¹ Kant, Immanuel, Critique of Pure Reason, trans. Smith, Norman Kemp (New York: St. Martin's Press, 1965), p. 593 (A737-38/B767-68).Google Scholar

² Kant, Immanuel, Critique of Judgment, trans. Pluhar, W. S. (Indianapolis: Hackett Publishing Co., 1987), section 8.Google Scholar

³ Michelman, Frank, “Justification (and Justifiability) of Law in a Contradictory World,” in NOMOS XXVIII: Justification, ed. Pennock, J. Roland and Chapman, John W. (New York: New York University Press, 1986), p. 73.Google Scholar

⁴ The most prominent representative of this view, of course, is Jurgen Habermas, who defends a “discourse theory of ethics” which takes public justification to be the core organizing idea of morality and politics; see Habermas, , Moral Consciousness and Communicative Action, trans. Lenhardt, C. and Nicholsen, S. W. (Cambridge, MA: MIT Press, 1991)Google Scholar, and Justification and Application, trans. Cronin, Ciaran (Cambridge, MA: MIT Press, 1993).Google Scholar Samuel Freeman recently argued that Thomas Scanlon's contractarian theory of morals is best read in this way; see Freeman, , “Contractualism, Moral Motivation, and Practical Reason,” Journal of Philosophy, vol. 88 (1991), pp. 281–303.CrossRefGoogle Scholar Christine Korsgaard claims that John Rawls's theory of justice is also committed to this view; see Korsgaard, , “The Reasons We Can Share,” Social Philosophy & Policy, vol. 10, no. 1 (1993), pp. 24–51.Google Scholar

⁵ See my essay “Public Practical Reason: Political Practice,” in NOMOS XXXVII; Theory and Practice, ed. Shapiro, Ian and De Cew, Judith (New York: Oxford University Press, 1994).Google Scholar

⁶ Wittgenstein, Ludwig, Philosophical Investigations (New York: Macmillan, 1953), paragraphs 138–242.Google Scholar

⁷ The argument I shall outline closely follows Pettit, Philip's recent work, The Common Mind (New York: Oxford University Press, 1993)Google Scholar, chs. 2 and 4, and is defended at length there. Subsequent references to Pettit's book will be given parenthetically in the text.

⁸ Any “nonbasic rule” will be dependent on more-basic rules which themselves are directly readable or are dependent on yet-more-basic rules which are directly readable.

⁹ Pettit, , The Common Mind, pp. 86–90.Google Scholar It is important to note, especially when we proceed to the next stage of Pettit's explanation, that the extrapolative inclination typically will be entirely transparent to the subject (ibid., pp. 89–90,103). That is, when faced with a series of examples, the subject will not self-consciously call in this inclination for consultation and upon its advice target a certain rule among the indefinite number of candidates; nor does the subject self-consciously consult the inclination again when he or she tries to apply the targeted rule. Rather, presented with the examples, this rule will present itself to the subject as the rule exemplified in them, and this will present itself as the correct application of the rule.

¹⁰ Anticipating the outcome of the argument I will offer presently for the necessarily interpersonal nature of this interaction, I will formulate these observations in terms of interpersonal interaction.

¹¹ Jay F. Rosenberg's discussion of the rule-following problem in One World and Our Knowledge of It (Dordrecht: Reidel, 1980)Google Scholar, chs. 4 and 7, brought this to my attention. See esp. ibid., pp. 158–59.

¹² Kant, , Critique of Pure Reason, p. 130 (A97). The influence of Kant's argument for the transcendental unity of apperception—that is, for the unity of experience as it bears on theoretical rationality—on the line of argument in the text is evident, but the argument in the text is not, strictly speaking, Kant's argument. I will not pause here to detail the differences.Google Scholar

¹³ See, e.g., Blackburn, Simon, “The Individual Strikes Back,” Synthese, vol. 58 (1984), pp. 281–301.CrossRefGoogle Scholar

¹⁴ I am not sure whether a solipsist could intelligibly be said to have a gender—surely, not one the solipsist could recognize. But to avoid the awkwardness of referring to our solipsist as “it,” I have opted for the masculine pronoun. At least I can expect that he will not object.

¹⁵ See, e.g., Williams, Bernard, Ethics and the Limits of Philosophy (Cambridge, MA: Harvard University Press, 1985), pp. 68–69.Google Scholar

¹⁶ The epigraph is drawn from Hume, David, Enquiry concerning the Principles of Morals, 3d ed., ed. Nidditch, P. H. (Oxford: Clarendon Press, 1975), p. 272Google Scholar (hereafter Enquiry, followed by page number).

¹⁷ Hume, David, A Treatise of Human Nature, 2d ed., ed. Nidditch, P. H. (Oxford: Clarendon Press, 1978), p. 582Google Scholarxs (hereafter Treatise).

¹⁸ This thought is also at the heart of Kantian reflections on judgment. See, for example, Kant, , Critique of JudgmentGoogle Scholar, section 8; see also Habermas, , Moral Consciousness and Communicative Action, pp. 48–49, 58–59.Google Scholar

¹⁹ Hegel, G. W. F., Philosophy of Right, trans. Knox, T. M. (Oxford: Oxford University Press, 1952), p. 258.CrossRefGoogle Scholar The argument of this paragraph roughly follows Hegel's critique of intuitionist moral theories. See Hegel, G. W. F., The Phenomenology of Spirit, trans. Miller, A. V. (Oxford: Clarendon Press, 1977), paragraphs 632–56, esp. 639–40, 643, 648.Google Scholar

²⁰ For a discussion of the distinction between “positional” and “nonpositional” objectivity, see Sen, Amartya K., “Objectivity and Position,” Lindley Lecture, University of Kansas, 1992, pp. 1–6Google Scholar; and Sen, , “Positional Objectivity,” Philosophy and Public Affairs, vol. 22 (1993), pp. 126–28.Google Scholar

²¹ Nagel, Thomas, The View from Nowhere (New York: Oxford University Press, 1986), chs. 1, 8, and 9.Google Scholar

²² My discussion draws on Hume, 's EnquiryGoogle Scholar, his Treatise, and his essay “Of the Standard of Taste,” in Hume, , Essays: Moral, Political and Literary, ed. Miller, E. F. (Indianapolis: Liberty Classics, 1985), pp. 226–49.Google Scholar No doubt my interpretation is controversial; however, this is not the occasion to defend it.

²³ This, I take it, is David Gauthier's proposal, following Hobbes. See his “Public Reason,” elsewhere in this volume.

²⁴ On this point, see Scanlon, Thomas, “Contractualism and Utilitarianism,” in Sen, Amartya K. and Williams, Bernard, eds., Utilitarianism and Beyond (Cambridge: Cambridge University Press, 1982), p. 117.Google Scholar

²⁵ Baier, Annette, Postures of the Mind (Minneapolis: University of Minnesota Press, 1985), p. 170.Google Scholar

²⁶ Hume, , “Standard of Taste,” p. 239Google Scholar; see also Kant, , Critique of JudgmentGoogle Scholar, section 40.

²⁷ Thus, on Hume's view, moral judgments and the point of view from which they are made depend in a way on “artifice.” (On this sense of “artifice,” see Treatise, p. 490.)Google Scholar For Hume, the same fundamental human capacities that make possible the emergence in human society of the conventions of justice, promising, contracts, and political authority are also at the root of the moral point of view. These capacities include reflective intelligence, the ability to communicate feelings, attitudes, and outlooks, the ability to understand the feelings, attitudes, and outlooks of others (Hume calls this “sympathy”), and the capacity to develop a “common sense of interest” over time through interaction and communication. (See Treatise, pp. 484–98.)Google Scholar

²⁸ Habermas also believes reflection from the moral point of view will yield highly general and abstract rules, but his reason is different. He believes that the moral point of view must abstract from all the particular circumstances of persons and community, and argues that norms can withstand dialogic assessment from the moral point of view as he conceives it only in a highly decontextualized form. See Habermas, , Justification and Application, p. 13.Google Scholar

²⁹ Kant, , Critique of JudgmentGoogle Scholar, section 8. See also Korsgaard, , “The Reasons We Can Share”Google Scholar (supra note 4); and Habermas, , Justification and Application, pp. 8, 24, 50, 52Google Scholar, and Moral Consciousness and Communicative Action, p. 65.Google Scholar Habermas writes: “Arguments by their very nature point beyond particular individual life worlds; in their pragmatic presuppositions, the normative content of presuppositions of communicative action is generalized, abstracted, enlarged, and extended to an ideal community encompassing all subjects capable of speech and action…” (Justification and Application, p. 50).Google Scholar

³⁰ This raises a difficult question to which I do not now have a fully satisfactory answer, namely, how are we to distinguish between a defeated hope of commonality and a yet unfulfilled one? If failure to secure wide agreement on particular norms or judgments were sufficient to warrant abandoning the hope, the moral point of view would be an impossible dream in any modern society. So we must resist the temptation to take persistent dissensus as a sufficient ground for regarding the hope as defeated. But it is fair to ask why must we resist, and for how long. Could it be that part of the vocation of morality is that we are required never to abandon attempts at moral discourse with even the most recalcitrant party? Or is this hopelessly unrealistic? If it is part of our responsibility as moral agents never to regard moral discourse as futile, then we may have to accord universality, like consensus, the role of a regulative ideal in our understanding of the moral point of view.

³¹ This is a central feature of Rawls's theory of justice. See Rawls, John, A Theory of Justice (Cambridge, MA: Harvard University Press, 1971)Google Scholar, section 87; “Justice as Fairness: Political Not Metaphysical,” Philosophy and Public Affairs, vol. 14 (1985), p. 229Google Scholar; and “The Idea of an Overlapping Consensus,” Oxford Journal of Legal Studies, vol. 7 (1987), p. 6.Google Scholar

³² A fuller sketch is given in my “Public Practical Reason: Political Practice,” section II.

³³ I borrow the term from Oakeshott, Michael, Of Human Conduct (Oxford: Clarendon Press, 1975), pp. 35–36Google Scholar, but I give it a quite different sense.

³⁴ Korsgaard makes this point in “The Reasons We Can Share,” pp. 25–26.Google Scholar

³⁵ Joseph Raz also reminds us that the agent-relative/neutral distinction must not be confused with the distinction between valued outcomes or states of affairs, and actions valued in themselves (regardless of their outcomes). See Raz, , The Morality of Freedom (Oxford: Clarendon Press, 1986), pp. 281ff.Google Scholar

³⁶ Korsgaard's critique of Nagel's agent-centered account of deontological constraints in “The Reasons We Can Share” (pp. 41–49, esp. 47–48)Google Scholar suggested this argument to me. However, because I do not have the space here to do justice to Korsgaard's (or Nagel's) discussion, I will not attribute the argument to her. I do believe, however, that Korsgaard slips back and forth between the two versions of the agent-relative/neutral distinction I have identified; but then, so does Nagel.

³⁷ Suppose that you are not particularly religious but that you have a devout neighbor. Suppose she violates an important religious duty she acknowledged, causing you some inconvenience, but no more than what you regard as part of the normal price of social interaction. Not sharing the perspective from which her action is regarded as wrongdoing, you could not honestly regard yourself as a victim of her action, even though you have suffered a loss.

³⁸ A parallel unclarity faces the substantive version of the agent-relative/neutral distinction.

³⁹ I hasten to add that this is not to say that they are reasons of some collective entity.

⁴⁰ We can also talk of “sharing” of agent-centered (as well as action/outcome-centered) reasons. For example, the reason a mother has to secure the education of her child may also be shared with her husband. Thus, reasons can be joint or public in their content as well as their practical force. I shall not discuss this dimension of sharing any further.

⁴¹ Korsgaard, , “The Reasons We Can Share,” pp. 40–41.Google Scholar Korsgaard speaks of sharing ends rather than sharing reasons, but the difference for our purposes is not material; for ends provide us with reasons for action, and if ends are shared so too will be the reasons they supply us.

⁴² However, the reason for each may not be a reason for any particular action. For whether each has a concrete reason may depend on whether there is a cooperative scheme in place to achieve the correlative collective goal. Where a scheme is in place, one will have a reason to do one's part in it. Where there is no such scheme in place, one's reason will be abstract and contingent upon some cooperative scheme actually being put in place. I am grateful to David Schmidtz for this clarification.

⁴³ Hobbes, Thomas, Leviathan, ed. Macpherson, C. B. (Baltimore: Penguin Books, 1968), p. 120.Google Scholar

⁴⁴ Peace, unlike a beach home, or first prize in a slam-dunk competition, is a public good in the standard economists' sense of that term. So the idea of peace-for-me-(alone) is not an intelligible idea.

⁴⁵ Bentham, Jeremy, Introduction to the Principles of Morals and Legislation, ed. Burns, J. H. and Hart, H. L. A. (London: The Athlone Press, 1970), p. 28.Google Scholar

⁴⁶ Cavell, Stanley, The Claim of Reason (New York: Oxford University Press, 1979), p. 312.Google Scholar

⁴⁷ The best example is Gauthier, David's Morals by Agreement (Oxford: Clarendon Press, 1986)Google Scholar, and his “Constituting Democracy,” Lindley Lecture, University of Kansas, 1989. Scanlon's contractarian theory may be more robust; see Scanlon, , “Contractualism and Utilitarianism” (supra note 24), pp. 103–28.Google Scholar In “The Significance of Choice,” in The Tanner lectures on Human Values, VIII, ed. McMurrin, S. (Salt Lake City: University of Utah Press, 1988), p. 166Google Scholar, Scanlon describes morality as “a system of co-deliberation.”

⁴⁸ Michael Walzer discusses Thucydides' account of this dialogue in Walzer, , Just and Unjust Wars (New York: Basic Books, 1977), pp. 5–13.Google Scholar

⁴⁹ Gauthier, David, “Constituting Democracy,” esp. pp. 10–11Google Scholar; see, more generally, his Morals by Agreement, ch. 5.

⁵⁰ Strawson, Peter, “Freedom and Resentment,” in Fischer, John Martin and Ravizza, Mark, eds., Perspectives on Moral Responsibility (Ithaca, NY: Cornell University Press, 1993), p. 52.Google Scholar

⁵¹ Rawls distinguishes between a community's adopting a political conception of justice (or social and political constitution) as a modus vivendi and its adopting it as a moral conception. See, e.g., Rawls, , “The Idea of an Overlapping Consensus,” pp. 1, 3, 11.Google Scholar Whether Rawls intended this or not, I think the difference consists in the fact that, if members of a community accept the conception as a modus vivendi, they, more or less for their oivn part, come to see that it is useful (perhaps prudent) to comply with the constitution, while if they accept it as a moral conception, they mutually acknowledge that the constitution is the right or proper one (perhaps under the circumstances and for them).

⁵² Rawls, John, “The Domain of the Political and Overlapping Consensus,” New York University law Review, vol. 64 (1989), p. 237.Google Scholar

⁵³ Berlin's explanation suggests a reason for us to regard dissensus as not entirely unwelcome. I discuss a further reason for welcoming dissensus in “Public Practical Reason: Political Practice,” section II, A.

⁵⁴ I use this term in the sense Kant introduces to describe the principle of the indefinite extendability of any empirical synthesis; see Kant, , Critique of Pure Reason, pp. 449–50 (A508-9/B536-37).Google Scholar

⁵⁵ Postema, , “Public Practical Reason: Political Practice,” section II, B.Google Scholar

⁵⁶ Ibid., section III, B.

⁵⁷ For his most recent discussions, see Habermas, , Moral Consciousness and Communicative Action (supra note 4), pp. 78–98Google Scholar, and Justification and Application, pp. 31, 76–88.Google Scholar See also Alexy, Robert, A Theory of Legal Argumentation, trans. Adler, R. and MacCormick, N. (Oxford: Clarendon Press, 1989), pp. 119–24.Google Scholar

⁵⁸ Scanlon, , “Contractualism and Utilitarianism,” pp. 116–17.Google Scholar

⁵⁹ Joseph Raz pointed this out to me.

⁶⁰ I leave aside the question of how Scanlon understood his observation.

⁶¹ Robin Dillon helped me see this alternative reading of the recognition thesis.

⁶² To my knowledge, the richest philosophical discussion of the variety of self-defeating attempts to tear recognition from the hands of an unresponsive world is Hegel's discussion of the “dialectic of recognition” in Phenomenology of Spirit, ch. 4.Google Scholar

⁶³ Dent, N. J. H., Rousseau (Oxford: Basil Blackwell, 1988), p. 23.Google Scholar

⁶⁴ See Rousseau, Jean-Jacques, On the Social Contract and Discourses, trans. Cress, D. A. (Indianapolis: Hackett, 1983), pp. 140–42.Google Scholar Hume may also have had such beings in mind in his discussion of “rational but inferior beings” to whom We cannot owe duties of justice (Enquiry, p. 190).Google Scholar

⁶⁵ Gaus, Gerald F. calls them “valuer-agents” in his Value and Justification: The Foundations of liberal Theory (Cambridge: Cambridge University Press, 1990), pp. 278–80.CrossRefGoogle Scholar

⁶⁶ Note that for someone to acknowledge that, from my perspective, I have reason to act in a certain way is not for that person to acknowledge the validity of that perspective or the reasons that it allegedly generates for me.

⁶⁷ The famous “stag-hunt game,” a version of the game theorists' assurance game, has its roots in Rousseau's discussion of this pre-moral, strategically rational being; see Discourse on the Origin of Inequality, pp. 141–42.Google Scholar

⁶⁸ Frankfurt, Harry, “Freedom of the Will and the Concept of a Person,” Journal of Philosophy, vol. 68 (1971), pp. 5–20, esp. section II.CrossRefGoogle Scholar

⁶⁹ See Raz, , Wie Morality of Freedom, pp. 288–313.Google Scholar

⁷⁰ This discussion of forms of recognition, of course, is not exhaustive. For example, we can be present in the lives of others as friends, lovers, parents, colleagues, or fellow citizens. Indeed, there are many other kinds of social relations which manifest richly divergent modes of mutual recognition, none of which fit neatly into the above four categories. I do believe, however, that mutual recognition as moral selves will prove to be a critical component of any of these other forms, and thus will prove to be presupposed by them. At the same time, that category promises, I suspect, to have the widest (if not entirely universal) scope of all the remaining categories that incorporate it. However, I do not regard these important assumptions as uncontroversial. I can only plead lack of space to defend them adequately.

⁷¹ Note again that I have in mind only the meta-ethical dimension of the distinction between agent-neutral and agent-relative reasons. I intend my argument to leave entirely open the question of the extent to which reasons in their substantive content are agent-relative or agent-neutral (that is, agent-centered or action/outcome-centered). Thus, so far as my argument is concerned, public reasons may in fact be agent-centered (or action/outcome-centered) in their substantive content. I do not attempt to trace out any connections that may obtain between the requirement of publicity of reasons as a meta-ethical concern and the various forms of agent-relative or agent-neutral reasons as a substantive concern.