential and plausible. Call it the “grounds objection.” It has two strands. It objects that there are justifying reasons for action other than that an action will maximize my self-interest. It also objects that sometimes, what makes an action right is a fact other than its maximizing my self-interest. I briefly explain and criticize many of the arguments given against egoism in the period, then explain and defend the grounds objection.

Any reader of Roger Crisp's Sacrifice Regained: Morality and Self-Interest in British Moral Philosophy from Hobbes to Bentham must be struck by the sympathetic attention the British moralists paid to various kinds of normative egoism. Some are rational egoists: I ought to perform some action if and only if, and because, the action maximizes my self-interest. Some are veto egoists, holding that maximal promotion of one's self-interest is a necessary condition on rational action. Some are dualists, holding that there are equally fundamental reasons of morality and self-interest. Almost everyone is very determined to argue that morality and self-interest do not conflict. One sees why, reflecting on this history, Sidgwick gave egoism such prominence in The Methods of Ethics.

What seems remarkable is how quickly the picture changed. Ross writes of “a generation for which the view that we should seek only our own pleasure is already out of date” – presumably referring to his own generation (Ross 1939: 274). After Sidgwick, egoism gets little support and, after Prichard, little attention. This does not seem to be due to Sidgwick's arguments. He saw egoism as undefeated.¹

I think the best explanation is an argument first given – though not in the best form – by Prichard. It continues through Carritt, Ross, and Ewing. There are slightly different versions of it. One strand objects that there are non-derivative justifying reasons for action other than that an action will maximize my self-interest.² Another objects that sometimes, what makes an action right – what, in Prichard's favorite terms, “renders” or “constitutes” an action right – is a fact other than its maximizing my self-interest. Since both strands concern the grounds for reaching a verdict, call the arguments together “the grounds objection.”³

I start with accounts of other arguments against egoism given in the period. They have obvious deficiencies (I). I then present the grounds objection (II) and worries about it (III).

Five preliminaries: First, I do not deny that factors other than the grounds objection played a role in the decline of egoism. Perhaps the view that the only “ought” is the moral “ought” took over. That view is prominent in Prichard and Ross. The issue is largely ignored by Crisp's British moralists and by Sidgwick, who grants that it “may be doubted whether [egoism] ought to be included among received ‘methods of Ethics,’” but includes it anyway (Sidgwick 1981: 119). Or perhaps the falsity of psychological egoism, or its irrelevance in arguing for normative egoism, did not sink in until after Sidgwick.⁴ Perhaps decline in belief in an afterlife (in which moral action was rewarded) made the counter-intuitiveness of egoism more apparent. Perhaps decline in looking for justifications of religious belief lessened the motivation to assert egoism, given that asserting egoism gave God the job of removing any conflict between morality and self-interest. Perhaps factors unrelated to philosophical arguments, such as the Great War, made egoism unpopular.

Second, while I suspect that the grounds objection was influential, it is not easy to show the influence. Prichard, Carritt, Ross, and Ewing knew one another and repeat very similar arguments. Reviewers of Prichard's Duty and Interest do not challenge Prichard's claim that showing that duty is in one's interest is irrelevant.⁵ Nor does Joseph in Some Problems in Ethics (largely directed at criticizing Prichard; Joseph 1931). Reviewers of Prichard's Moral Obligation, Carritt's The Theory of Morals, and Ross's The Right and the Good and Foundations of Ethics do not challenge the discussions of egoism.⁶ Nor does Johnson in his book on Prichard, Carritt, and Ross, Rightness and Goodness (Johnson 1959). Later comprehensive works such as Mabbott's An Introduction to Ethics and McCloskey's Meta-Ethics and Normative Ethics ignore egoism (Mabbott 1966; McCloskey 1969). Neither Ewing's mid-century review of British ethical thought nor Warnock's Ethics since 1900 mentions it (Ewing 1966; Warnock 1960). Both editions of Sellars and Hospers's Readings in Ethical Theory include discussions of egoism by Russell, Prichard, and Ross, but no others (Sellars and Hospers 1952). But one does not find many later writers saying that they put aside egoism because of the grounds objection. One exception is W. K. Frankena, who writes that “it is hard to believe that there are no other moral virtues or obligations which are independent of prudence or our own welfare” (Frankena 1963: 18).⁷

Since I take the grounds objection to be plausible, it is worth considering it, even if one cannot prove its influence.

Third, I do not claim that Prichard was the first to suggest the grounds objection. Sidgwick, for example, writes in passing that “common moral opinion certainly regards the duty or virtue of Prudence as only a part – and not the most important part – of duty or virtue in general” (Sidgwick 1981: 7–8). But Prichard was the first to make this strategy central.

Fourth, the egoism I consider holds that I ought to perform some action if and only if, and because, the action maximizes my self-interest. Sidgwick's egoist took self-interest to consist in pleasure. Ewing's egoist justifies action by appeal to her own “pleasure” or “happiness” (Ewing 1965: 21–31). Ross's egoist aims at “the advantage or pleasure of the agent”; his target is “Egoistic Hedonism” (Ross 2002: 16; 1939: 65; also Ross 1939: 273–74). Carritt takes the egoist to aim at “happiness” and discusses egoism in a chapter entitled “Hedonism” (Carritt 1928: 11–12; also 1947: 49–55). Prichard's target justifies moral claims by appeal to her own “happiness,” “advantage,” or “interest,” all of which Prichard seems to understand as feelings of “satisfaction” or “enjoyment” (Prichard 2002: 174). All note that egoism has been supported by some form of psychological egoism.

Sometimes, however, “happiness” or “interest” involves something other than pleasure or desire satisfaction. For example, Ewing considers an egoist who includes virtue as part of one's interest. He attributes the view to Plato and Aristotle. He gives the grounds objection to this “higher level” egoism as well: “Would not a man be a prig rather than a saint if he decided all actions by reference only to their effects on his own character?” (Ewing 1965: 31–32).⁸ Whether the objection rightly targets Plato, Aristotle, Green, etc. depends on whether they are best read as holding that there are non-derivative reasons for action other than reasons for me to bring about some benefit to myself. Since this is a thorny issue, I put it aside.⁹

Fifth, one might expect the rejection of egoism to discuss Sidgwick's “distinction passage,” which some think is his argument for egoism.¹⁰ It is ignored, however, other than by Hayward. I suspect that this is because the passage is both obscure and (as a referee noted) sometimes read as illegitimately deducing an “ought” from an “is.” Sidgwick's positive case for egoism is that it seems plausible to him (perhaps self-evident) and has widespread support. These claims are addressed.

I

The grounds objection is but one argument given in the period against egoism. Here I sketch some of the alternatives and indicate their difficulties.

(1) The move away from Sidgwickian egoism comes quickly.¹¹ Hayward, in The Ethical Philosophy of Sidgwick – published a year after Sidgwick's death – argues that “‘I ought to seek my own happiness’” is “a most violent and gross case of abuse of language. The ‘ought’ is absolutely out of place. What is meant is, ‘I should like to have pleasure, pleasure is pleasing,’ statements which … have nothing to do with morality” (Hayward 1901: 124–25). “‘Ought’ always implies some purpose connected with the higher living of self and others; in short some ideal of social and personal excellence” (127). Hayward usually relies on his sense of what is an abuse of language, but at one point he gives an argument. “Pleasure … is the most absolutely ‘subjective’ of all things. Now, inasmuch as ‘ought’ and ‘right’ imply an ‘objective’ standard, it is difficult to see how any ‘oughtness’ can attach to pleasure” (128–29).

The last argument is terrible. “I ought to promote my pleasure” is objective, in the sense that if two parties disagree about it, at least one must be wrong. This is so even if pleasure is subjective in the sense that it depends on the desires of each party. Hayward's argument would rule out not just egoism, but utilitarianism as well.

Hayward's appeal to language use is also easily met. The egoist can say that her “ought” is the notion of a justifying reason, with no necessary connection to “higher living of self and others.”

One distinctive feature of Hayward's treatment is that he thinks the problem for the egoist arises because the egoist says she ought to seek her own happiness. If one says she ought to seek her own perfection, he has no objection. His argument seems to be that ordinary thought often calls for sacrifice of one's happiness, but is uncertain about calling for sacrifice of one's perfection (Hayward 1901: 132). He may also be thinking that my perfection is partly constituted by my carrying out duties to others, and that “ought” implies a concern for others. (Seth, criticizing Sidgwick on similar grounds, writes that “the point of view of Duty is always … the point of view of Society” (Seth 1901: 184).)

(2) Moore argues that egoism is self-contradictory. The egoist holds (i) that I ought to maximize my good. He also holds (ii) that others should not maximize my good; they should maximize their own goods. (iii) “x is my good” means “my possessing x is good.” (iv) If my possessing x is good, others ought to maximize my possession of it. But then the egoist has contradicted himself, holding that others should and should not maximize my good (Moore 1903: 98–102).¹²

There are two standard worries.

First, egoism might be stated without using “my good.” Sidgwick's official statement is that the egoist adopts “his own greatest happiness as the ultimate end of each individual's actions. … [I]t is reasonable for a man to act in the manner most conducive to his own happiness” (Sidgwick 1981: 119; also xix, 11, 418, 497, 498). Putting egoism in terms of “ought” or “reasonable” and happiness avoids Moore's argument. It is easy for the egoist to deny the equivalent of (iv), “if I ought to pursue my happiness, others ought to maximise my happiness.”¹³ Thus Prichard criticizes Moore by noting that “what [egoism] represents as what a man ought to aim at is not his own goodness … but his own happiness” (Prichard 2002: 176).

Second, one might keep “my good” but deny (iv). One might do so either by denying that the goodness of my happiness by itself gives anyone a reason to promote it or by thinking that it gives a reason only to me. Broad seems to give the former objection to Moore. The egoist can hold that

equally pleasant states of mind are equally good things, no matter whose states of mind they may be. But he holds that each of us is properly concerned, not with all good things, but only with a certain restricted class of good things, viz., those which are states of his own mind … [I]t is unfitting for A to desire as an end anything that falls outside this class, no matter … how much better it may be than anything that falls within the class. (Broad 1930: 243)

The egoist can hold “the extreme deontological view that such a [self-sacrificing] action would be unfitting, and that its unfittingness suffices to make it wrong on the whole no matter how intrinsically good its consequences might be” (Broad 1930: 244). Moore is assuming the “purely teleological view” that “‘right’ and ‘conducive to intrinsically good results’ are mutually equivalent” (Broad 1930: 244; also Broad 1968: 44–46; 1971a: 347–50; 1971b: 266; 1985: 212–14; Ewing 1965: 33). Similarly, Prichard objects to Moore that whether I ought to promote the good of A or B depends on my relation to A and B – on whether, for example, A is me and B is another person – not just on the amounts of good produced (Prichard 2002: 217).

Many have given the second objection: one might think that “my good” should not be analyzed in terms of “good” but rather is a distinct concept, linked only to agent-relative reasons (e.g., Mackie 1976, Smith 2003).

Moore's argument seems to have convinced few. Rashdall is one exception. He repeats and endorses Moore's argument (Rashdall 1913: 63 n.2; 1916: 119). He also gives a version that does not depend on “my good.” He asks the egoist

why Reason should attach more importance to one man's pleasure than to another's. If it is pleasure that is the end, it cannot matter … whose pleasure it is that is promoted. The greater pleasure must always be preferable to the less pleasure, even though the promotion of the greatest pleasure on the whole should demand that this or that individual should sacrifice some of his private pleasure. (Rashdall 1924: 45; also 48)

This, like Moore, seems to depend on the “purely teleological view,” applied to “ends” rather than “good.”

Of course, one might think that the purely teleological view is correct. Russell writes that

if good is to be pursued at all, it can hardly be relevant who is to enjoy the good. It would be as reasonable for a man on Sundays to think only of his welfare on future Sundays, and on Mondays to think only of Mondays … It is … so evident that it is better to secure a greater good for A than a lesser good for B, that it is hard to find any more evident principle by which to prove this. (Russell 1966: 47–48; also Moore 1965: 100; Laird 1926: 226; Ewing 1947: 188; 1965: 76; 1966: 69)

But not everyone agrees with the purely teleological view. Like the egoist, Prichard, Ross and Carritt think that an action can be wrong even when its consequences are optimific. For all, this is sometimes because a prior act invokes a duty other than beneficence: I ought to keep a promise, even when doing so produces less good than breaking it does. For some, this is because beneficence itself can be a duty to bring about less good: Prichard (though not Ross) thinks that my relation to the good of my father can by itself justify bringing about less good rather than more (Prichard 2002: 2, 217; also Broad 1930: 243–45; 1971b: 279–80; 1985: 221–22). The similarities between deontology and egoism noted by Broad show that they cannot reject egoism as Moore, Rashdall, and Russell do.¹⁴

(3) Prichard, especially, gives many other arguments.

If we ask why we should keep our engagements, one answer is that doing so helps us satisfy our desires. Prichard objects that this

answer is, of course, not an answer, for it fails to convince us that we ought to keep our engagements; even if successful on its own lines, it only makes us want to keep them. And Kant was really only pointing out this fact when he distinguished hypothetical and categorical imperatives … [T]hough we shall lose our hesitation to act … we shall not recover our sense that we ought to do so. (Prichard 2002: 9)

[W]e need only imagine ourselves unwilling to do some action, e.g. to get up early in the morning, and then thinking of some substantial gain which it would bring us which we consider would more than outweigh the loss of comfort. For we then find that though the thought might make us less unwilling to do it, it would do nothing to make us think that it was a duty on our part to do it. And by doing this we can come to recognize that conduciveness to our advantage is simply irrelevant to the question whether it is a duty to do some action. (Prichard 2002: 171)

[W]e have only to ask ourselves whether some act's being that which would do most to make us happy would render it what we are bound to do, to know that it would not. [But] [t]he fatal objection … is that it resolves the moral “ought” into the non-moral “ought,” representing our being morally bound to do some action as if it were the same thing as the action's being one which we must do if our purpose is to become realized. And in consequence, strictly speaking the theory is not a theory of obligation, or duty, at all. (Prichard 2002: 188; also 43)

The argument seems to be that if I ought to do some action, or the action is my duty, this is so categorically – it would be my duty regardless of my desires, and regardless of any happiness I would gain or lose.¹⁵

One problem with the argument is that, as Sidgwick thought, egoism can be seen as making categorical “ought” claims (Shaver 2014: 316). If, in Prichard's example, I remain unwilling to get up early, the egoist will say that, regardless of what I want now, I ought to get up.

Prichard slides from thinking of the egoist as justifying duties on the basis of the agent's own “good,” “advantage,” and “happiness” (Prichard 2002: 8) to thinking of doing so on the basis of the agent's desires: “if … we put to ourselves the question ‘Why should I do so and so?’, we are satisfied by being convinced either that the doing so will lead to something which we want … or that the doing so itself … is something that we want or should like” (Prichard 2002: 8). He then notes that the “ought” justified is the “ought” of the hypothetical imperative. But a justification on the basis of one's good, advantage, or happiness need not appeal to what on balance we want, either at the time of acting or later. Sidgwick, for example, thinks of happiness as a matter of those mental states which, at the time of experiencing them, one finds desirable for their own sakes. He admits that one might not, at any time, most desire these mental states. Hedonists who think of pleasure as independent of desire (and desirability) – perhaps it is a distinct feeling or attitude – admit the same possibility. So too do Ewing's “higher” egoists: they justify my doing my duty by appeal to the virtue I would gain, not by the satisfaction of my desires. In all these cases, the egoist's “ought” is categorical.

(4) Prichard sometimes relies on our “feeling” or “sense” of obligation (2002: 1, 9, 11, 12, 15, 16, 17, 55, 59, 73), our “sense that we ought to do” an action (9, 10, 12, 13, 19), our “sense of the rightness of an act” (14), “the feeling of imperativeness” (10, 55). He often contrasts this with desire (1, 9, 11, 15, 16, 55, 59, 73, 159–60). A “morally good action is morally good … because it is done because it is right, i.e. from a sense of obligation” (14). The “apprehension of a moral principle involves a certain emotion, viz. the sense of obligation” (55).

There is no such thing as the mere thought of the moral law; what exists is the sense of obligation. In other words, our recognition that an act of a certain kind ought to be done involves of its very nature emotion. (Hence the phrase sense of obligation.) It is not that the recognition is an act of intelligence and the emotion a non-rational feeling joined to it. The emotion is through and through rational, and the recognition and the emotion, though distinguishable, are inseparable. (Prichard 2002: 73)

Prichard notes repeatedly that various attempts to explain the basis of moral obligation fail to capture this sense. If so, the objection to egoism is that showing that an action is in my interest fails to produce this sense. He could make this claim without any views about the relevant “ought” being categorical.

The problem is that one wants some explanation for why the sense of obligation occurs when it does. Taking it to be a “rational” emotion gives that explanation: there is a belief component to the emotion. But if the belief is that one must categorically do the action, the sense of obligation argument inherits whatever problems that suggestion encounters.

II

I turn to the grounds objection. It is a more successful pair of objections.

Egoism says (1) that the only justifying reason for action is that the action will maximize my self-interest. One objection is

(O1) there are reasons for action other than that an action will maximize my self-interest (e.g., that it maximizes the pleasure of others, that it is the fulfilment of a promise).

Egoism also says (2) that what makes an action right is that it maximizes my self-interest. One objection is

(O2) sometimes, what makes an action right is a fact other than its maximizing my self-interest (e.g., that it maximizes the pleasure of others, that it is the fulfilment of a promise).

O1 concerns “flavorless” (and so not specifically moral) reasons for action. They play the role of Ross's prima facie duties: they count in favor of or against actions, but must be weighed against one another to see what one ought to do all things considered.¹⁶ They are factors capable of making it true that I ought to do some action. O2 concerns what I ought to do, in a flavorless sense, all things considered. It concerns what actually makes it true that I ought to do some action. One who gives O2 will endorse O1: if, say, being the fulfilment of a promise makes an action right, being the fulfilment of a promise must be a reason for action. One could, however, endorse O1 without O2 (though this is unlikely): one could think being the fulfilment of a promise (and all other non-egoistic reasons) are reasons, though the decisive reason is always provided by self-interest.

Prichard and Ross stress that egoism makes a mistake about grounds. But they fail to clearly give O1 and O2, in two ways. They often claim that maximizing self-interest is not a reason at all, and they seem to take the reasons in question to be moral rather than flavorless. O1 and O2 are not clearly stated until Ewing.

Thus Prichard objects to Sidgwick that:

we do not … think it a duty on our part to aim at … our happiness; nor do we … think it is a duty to … do those acts which we think will lead to our happiness … [W]e may … think that it will pay us to get up; we may also think that it is weak … to stay on in bed, or even that we shall be ashamed of ourselves if we do; but we shall not think it a duty to get up unless there supervenes some consideration other than that of conduciveness to our happiness. (Prichard 2002: 135; also 204, 206)

Prichard relies on a distinction between “duty” and other forms of appraisal. Exactly how the distinction should be drawn is unclear, but it must follow from it that it is a necessary condition on a duty that it is grounded in something other than conduciveness to my happiness.¹⁷

This is a frequent theme in “Duty and Interest”:

though we may find ourselves quite unable to state what it is that does render an action a duty, we ordinarily think that, whatever it is, it is not conduciveness to our advantage … [W]hen we seriously face the view that unless an action be advantageous, it cannot really be a duty, we are forced both to abandon it and also to allow that even if it were true, it would not enable us to vindicate the truth of our ordinary moral convictions … [W]herever in ordinary life we think of some particular action as a duty, we are not simply thinking of it as right, but also thinking of its rightness as constituted by the possession of some definite characteristic other than that of being advantageous to the agent … [I]f we were to maintain that conduciveness to the agent's advantage is what renders an action right, we should have to allow that any of our ordinary moral convictions … is simply a mistake, as being really the conviction that some particular action is rendered a duty by its possession of some characteristic which is not that of being advantageous. (Prichard 2002: 26, 27, 29, 30)

As Prichard puts the point later, “no one … really thinks that the fact that a certain action would be advantageous … to himself constitutes the action a duty on his part” (Prichard 2002: 123; also 122).

Ross gives a similar argument. Egoism

comes to grief over the fact, which stares us in the face, that a great part of duty consists in an observance of the rights and a furtherance of the interests of others, whatever the cost to ourselves may be. Plato and others may be right in holding that a regard for the rights of others never in the long run involves a loss of happiness for the agent … But this, even if true, is irrelevant to the rightness of the act. (Ross 2002: 16)¹⁸

The egoist can reply (as against Hayward) that she is proposing an account not of duty, but of what one has a justifying reason to do. Prichard and Ross might be correct that considerations of self-interest are irrelevant to “duty” or moral rightness. But such considerations are relevant to what I ought to do in a flavorless sense. When (say) I save a drowning person to get a reward, the reward gives me a reason for the act.¹⁹

When Prichard and Ross argue that considerations of self-interest are irrelevant, their argument seems to depend on thinking that the only reasons are moral ones. The egoist can reply that she is concerned with flavorless reasons. But Prichard and Ross also argue that considerations of self-interest are not the only considerations. There is no reason to think that here they are thinking of flavorless reasons. But their argument does not depend on thinking that the only reasons are moral ones. That an action would help another seems not only to give a moral reason to do it; it seems to give a flavorless reason as well.

Thus Prichard writes that the rightness of an action is “constituted by the possession of a certain characteristic other than that of being advantageous to the agent,” such as “fulfilling the promise” (Prichard 2002: 122). We think an action is made right by having characteristics such as “fulfilling the promise which we made to X yesterday” or “looking after our parents” (Prichard 2002: 29, 121; also 4, 12–13).

[T]hat advantageousness is what renders an action right … is obviously something which no one is going to maintain … For he will be involved in maintaining not only that it is a duty to do whatever is for our advantage, but that this is our only duty. And the fatal objection to maintaining this is simply that no one actually thinks it. (Prichard 2002: 29; also 122)

Similarly, Ross writes that:

[e]goistic Hedonism is put out of court by the fact that stares us in the face, that it is consideration for the rights or interests of others, far more often (to state the matter very mildly) than consideration for our own interests or rights, that makes us think it is our duty to behave in a certain way. Whether consideration of our own rights or interests ever gives rise to the thought that we ought to behave in a certain way, in distinction from the thought that it would be prudent or sensible to behave in a certain way, is a question to which I hope to come later. But that it is the sole consideration which gives rise to the thought of duty is too palpably untrue to need serious discussion. (Ross 1939: 65)

Here, Ross is neutral about whether considerations of self-interest could make it true that I have some duty.²⁰ His argument against egoism resembles his argument against ideal utilitarianism: he need not deny that my welfare, or facts about what is good, give me reasons or make something my duty; what he denies is that these are the only things that give me reasons or make something my duty.

This is close to O2. But Prichard and Ross still seem to think of the reasons as moral. Indeed, they seem not to distinguish a moral and a flavorless “ought.”²¹

Ewing gives the best statement of O1 and O2:

[A]ccording to egoistic hedonism the only reason why anything is wrong is because it is not conducive to the agent's greatest pleasure. Even if in fact it is the case that it is never conducive to my own greatest pleasure to hurt others, it should be plain that this is not the main reason why it is wrong. If we can see clearly that our own pleasure is good, we can see just as clearly that the fact that an action needlessly and intentionally hurts another is quite sufficient to make it wrong, whether it also hurts me or not. (Ewing 1965: 28)

Ewing suggests that egoism has been popular because

the fact that a course of action is conducive to one's own happiness is … a reason for adopting it. It is a subject for dispute whether this makes it morally obligatory or merely prudent to act in the way proposed, but at least it is a good reason of some sort for doing so … Now the project of bringing all Ethics under a single principle so that there is just one kind of circumstance which decides whether an act is right or wrong is very attractive to thinkers, and so when we have found a principle which obviously does give valid reasons for action, there is a temptation to bring all ethical judgments under it. (Ewing 1965: 28)

Ewing is neutral here on whether promoting one's happiness is “morally obligatory or merely prudent.” Elsewhere he argues, against Prichard and Ross, that one ought, in the sense of “fittingness,” do so (Ewing 1939: 3, 8, 20; 1947: 133, 1 160, 162–63, 194; 1948: 108; 1959: 92, 111–14; 1970: 3n.). My interest is “a good reason of some sort.” Ewing employs a concept of “fittingness” that is flavorless. Some fittingness claims need not be thought of as claims about moral reasons (though all claims about moral reasons involve claims about fittingness). For example, “certain actions are fitting simply because they are conducive to the agent's happiness, yet it may at least be doubted whether this is a sufficient ground to make them morally obligatory” (Ewing 1947: 133; 1939: 20; also 1947: 160, 163; 1959: 91–2).²² Ewing can then say that there are factors which make an action fitting other than its conduciveness to one's happiness, such as that the action helps another.

I have not noted Carritt. I think he gives O2. He argues against thinking that acts are right only because they benefit the agent:

the phrase “I ought,” if it has any meaning at all different from “I want,” implies “Whether I want or not.” And it seems, to the plain man at least, absurd to tell him that he ought to try to get the greatest pleasure whether he want it or not. He might admit that he ought to try to get his own greatest happiness rather than the second pint of beer which at the moment he desires more. But he may happen at the moment to desire the welfare of his country or of his children, or the advancement of knowledge, or the leaving of an unsullied reputation, more than he desires what appears to him likely to lead to his happiness upon the whole; and it will be hard to convince him that patriotic or parental or scientific effort, or even the pursuit of posthumous fame, is wrong. And he often thinks he ought to sacrifice the happiness he does want. (Carritt: 1928: 11)

Carritt does not deny that it makes sense to say “you ought to get the greatest pleasure for yourself, whether you want that now or not.” Nor does he deny that the claim is sometimes true.²³ His point is that we think that the claim is often false, since we think considerations other than my pleasure give me reasons.²⁴

I have three observations. First, O1 and O2 can be given whether or not one thinks that the right act must coincide with the act that produces the most good. They are not answered by an egoist's arguing that morality and self-interest coincide. They do not depend on taking the egoist to offer only hypothetical imperatives. They do not depend on restricting “ought” or “duty” to morally loaded senses.

O1 and O2 can also be given without agreeing with Prichard that it is a mistake to ask for a justification of morality. T. H. Irwin objects that while Prichard is right that one should not, in deciding what to do in particular cases, always be asking about “the point of morality,” he is wrong to rule out that question asked by a “reflective agent” in “a cool hour” (Irwin 2009: 885). One might agree that Prichard should allow the question. What he is right to reject is the answer to the question given by the egoist.

Second, O1 and O2 do not claim that egoists make false claims about what is right. O1 and O2 concern the grounds of judgments of rightness. The objection that egoists make false claims about what is right was made, though less commonly. Carritt suggests it in the passage quoted above (see also Carritt 1947: 50). Ewing writes that “some of the worst acts ever done could be justified if egoistic hedonism were true.” He gives a case of deathbed revenge that maximizes the happiness of the person dying (Ewing 1965: 24–25; also Broad 1971b: 272).²⁵ Stressing objections to the egoist's grounds lets one avoid engaging with particular egoist responses intended to capture common-sense judgments about rightness – as Prichard writes, “[t]here is really no need to consider in detail whether these arguments are successful, for even if they are successful, they will do nothing to prove what they are intended to prove, viz. that the moral convictions of our ordinary life are true,” since these moral convictions include thoughts such as “the action [is] rendered a duty by its being the keeping of our promise” (Prichard 2002: 30, 29).

For some, there is a further reason for preferring objecting to reasons, as in O1, to objecting to verdicts about what is right. Ross holds that our “judgments about our actual duty in concrete situations have none of the certainty that attaches to our recognition of the general principles of duty” (Ross 2002: 30). Ewing held that the “primary ethical intuition is not that an action is fitting or unfitting as a whole, but that it is fitting or unfitting in certain respects” (Ewing 1947: 195; 1966: 80). If so, it makes sense to argue that the egoist is wrong about what reasons there are (O1) rather than about our “actual duty in concrete situations.”²⁶

Third, my interest is egoism. Dualists and veto egoists can accept at least O1. They agree that there are non-egoistic reasons. Carritt, however, can be read as arguing that considerations of one's own happiness are no more important than those of the welfare of one's country or children, or of reputation or knowledge. There is nothing special about the category of egoistic reasons, such that they form a dualism with or can veto non-egoistic reasons. Perhaps this is why many who now admit something like a dualism replace egoistic reasons with agent-relative ones.

III

I turn to objections to O1 and O2. The first is an objection to taking them as so important; the second and third are to their adequacy as a reply to the egoist.

Egoism and utilitarianism

O1 and O2 are variants on a familiar argument against ideal utilitarianism. Prichard, Carritt, and Ross argue that even if ideal utilitarianism captured all the right verdicts on cases, (i) it holds, wrongly, that there is only one duty, or (ii) we know that the reason for, say, keeping a promise is not that doing so maximizes good, or (iii) we know that a reason for, say, keeping a promise is simply that I have made it. These arguments have been influential, but certainly did not kill off ideal utilitarianism. Egoism has fared less well. This suggests that I am wrong to highlight the parallel arguments against egoism.

I think, however, that O1 and O2 are better arguments against egoism than against at least the popular versions of ideal utilitarianism. Ideal utilitarians usually do not limit goods to those connected to welfare.²⁷ In adding goods to pleasure, they do not say that knowledge, for example, is good only insofar as it benefits the knower. This makes it easier for the ideal utilitarian to capture the reasons and truth-makers O1 and O2 rely on. Say, with Ewing, Johnson, and Blanshard, that the state of affairs in which a promise has been broken is bad. Avoiding this badness is a reason to keep the promise. Prichard and Ross say that a reason to keep a promise, or what makes it true that one ought to keep a promise, is “because you made the promise”; Ewing says it is “because the state of affairs in which a promise has been broken is bad.” It is doubtful that common sense sees the former as obviously right and the latter as obviously wrong.²⁸

Egoism, however, is almost always put in terms of welfare. And few egoists think of welfare as constituted (in part) by things such as “I have kept a promise.” They instead argue that keeping the promise has a causal effect on one's welfare; the reason to keep stems from this effect. This makes them more vulnerable to O1 and O2.²⁹

Utilitarians who limit goods to welfare are also more vulnerable. But they are still less vulnerable than egoists. In part this is because they agree with Ross that there is an underived reason to help others. In part this is because it seems more plausible to derive Ross's “special” duties, such as fidelity, from their role in producing general welfare than it does to derive them from their role in producing welfare in the agent.

Begging the question

An egoist might object that, in an argument against egoism, one cannot start by assuming that an action is made right by something other than being advantageous to the agent.³⁰

This misunderstands at least Prichard's framework. He sees the egoist as trying to “vindicate the truth of our ordinary moral convictions.” Moral philosophy arises from doubt about these convictions, given the costs of acting on them (Prichard 2002: 7, 8, 19, 23, 26, 120–21). Egoism is offered as removing the doubt by showing that really there is no cost. Prichard's point is that egoism will not remove the doubt, even if it shows that there is no cost, since it does not vindicate the convictions – given that the convictions include beliefs about grounds.

It does not follow that Prichard et al. have nothing to say against an egoist who is not interested in vindicating these convictions. That egoist must hold that we are mistaken in thinking that, for example, that an action benefits another is a reason for doing the action. Given that this is something most of us are confident about, the egoist needs to produce an argument to show that we are mistaken.³¹ Perhaps the egoist can discredit this belief – perhaps as Peter Singer and others have tried to discredit anti-utilitarian beliefs, by explaining them on the basis of religion or emotion or evolution. That project is very different from the traditional egoist project of showing that doing our duty is in our interest. Prichard can ask to see the argument (and argues that no argument that relies on psychological egoism is successful). I suspect that Prichard et al. place more confidence in the belief that (say) that an action benefits another is a reason for doing the action than they do in the premises of any discrediting explanation. But of course whether they are correct depends on the explanation offered.

Disagreement

Alison Hills objects that a claim frequently used against the egoist in O1 –

(H) “I have a reason to help those in need”

– is subject to peer disagreement and that peer disagreement, at least here, justifies suspending judgment. She grants that in some cases of peer disagreement, where there is “only the odd disagreement or mistake,” it is not rational to suspend judgment (Hills: 2010: 162). If so, a moralist could avoid suspending judgment about H in response to egoist denials of it, since egoists are hardly common (and perhaps not peers, insofar as they depend on psychological egoism). But H, she argues, is subject to much wider disagreement, since moralists who believe H “think that it is true for different reasons.” For example, utilitarians think H is true because helping maximizes happiness; Kantians think H is true because not helping cannot be willed as a universal law. She concludes that the agreement on H is “superficial” and so moralists should suspend judgment about it (163). This would rule out O1, at least for a version which gave “that another would be helped” as a non-egoist reason.

Here are some replies.

1. (i) A utilitarian might say that a Kantian is not justified in believing H, given that the Kantian believes H on the basis of false beliefs. But I do not think it follows that the agreement is “superficial” or really absent. The utilitarian would say “I agree that I have a moral reason to help those in need, but not for the reasons you give.”

2. (ii) Utilitarians and Kantians seem to agree on some of what makes H true. They agree on a duty of beneficence; they agree that (say) the death of the person in need would be bad, and that that is part of the reason why one has a reason to help; they agree that it is virtuous to help. Each might say of the other that he or she has got something right.

3. (iii) One can construct an account of what makes H true on which non-egoists agree – the disjunction of the utilitarian and Kantian (and other) explanations.

***

O1 and O2 have an obvious drawback: they say nothing to an egoist who denies the intuitions they rely on. In this they differ from more ambitious arguments, such as the argument from personal identity suggested by Sidgwick and developed by Derek Parfit.³² But this also may explain their role in the decline of egoism. The ambitious arguments depend on controversial premises. I suspect that positions usually decline when they conflict with intuitions which are widely shared; when the intuitions are not undercut; and when the historical arguments for the position – here, almost always from psychological egoism – have been acknowledged by both sides to be bad.³³