¹ The term is used by Nozick, Robert in Anarchy, State, and Utopia (New York: Basic Books, 1974)Google Scholar. I should emphasize that it is the term that I am criticizing here. Nozick's account of side constraints anticipates some of what I will say in this essay about deontological reasons: in particular, that they are based on the Kantian notion that people must not be treated as means (ibid., p. 30), and that they will seem puzzling only to someone who assumes that “a moral concern can function only as a moral goal” (p. 28).Google Scholar

² Rawls, John, A Theory of Justice (Cambridge: Harvard University Press, 1971). See especially pp. 139–42.Google Scholar

³ This formulation may give rise to the misapprehension that I do not think that there can be duties to the self, or that questions of value cannot arise for the self. What I actually think is that the relations between stages of a self have many of the same features as the relations between separate persons; if stages of the self are to lay each other under normative demands, they too owe each other reasons they can share. But, for reasons indicated in Section IV of this essay, it follows that the self cannot have a reason it could not, in principle, share with others. This gives the question of the reasons we can share with others a certain priority, and that is the focus of this essay. Duties to the self do not get an adequate treatment here.

⁴ In this essay, references to Nagel's works will be inserted into the text. The abbreviations used are: “PA” for The Possibility of Altruism (Princeton: Princeton University Press, 1970)Google Scholar; and “VFN” for The View from Nowhere (New York: Oxford University Press, 1986).Google Scholar

⁵ In The Possibility of Altruism, Nagel uses the terms “subjective” and “objective.” But these terms are awkward because they are used in so many different ways. “Subjective” may be used in a metaphysical sense, to refer to how things are for someone, assuming that things might be different for others. Or it may be used in an epistemological sense, to refer to how things seem to someone, assuming that things might in fact be different from the way they seem. To avoid confusion, notice that in this sense the subjective need not be personal or individual. Something could seem the same way to every human being and yet not be that way from some more objective point of view. A mirage, although seen by everybody, is in this sense a subjective illusion; more controversially, one might say that colors are a feature of the subjective experience of creatures with color vision. In The Possibility of Altruism, Nagel uses “subjective” in a metaphysical sense: a subjective value is one that is good-for some individual. In The View from Nowhere, however, Nagel uses that term to refer to what seems to be a reason. Here his project is first to assert that it seems to us as if we had reasons and values (from a subjective or personal standpoint), and then to raise the question whether or not, from a more objective or impersonal standpoint, that appearance reveals itself as an illusion of the subjective standpoint (VFN, ch. 8). For this reason, Nagel borrows Derek Parfit's terms to cover his earlier distinction. What he had called a subjective value becomes an agent-relative value, which is a source of reasons for a particular agent, but not necessarily for others. What he had called an objective value becomes an agent-neutral value, which is a source of reasons for any agent (VFN, p. 152)Google Scholar. Parfit introduces these terms in Reasons and Persons (Oxford: Oxford University Press, 1984), p. 143.Google Scholar

⁶ James Dreier has pointed out to me that in styling my project an attack on the distinction between relative/subjective and neutral/objective, I might give the impression that I think this logical distinction in not exhaustive, which it obviously is. My quarrel, as will emerge, is really with Nagel's account of the source of these reasons, which suggests that values and reasons originate either from personal, idiosyncratic desires or from metaphysical realities of some kind. I thank Dreier for the point.

⁷ More accurately, Nagel's view is that if I do not, I will suffer from dissociation between the personal and impersonal views I can take of myself (PA, ch. 11).

⁸ Or, as one might put it, that every person, being equally real, is a source of value. But Nagel does not put it that way: he moves, as we shall see, from a focus on the (equal) reality of people to a focus on the reality of their reasons. In one sense, I believe his mistake lies here, and that he would have arrived at a more Kantian and, as I think, more correct position if he had not made this move.

⁹ See especially Moore, G. E., “The Conception of Intrinsic Value,” in Philosophical Studies (London: Routledge and Kegan Paul, 1922)Google Scholar. Values could be independent of agents in this sense and still always involve agents in another sense: agents and their experiences might always be parts of the complex “organic unities” which G. E. Moore thought were the loci of value. See note 24. I thank Arthur Kuflik for prompting me to be clearer about this.

¹⁰ Another view makes good-for-ness objective in this sense. It is a fact about the universe that a certain thing is good for me or for you. I think that this is the view that G. E. Moore, from whom I borrow the idea of formulating these notions in terms of good-for and good-absolutely, found incoherent. See Moore, , Principia Ethica (Cambridge: Cambridge University Press, 1903), pp. 97ffGoogle Scholar. I do not know whether it is incoherent, but it is not tempting.

¹¹ For another account of Intersubjectivism, see Darwall, Stephen, Impartial ReasonGoogle Scholar, part III.

¹² It may help to give examples of the sort of position I have in mind here. I am thinking, as will become clear, of Kant's claim that respect for the humanity in the person of an other requires you to share his ends; or of Hume's view that the virtues get their value from a shared evaluative standpoint. According to Kant's argument, a person's subjective ends become objective ends in the eyes of those who respect his humanity; according to Hume's, the character traits subjectively valued by the members of a person's own “narrow circle” become objectively valued when viewed from a general point of view which we share. As I suggest below in the text, one may read Nagel's projects as forms of Intersubjectivist constructivism as well. I do not know whether an Intersubjectivist position must be one in which objective values are constructed from subjective ones, but the Intersubjectivist positions with which I am familiar do take this form.

¹³ This is not to say that there cannot be values that are best understood as “good for us.” But these will not be the results of addition. They will exist when the two of us stand in a relationship to which the value in question is relevant. In this way the birth of a child might be good for a couple, or the conclusion of a treaty might be good for a nation. These are collective, not aggregative, goods.

¹⁴ Obviously, the array of logically possible positions goes far beyond the two that are schematically described in the text. One could be an Intersubjectivist and yet think that values can be added across the boundaries of persons. One could be an Objective Realist and yet deny that values can be added-not only across the boundaries of people, but at all. In Reasons and Persons, for example, Derek Parfit explores the possibility that weighing and compensation cannot take place even within the boundaries of a life (pp. 342–45). I am not concerned to discuss all possible theories of neutral value, but only the two I find most natural. I shall assume throughout this essay that if there is any objection to adding values, it comes from the consideration that everything that is good or bad is good or bad for somebody, and that values can be added within individual lives. I shall also assume that the view that everything that is good or bad is good or bad for someone, is most naturally associated with some form of Intersubjectivism.

¹⁵ Sometimes, Nagel seems to imply that all it amounts to for a reason to be “really there” is that it can be assimilated to the objective standpoint without contradiction or incoherence. This unites the practical and the epistemological projects described in the text, and the result would be an Intersubjectivist form of realism. Nagel's values would be part of reality because we put them there, rather the way that, according to Kant, causes are part of empirical reality. This view would have the merit of giving us realism without metaphysics. But it would require a transcendental argument for the category of objective value, and I do not myself see how, in the absence of Kant's own firm division between theoretical and practical reason, this is to be achieved.

¹⁶ Nagel might reply that all that follows is that, if we exist, we have reason to stop the animal's pain. But if pain has a value of its own it seems more natural to say that there just is a reason to stop the animal's pain, although the animal cannot see and respond to it.

¹⁷ One reason I take this option to be important is this: I think that its lack of ontological or metaphysical commitments is a clear advantage of Intersubjectivism; we should not be Objective Realists unless, so to speak, there is no other way. This is not just because of Ockham's razor. A conviction that there are metaphysical truths backing up our claims of value must rest on, and therefore cannot explain, our confidence in our claims of value. Metaphysical moral realism takes us the long way around to end up where we started-at our own deep conviction that our values are not groundless-without giving us what we wanted-some account of the source of that conviction.

¹⁸ Nagel backed off from his earlier position by degrees. At the time he added the postscript to The Possibility of Altruism (quoted above), he had decided that it was possible that an individual's subjective reasons may sometimes have a legitimate normative force for her that goes beyond that of their objective correlates. If my happiness is good-absolutely, we both have a reason to pursue it, but perhaps I find an additional or a stronger reason in the fact that it is good-for me. This seems to be an intermediate position between the views of The Possibility of Altruism and The View from Nowhere.

¹⁹ See Smart, J. J. C. and Williams, Bernard, Utilitarianism: For and Against (Cambridge: Cambridge University Press, 1973), pp. 100ff.CrossRefGoogle Scholar; and Williams, , Moral Luck (Cambridge: Cambridge University Press, 1981), pp. 1–19.CrossRefGoogle Scholar

²⁰ See Scheffler, Samuel, Tlie Rejection of Consequentialism (Oxford: Oxford University Press, 1982), p. 41Google Scholar. I discuss Scheffler's views briefly in Section V below.

²¹ Many of which can be found in the two pieces by Williams cited in note 19.

²² Several readers have pointed out to me that this label, together with the example I go on to discuss, might suggest that all personal projects are in some way competitive. I do not mean to imply that, and in fact I discuss some noncompetitive ones below. But the choice of an example of a personal project which is competitive seems to me to be useful, since such projects are especially resistant to objectification of either an Objective Realist or an Intersubjectivist kind.

²³ Barbara Herman has pointed out to me that the external account works better for natural beauty than for art, since works of art are socially embedded and therefore their value seems more relative to our interests.

²⁴ One may wonder whether an Objective Realist can accommodate the cases of clearly relational value, like the case of chocolate. The answer is yes. The Objective Realist does not have to place the intrinsic value in the chocolate. He can place it in the experience of a human being enjoying eating the chocolate. That is to say, he can construct what G. E. Moore called an organic unity and place the value-creating relationship inside of it. (See Moore, , Principia EthicaGoogle Scholar, ch. 6.) The trouble with this strategy is that it conceals the fact that the value is really relational, and the possibility, embraced by Intersubjectivism, that all values are really relational. For further discussion, see my “Two Distinctions in Goodness,” Philosophical Review, vol. 92 (1983), pp. 169–95, especially 190–93.Google Scholar

²⁵ Of course, this way of putting it assumes that no one else has any desires about the campus that could weigh against mine. In that sense, it assumes that I am the only person in the world who cares about the campus. Some people, when they realize that, are tempted to think that under those improbable circumstances I would be the right person to determine what counts as a good state of the campus.

²⁶ This includes your happiness or pleasure, which perhaps makes what I say here controversial. I am claiming that if I care about you I want your ambitions to be fulfilled, and not only in order to make you happy. I want them to be satisfied simply because you do. This is why deathbed wishes are entrusted to loved ones. Of course, this does not mean that I will never oppose your pursuit of an ambition if I foresee that it will make you miserable. But that is a matter of weighing, not a matter of refusing to give the ambition any weight of its own. Something here depends on one's views about rationality. There are people who hold that it is only rational to fulfill those ambitions that will make us happy. If you hold this view about rationality, you are likely to encourage and help your friends only to do what will make them happy, just as you are likely to give up your own more dangerous ambitions. But if you hold that it is sometimes rational just to do what you think is important without regard for your happiness, you are likely to respect a friend's desire to do what he thinks is important without regard to his happiness as well. Of course, if you hold the view that happiness just consists in doing what you think is most important, these issues can not even arise.

²⁷ On this point, see also Darwall, Stephen, Impartial Reason, p. 139.Google Scholar

²⁸ When introducing the idea (in PA, p. 29)Google Scholar, Nagel writes as if a motivated desire were one arrived at through deliberation. But on his own view prudence is a motivated desire, and most of us can hardly be said to have arrived at it through deliberation. You arrive at it through the simple recognition of a reason-that it is your own future-without deliberation. I am using the term in this looser sense; I do not think that most people arrive at their ambitions through deliberation.

²⁹ I am not suggesting that there is something perverted about sexual jealousy. The desire to make love to someone is not primarily the desire to be the one who provides him with a certain kind of experience. The desire to make someone happy can be an expression of either morality or of love, but in neither case is it their essence. For further discussion, see my “Creating the Kingdom of Ends: Reciprocity and Responsibility in Personal Relations,” in Philosophical Perspectives 6: Ethics, ed. Tomberlin, James (Atascadero, California: Ridgeview Publishing Company, 1992).Google Scholar

³⁰ One may say that human talents and powers are developed and refined by these activities, and that this is an objective human good. Indeed, when people talk about what they like about these activities, these are the things they talk about. But this does not mean that what they care about are these supposedly objectively valuable features of their chosen activities rather than the particular activities themselves. Other activities, which people are not always prepared to substitute for the ones they actually choose, may refine and develop similar human powers. And one may even accept these other activities as substitutes if it is necessary (as when one turns to a less strenuous sport in old age). But we should not take that to mean that the “objective” goods embodied in the activities were all that the actors cared about. The problem here is like the problem associated with the fact that we love particular people even though what we can say we love about them is general. You love a particular person, not just his warmth, intelligence, and sense of humor. It is not true that any other person with these attributes would do just as well, even though it is true that if he leaves you, you may seek another person with these attributes to replace him. No adequate theory of value can ignore these complex facts.

³¹ There are several ways to motivate this thought. Daniel Warren has pointed out to me, in conversation, that without this thought the requirement to share ends could be met by someone who took a sort of patronizing attitude towards the ambitions of others: “Oh, well, you like it, so I suppose we shall have to count it as good.” Scott Kim points out that a parallel problem exists on the recipient's side: If you accept help from someone who does not in any way enter into your ambitions, you may be regarding him somewhat instrumentally. The point of these remarks is not to show that there is something wrong with either helping or accepting help among those who do not really enter into each other's interests, but that the moral attitude required of us is less than perfectly realized in such cases. This in turn shows that there is a kind of continuum between the sense of “shared ends” defined in the previous paragraph and the sense defined in this paragraph. One may share the ends of others in the sense of (i) agreeing to promote them because they are another's ends; (ii) trusting that there must really be something interesting about them because they are another's ends; (iii) seeing what is interesting about them; and (iv) coming to have them as your own ends. I thank Thomas Scanlon for prompting me to be clearer about this point, and Amélie Rorty for reminding me of the importance of the possibility that one may stop at step (iii); e.g., one may come to have a much better appreciation of what a certain school of art was trying to do without actually coming to enjoy the works or find them beautiful.

³² See Mill, John Stuart, Utilitarianism (Indianapolis: Hackett Publishing, 1979), pp. 23–24.Google Scholar

³³ Ibid., p. 22.

³⁴ Williams, , in Smart, and Williams, , Utilitarianism: For and Against, p. 98.Google Scholar

³⁵ I think this point is sometimes overlooked in discussions of this example. Williams, to be fair, specifies that the Indians are begging Jim to accept the offer (ibid., p. 99). But he obscures the importance of this point when he says that this is “obviously” what they would be doing.

³⁶ This is a familiar move: when reminded that a person is likely to experience a negative moral emotion such as guilt, regret, hesitation, or squeamishness about doing something which according to our theory is right, the philosopher points out that the action in question is usually wrong and that it is therefore healthy to be equipped with some reactions which will make it hard for us to do it or will make us think twice before doing it. The assumption seems to be that our emotions are clunkier, more mechanical, less sensitive to the details of a situation, and altogether less refined than our thoughts. This view seems to be a byproduct of the modem conception of the emotions; the emotions are conceived as feelings or reactions, not as perceptions. Aristotle, for instance, would not have said this about the trained emotions of the virtuous person.

³⁷ Scheffler, , The Rejection of Consequentialism, pp. 14ff.Google Scholar

³⁸ Ibid., p. 82.

³⁹ See note 1.

⁴⁰ See Kant, Immanuel, Grounding for the Metaphysics of Morals (Indianapolis: Hackett Pub lishing, 1983), p. 37Google Scholar; Prussian Academy edition, p. 430. For interpretation, see my “The Right to Lie: Kant on Dealing with Evil,” Philosophy and Public Affairs, vol. 15 (1986), pp. 325–49Google Scholar; and O'Neill, Onora, “Between Consenting Adults,” Philosophy and Public Affairs, vol. 14 (1985), pp. 252–77.Google Scholar

⁴¹ There are familiar philosophical puzzles about all of these notions. This is perhaps especially true of coercion, notoriously hard to distinguish in any formal way from bribery or the mere offer of an incentive. This is not the place to take these puzzles up, but this should pose a problem only for readers who are actually skeptical about whether there is such a thing as coercion.

⁴² That is a remark that needs many qualifications. Actual consent-in the sense of saying yes-can easily be spurious. As Onora O'Neill argues, a better test of whether someone was able to consent is whether the person had an authentic opportunity to say no. See O'Neill, Onora, “Justice, Gender, and International Boundaries,” in The Quality of Life, ed. Nussbaum, Martha and Sen, Amartya (Oxford: Clarendon Press, 1992).Google Scholar

⁴³ Thomas Scanion has drawn my attention to a footnote in Nagel's paper “War and Massacre” in which Nagel mentions that Marshall Cohen says that according to Nagel's view, shooting at someone establishes an I-thou relationship. See Mortal Questions (Cambridge: Cambridge University Press, 1979), p. 69Google Scholar. I agree with Cohen, and think that so interpreted Nagel is right.

⁴⁴ Strictly speaking, this is only an account of what Kant would call the “incentive” of morality; we are not obligated until we acknowledge the necessity of adopting this incentive as law. A related point is this: several readers, among them Barbara Herman and Arthur Kuflik, have pointed out to me that this account says nothing about why I must recognize the other as a person, only about what follows from the fact that I do. For now, I can only acknowledge that the argument is incomplete in these ways. I hope to say more on these points elsewhere.

⁴⁵ A similar point, I think, can be made about Scheffler. He says that it is “natural” to in terpret Nozick's defense of side constraints as an appeal to the disvalue, the badness, of violating those constraints (Scheffler, , The Rejection of Consequentialism, p. 88)Google Scholar. But it is only “natural” if you ignore Nozick's reminder that a moral constraint does not have to function as a moral goal-that is, only if you presuppose that the business of morality is the realization of goals.

⁴⁶ Several readers have suggested to me that I am not really rejecting consequentialism but only proposing an alternative account of what we should aim at: decent human relationships. This suggestion is similar to the familiar consequentialist reply to standard counterexamples: “If justice matters, we can include it among the results.” That kind of inclusion results in the curious view discussed above: that we should commit injustice if it will bring about more justice. Scheffler imagines his consequentialist saying: “And if you are worried that a violation of R [the deontological requirement] corrupts the relationship between the agent and the victim, and that the corruption of a human relationship is a bad thing, then why isn't it at least as permissible to corrupt one valuable relationship if that is the only way to prevent the corruption of five equally valuable human relationships?” (Scheffler, , The Rejection of Consequentialism, pp. 89–90)Google Scholar. A commitment to mutual respect in human relationships is not merely a commitment to bringing respectful relationships about, any more than a commitment to justice is merely a commitment to bringing justice about. For example: In the early stages of our friendship, I might be tempted to conceal things from you in order to help bring about a condition of mutual trust; I might be afraid that you will reject me too quickly if you find certain things out before you know me better. But if mutual trust is ever to be achieved, the day must come when my calculations about the effects of my telling you things stops: that is what it means for me to trust you. The point here is that having decent relationships with people is not the same as bringing them about, and to some extent is in consistent with regarding them as things to be brought about. And my suggestion in this essay is that having decent human relationships, not bringing them about, is the primary concern of morality.

⁴⁷ Nor is it an accident that many of my own examples in this essay, especially the ones concerning Jim and the Indians, focus on what the protagonists might say to each other. Many of Rawls's arguments invite us to imagine people talking to each other, to consider what it would be like to say certain things to another person. His argument against the utilitarian account of what is wrong with slavery, in “Justice as Reciprocity” (in Utilitarianism with Critical Essays, ed. Gorovitz, Samuel [Indianapolis: Bobbs-Merrill, 1971], pp. 242–68)Google Scholar in effect invites us to consider the absurdity of a slaveholder who says to a protesting slave: “But my gains outweigh your losses!” His consideration of the effects of publicizing principles of justice on people's self-respect are also related to this theme (see Tiieory of Justice, pp. 177ff)Google Scholar. Part of the appeal of the difference principle is that it is the source of justifica tions which you can offer to anyone without embarrassment.