1 See Rawls, John A Theory of Justice (Cambridge: Harvard University Press 1971), 46–53Google Scholar; and Daniels, Norman Justice and Justification: Reflective Equilibrium in Theory and Practice (Cambridge: Cambridge University Press 1996)CrossRefGoogle Scholar. For a useful survey of recent literature on intuitionism, see McMahan, Jeff ‘Moral Intuition,’ LaFollette, Hugh ed., The Blackwell Guide to Ethical Theory (Oxford: Blackwell 2000)Google Scholar. By ‘intuitionism’ I mean the method of justification that came to be known by that name during the last twenty years or so, the method of which Rawls's notion of reflective equilibrium is a development; I am not, for instance, using the term ‘intuitionism’ here to refer to anti-consequentialism or Ross-style pluralism (which is, ironically, how Rawls uses the term in A Theory of Justice), or to restrict it to the very specific sort of program in moral epistemology that Sidgwick discusses in The Methods of Ethics (London: Macmillan 1907). Instead, I simply mean the general thesis that moral theories must be checked against our pre-theoretical moral intuitions, or at least some subset of them, and the idea that those intuitions form one of the key starting points for moral theorizing.

2 For nature to morality arguments that rest on claims about personal identity, see Parfit, Derek ‘Later Selves and Moral Principles,’ Montefiore, Alan ed., Ethics and Personal Relations (Montreal: MeGill-Queen's University Press 1973)Google Scholar, and part 3 of Reasons and Persons (Oxford: Oxford University Press 1984). For arguments that rest on claims about the separateness of persons, see Rawls, A Theory of Justice, 23-4Google Scholar. Compare Nozick, Robert Anarchy, State, and Utopia (New York: Basic Books 1974), 32–3Google Scholar. For arguments that draw on claims about our motivational capacities, see Williams, Bernard ‘Persons, Character, and Morality,’ Moral Luck [Cambridge: Cambridge University Press 1981)CrossRefGoogle Scholar) and ‘A Critique of Utilitarianism’ in Smart, J.J.C. and Williams, B. eds., Utilitarianism: For and Against (Cambridge: Cambridge University Press 1973)CrossRefGoogle Scholar and Scheffler, Samuel The Rejection of Consequentialism (Oxford: Oxford University Press 1982)Google Scholar. Proponents of some versions of feminist ethics also base ethical claims or ethical theories on facts about human biology or psychology. See, for example, Gilligan, Carol In a Different Voice (Cambridge: Harvard University Press 1982)Google Scholar; Noddings, Nel Caring (Berkeley: University of California Press 1984)Google Scholar; Held, Virginia Feminist Morality: Transforming Culture, Society, and Politics (Chicago: University of Chicago Press 1993)Google Scholar; Ruddick, Sara Maternal Thinking (Boston: Beacon Press 1989)Google Scholar. See also Flanagan, Owen The Varieties of Moral Personality (Cambridge: Harvard University Press 1991), 196–252.Google Scholar

3 Some discussion appears in Samuel Scheffler, The Rejection of Consequentialism; Kagan, Shelly The Limits of Morality (Oxford: Oxford University Press 1989)Google Scholar and Normative Ethics (Boulder, CO: Westview 1998), 280-94; Owen Flanagan, The Varieties of Moral Personality; Conee, Earl ‘On Seeking a Rationale,’ Philosophy and Phenomenological Research 45 (1985)CrossRefGoogle Scholar; Montague, Phillip ‘The “Negative” and “Positive” Arguments of Moral Moderates,’ Pacific Philosophical Quarterly 77 (1996) 37–44CrossRefGoogle Scholar; McMahon, Christopher ‘Expression Arguments in Ethics,’ Pacific Philosophical Quarterly 69 (1988) 325–41CrossRefGoogle Scholar; John Rawls, ‘The Independence of Moral Theory’ (Presidential Address, American Philosophical Association 1974); Daniels, Norman ‘Moral Theory and the Plasticity of Persons,’ Monist 62 (1979)CrossRefGoogle Scholar; and Stem, Robert ‘The Relation between Moral Theory and Metaphysics,’ Proceedings of the Aristotelian Society 92 (1992) 143–59Google Scholar.

4 Or, as Rawls succinctly puts it, that ‘the correct regulative principle for anything depends on the nature of that thing’ (A Theory of Justice, 29).

5 See Bernard Williams, ‘Persons, Character, and Morality’ and ‘A Critique of Utilitarianism,’ and Samuel Scheffler, The Rejection of Consequentialism. A ‘personal point of view’ is a ranking of states of affairs, connected with a specific person, which (typically) differs from an impersonal, objective, or impartial ranking.

6 See Rawls, A Theory of Justice, 23-4Google Scholar; Nozick; and McKerlie, Dennis ‘Egalitarianism and the Separateness of Persons,’ Canadian Journal of Philosophy 18 (1988) 205–26.CrossRefGoogle Scholar

7 See Daniels, Norman ‘Wide Reflective Equilibrium and Theory Acceptance in Ethics,’ Journal of Philosophy 76 (1979) 256–82)CrossRefGoogle Scholar. A moral theory that bears a suitable linking relation to the nature-claims made by our best theories of human nature would, ipso facto, cohere with at least one background theory that seems relevant to morality.

8 To take just to prominent examples, Michael Sandel can be plausibly read as making this kind of criticism against the contractarian liberalism of Rawls in Liberalism and the Limits of Justice (New York: Cambridge University Press 1982), esp. 14; and Virginia Held can be read as making this kind of criticism against contractarian (and other) kinds of moral theories in Feminist Morality: Transforming Culture, Society, and Politics (see esp. 57-63 and 193'-214). For a fuller discussion of some classic communitarian literature that reconstructs its criticisms of contractarianism in this way, see Kymlicka, Will Contemporary Political Philosophy (Oxford: Oxford University Press 1990), 207–15Google Scholar. For a fuller discussion of some classic feminist literature that reconstructs criticisms of contractarian (and other) approaches to morality in this way, see Tong, Rosemarie Feminine and Feminist Ethics (Belmont, CA: Wadsworth 1993), 49–63Google Scholar.

9 See Rawls, ‘Justice as Fairness: Political Not Metaphysical,’ Philosophy and Public Affairs 14 (1985) 223–51Google Scholar; and Gauthier, David Morals by Agreement (Oxford: Oxford University Press 1986), 319–39, esp. 338Google Scholar. Such replies may involve showing that while the theory in question does make use of some model of human nature and/ or the nature of persons, that model is not meant to embody any claim on the part of the theory about what persons are really like. Such claims will be most compelling when they are accompanied by an alternative explanation of the work that the model does within the theory. Thus Gauthier's remarks about the role of homo economicus in his theory suggest that the homo economicus model of human nature is meant as a counterfactual that is necessary to show that moral constraints do not derive their rationality from social motivations, since they would be rational even if (contrary to fact), persons were asocial.

10 Of course it may be possible to produce a compelling analysis of the moral theory under attack that makes it difficult to deny the link between it and some false nature-claim. However, such arguments are likely to face an uphill battle, since they will in effect be analyzing the ‘intent’ of someone else's theory. That is, they will have to claim, more or less, that a theory is ‘intended’ to capture or reflect some fact about human nature or the nature of persons. In general, it will be possible for the proponent of the theory to give an alternate account of the true intent of the theory, and in general when two people disagree about the intent of a theory, it is probably best to believe the person who holds that theory.

11 I shall use the terms ‘deontologist’ and ‘consequentialist’ so that they correspond more or less with Kagan's moderate and extremist. The deontologist, as I am thinking of her, denies that an agent is always morally required to (as Kagan puts it) ‘make her greatest possible contribution to the [impartially specified] greatest good.’ She may base this denial on an acceptance of agent-centered options and constraints, or on a skepticism about the existence of a single, rationally privileged impartial evaluative point of view that ranks states of affairs in such a way as to give all agents a moral reason to bring about the state of affairs that it ranks as optimal.

12 It turns out, however, that it would not be plausible to make the conditional claim about logical consistency. For making the conditional claim about the relation of consistency would imply that a moral theory would become less plausible if it was consistent with any false nature-claims. This, implication, however, is absurd. Utilitarianism, for example, is not made less plausible simply because it is logically consistent with the false nature-claim that human beings are feathered tripeds. So we cannot infer the conditional claim from the unconditional claim: the fact that it would be plausible to make the unconditional claim about a given linking relation does not entail that it would also be plausible to make the conditional claim about that same linking relation. I thank an editor of The Canadian Journal of Philosophy for helping me to see this complication (and avoid a logical glitch in an earlier way of formulating the ideas in this section of the paper).

13 See Quine, W.V.O. ‘Existence and Quantification’ in Ontological Relativity and Other Essays (New York: Columbia University Press 1969), 94ff.CrossRefGoogle Scholar

14 Versions of this argument appear both in Reasons and Persons and in ‘Later Selves and Moral Principles.’ The actual argument is somewhat more subtle than the version given here, but this rough characterization should be sufficient to make the point.

15 Yet another example of this argument was suggested to me by an editor for Canadian Journal of Philosophy. Virtue theory arguably assumes that there are relatively fixed character traits that can be the fundamental subjects of moral assessment. However, there is apparently a fair amount of psychological data that calls into question whether we do have anything like the kind of stable character traits that could qualify as virtues and vices in the way that virtue theory normally assumes. Thus it may be that the psychological facts are logically inconsistent with the key tenets of (at least some forms of) virtue theory.

16 Flanagan, Owen Varieties of Moral Personality, 27Google Scholar; see also 15-55.

17 And one might reasonably distinguish between whether a theory is self-defeating and whether it is true. One might hold that if morality turns out to be so demanding as to be practically self-defeating, then that would be a reason to think that morality should only require us to promulgate a moral code that is as demanding as possible without causing people to give up on morality. Many people hold that an indirect or esoteric approach to morality like this would violate a so-called ‘publicity requirement’ that rules out any moral theory that argues against publicizing its own actual demands. However, it seems fair to say that arguments in favor of such publicity requirements have been inconclusive at best.

18 See Wolf, Susan ‘Moral Saints,’ The Journal of Philosophy 79 (1982) 419–39CrossRefGoogle Scholar, and Bernard Williams. For the record, Williams also suggests that Kant's ethics is similarly alienating. I think that he is on somewhat less firm ground with this claim, for the jury is still out on the demandingness of Kant's ethics. A good recent discussion of that issue is Baron's, Marcia Kantian Ethics (almost) without Apology (Ithaca: Cornell University Press 1995)Google Scholar.

19 Kagan, The Limits of Morality, 271-8; 280–330Google Scholar, esp. 324.

20 Ibid., 271

21 Ibid, 332. Rawls makes a similar point, noting that ‘moral conceptions regard persons differently and prize different aspects of their nature’ (‘The Independence of Moral Theory,’ 17).

22 I take my remarks here merely to summarize Kagan's remarks on pages 258-79 of The Limits of Morality, and to translate his point into my own terms.

23 See Taurek, John ‘Should the Numbers Count?’ Philosophy and Public Affairs 6 (1977) 293–316Google ScholarPubMed; and Kamm, F.M. Morality, Mortality, vol. 1 (New York: Oxford University Press 1993)Google Scholar.

24 Scheffler, 58-60; Scheffler uses different words, but I trust that I have presented the spirit of his idea. Cf. Kagan, The Limits of Morality, 345Google Scholar.

25 At least in ‘normal’ circumstances. As an anonymous referee pointed out to me, a moral theory might call for us to give up what it sees as a good characteristic in very exceptional situations. I add the qualifications ‘permanently and completely’ to deal with similar worries about moral theories that might sometimes ask us to temporarily subordinate one valuable aspect of our nature to some other value. Finally, it is important to note that a theory may be directly inaccessible by explicitly requiring us to change facts about ourselves, or it can be indirectly inaccessible by imposing requirements that can only be met if we change such facts.

26 Of course more would need to be said about what counts as ‘giving up’ a feature. Presumably the advocate of a partiality-respecting morality would want to allow that morality could, in some situations, require an agent to overcome or put aside her natural partiality. What she would want to rule out is a morality that made a permanent and pervasive impartiality a condition for being moral. She might also have to address indirect consequentialist arguments to the effect that some partiality may be an efficient means to promoting the impartial good. She may need to explain why retaining whatever degree of partiality is justified by these indirect arguments does not satisfy the condition that morality not require the giving up of one's natural partiality.

27 A Theory of Justice 186

28 Ibid., 189

29 However, there may be competing ways to model a given nature-fact. In The Possillility of Altruism (Oxford: Oxford University Press 1970), Thomas Nagel suggests an initial choice situation in which the chooser will ‘branch’ into each separate person who will eventually live in the society. He claims that a chooser in such a situation would endorse utilitarianism. Nagel's suggestion explicitly recognizes the separateness of persons. But unlike the chooser in Rawls's Original Position, Nagel's chooser does not become one of the persons but all of them. Hare's argument for utilitarianism based on the (supposed) requirement to identify with each person would seem to be a development of this or a similar idea. See R.M. Hare, Freedom and Reason (Oxford: Oxford University Press 1963), esp. chs. 5 and 6. While Nagel's proposed initial choice situation does ostensibly recognize the separateness of persons, it is unclear that a ‘branching’ entity of this sort really is a good approximation of separate individuals. The branches of such an entity are less separate than actual persons. In Equality and Partiality (Oxford: Oxford University Press 1991) Nagel seems to be suggesting that some version of egalitarianism unites an impartial moral concern with a respect for the separateness of persons.

30 The most serious worry concerns the maximin decision rule which plays a crucial role in Rawls's derivation of a deontological system from the Original Position. Cohen's, Joshua ‘Democratic Equality,’ Ethics 99 (1989) 727–51CrossRefGoogle Scholar) is a nice discussion of these issues with references to the relevant literature.

31 It is important to note that what I am calling the Rawlsian Relation is meant as a way to get a moral theory to embrace claims about human nature or the nature of persons, rather than a method for making a moral theory cohere with our considered moral intuitions about right and wrong. Although Rawls's own methodology emphasizes the task of getting moral theory to cohere with moral intuitions, it apparently does not rule out making it cohere with nature-facts. Indeed, Rawls seems to suggest (sec. 30 of A Theory of Justice) that we can rule out the initial choice situation that leads to utilitarianism on the basis of objective facts about the separateness of persons. So it is possible to see the nature to morality methodology as an extension or elaboration of Rawls's own methodology. On the other hand, Rawls seems to disavow the nature to morality methodology in ‘The Independence of Moral Theory.’ I cannot hope to resolve this tension in Rawls's thought here, but I refer the interested reader to the superb discussion of this issue in Robert Stem, 146-51. As mentioned before, the nature to morality methodology is independent of any claim about how much stock we should put in a moral theory's ability to cohere with our moral intuitions.

32 In order to avoid repeating various qualifications, I will restrict myself here to that part of morality that concerns our treatment of persons. I do not mean to suggest that morality is only concerned with persons. In fact, the present line of thought is quite compatible with the idea that morality may be concerned with many other kinds of beings, and that its function is to show us the appropriate way to manifest our concern for each kind of being that has moral status.

33 This is not to say that the further specification of what a person is fully determines what the moral theory will say; the moral theory will typically have to rely on other claims as well.

34 Haworth, Lawrence ‘Autonomy and Utility,’ Ethics 95 (1984)CrossRefGoogle Scholar, and Will Kymlicka (ch. 2) develop explanations of utilitarianism along these lines.

35 There seem to be at least two different ways that Kantians try to move from this picture of persons to prohibitions against sacrifices. The first is to argue that because each person is an individual moral subject, she does not partake in a larger collective moral subjectivity. For that reason, benefits to other persons do not compensate her for the sacrifices she must make to achieve them. The second is to argue that the status of persons as autonomous moral subjects gives them a status that makes them unlike anything else in the world. And whereas most other things in the world are available for use as tools because they are mere objects, persons, because of their various properties (rationality, consciousness, autonomy, etc.) are not objects at all, but rather subjects. And while it is proper to treat objects as tools, it is not proper to treat subjects as tools, for they are of a different order of being, as it were.

36 Not all utilitarians see their theory in these terms, however. I discuss the relation of utilitarianism to conceptions of what is valuable about persons in ‘On the Cross of Mere Utility: Utilitarianism, Sacrifices, and the Value of Persons,’ Utilitas 12 (2000) 1-24. In that paper, I note that some utilitarians may accept what may be thought of as a somewhat more Kantian conception of persons. Part of my argument in that paper is that if a utilitarian does accept a more Kantian picture of persons (and in particular, the relation of persons to value), then utilitarianism may not turn out to have all of the (generally counter-intuitive) implications that have often been attributed to it. I suspect that the Kymlicka/Haworth construal of utilitarianism would fit this model.

37 I thank an editor for Canadian Journal of Philosophy for this point.

38 One could draw on work of Kant and Alan Gewirth for arguments along these lines. Gewirth's arguments (which are embedded in a very different overall method of justification from the one I am discussing here) can be found in Reason and Morality (Chicago: University of Chicago Press 1978).

39 An argument of roughly this sort appears in Lippke, Richard ‘Why Persons Are the Ground of Rights (and Utility Isn't),’ Journal of Value Inquiry 18 (1984) 207–17CrossRefGoogle Scholar, at209ff.

40 I suspect that arguments relying. on premises about fundamental features of persons will be more convincing than those relying only on premises about features that are fundamental to human beings, but which could easily be imagined to be absent in non-human persons, but I won't pursue this issue further here.

41 I am assuming that the concept of personhood is roughly equivalent to (or at least includes) that of moral agency. Readers who distinguish sharply between persons and moral agents should take my remarks about persons to refer to moral agents.

42 The consequentialist could claim that separateness is neither a positive fact nor a metaphysically or conceptually fundamental fact about persons. Taking this line would not necessarily involve rejecting the claim that morality should embrace the most significant facts about persons. Instead, it would simply deny that separateness is one of the most significant facts about persons.

43 I thank an editor for the Canadian Journal of Philosophy for pressing this point.

44 Indeed, sometimes consequentialists will be pushed to this view of the underlying rationale for consequentialism when they are faced with arguments that consequentialism neglects persons in favor of utility or value, or that it treats persons as mere containers for value. When faced with such arguments, consequentialists often claim that this mistakes the consequentialist conception of value—that utility is not some thing over and above persons, but rather that we care about utility because we care about persons. I discuss this further in ‘On the Cross of Mere Utility.’

45 Notice that the major competing method of moral inquiry actually requites much more in terms of our taking our traditions for granted. The method of reflective equilibrium requires that the content of a theory not depart too radically from the content our ordinary moral thinking. The nature to morality methodology, on the other hand, requires only continuity with a conception of the nature and point of morality, i.e., the conception of morality as being for and (at least mainly) about persons.

46 I am grateful to anonymous reviewers and an editor for Canadian Journal of Philosophy for extremely generous and helpful suggestions. I am also grateful to Bob Stecker for several helpful comments and Shelly Kagan for some early words of encouragement on this project. Much of my thinking about this topic took shape while I was teaching the work of Scheffler and Kagan in classes at Simon Fraser University in the fall of 1997 and at Central Michigan University in the fall of 2000. I thank both of these truly wonderful groups of students, and especially Tim Christie and Lou Blouin, for conversations that helped me work out some of the ideas contained here.