1 Gauthier, David, Morals by Agreement (Oxford: Oxford University Press, 1986). Hereafter MA.)Google Scholar

2 MA, p.78. Gauthier's remark echoes that of Rawls, John: “The theory of justice is a part, perhaps the most significant part, of the theory of rational choice.” A Theory of Justice (Cambridge, MA: Harvard University Press, 1971), p.16.Google Scholar While Gauthier's theory concerns the internal constraints that rational individuals need to accept in order to bring about mutually beneficial outcomes, Thomas Hobbes's classic theory tries to make do with external constraints alone. Buchanan's, JamesLimits of Liberty (Chicago: University of Chicago Press, 1975)Google Scholar tries to reach less drastic conclusions than Hobbes without requiring the internal constraints that Gauthier argues are necessary.

3 A Prisoner's Dilemma is a type of situation where the outcome of individually rational choice is inefficient in the weak Pareto sense that another outcome would be better for each individual. Virtually any text in decision or game theory will discuss this dilemma.

4 Gauthier exaggerates the “Lockean” nature of these rights. I have argued elsewhere that rights such as the ones he defends are best understood as “semi-natural,” a notion I define along with that of a natural right. See my “Natural Rights and Public Goods,” Attig, Thomas, Callen, Donald, & Gray, John, eds., The Restraint of Liberty, Bowling Green Studies in Applied Philosophy VII (Bowling Green, OH: Bowling Green State University, 1985), pp.102–117.Google Scholar

5 I also believe that (3), defended mainly in Chapter VIII of MA, should be distinguished more clearly from (1) and (2) than is actually done by Gauthier. The actual presentation of morals by agreement may lead many readers to confuse these very different stages of the argument.

6 I have been convinced of this in part by Peter Danielson. It is, of course, a theme of James Buchanan's writings, especially his essays on Rawls, that contractarians should not seek to establish particular principles for all times and places. See The Limits of Liberty and Freedom in Constitutional Contract (College Station, TX: Texas A & M Press, 1977).Google Scholar This is also a feature of Gilbert Harman's contractarianism. An accessible presentation of his views can be found in The Nature of Morality (New York: Oxford University Press, 1977).Google Scholar Indeterminateness may also be a feature of morals by agreement once the assumption of self-interest is dropped.

7 It should be noted, as Gauthier does (MA, pp.5 Iff.), that morals by agreement doesn't require a subjectivist account of value, only a relativist (or individualist) one, and that some objectivist theories construe value as agent-relative. I shall come back to this point at the end of my essay when I conjecture that Gauthier can defensibly assume self-interestedness only by abandoning his subjectivism about value.

8 Gauthier, David, “The Incompleat Egoist,” The Tanner Lectures on Human Values, Stanford University, 05 10, 1983, p.73.Google Scholar

9 The phrase, of course, is from Rawls, A Theory of Justice, p. 13.

10 ibid.

11 This would accord with much of what Gauthier says about economic man in Chapters X-XI of MA, as well as in his earlier “The Social Contract as Ideology,“ Philosophy C Public Affairs, Vol. 6 (1977), pp.130–164.Google Scholar It would not, however, be consistent with his characterization of “economic rationality” in “Economic Rationality and Moral Constraints”, Midwest Studies in Philosophy, Vol. 3 (1978), pp.75–96.CrossRefGoogle Scholar

12 Laver, Michael, The Politics of Private Desires (Harmondsworth, Middlesex: Penguin, 1981), p.13.Google Scholar

13 To be disguished from asociality is what I would call asociability. Someone is asociable if she takes no pleasure in the company of others. Someone takes an interest in another's interests or preferences only if the former cares about die latter's well-being or preferences. Sociability, by contrast, I think of as taking pleasure in another's company. (Gauthier uses die term more or less indifferently with the notion of taking an interest in others. See MA, pp.11, 19, 254). Someone may lack concern for others, yet be sociable. It may be that people are rarely asociable without also being self-regarding. Nonetheless, die concepts are not die same.

14 For the latter see MA, pp.205, 220, 259.

15 “In moral theory, die Archimedean point is that position one must occupy, if one's decisions are to possess the moral force needed to govern die moral realm” (MA, p.233).

16 MA, p.235, emphasis added.

17 Kavka, Gregory, Mind, Vol. XVCI (January 1987), pp.117–121.CrossRefGoogle Scholar

18 My grouping of several different motivational assumptions under the umbrella term ‘self-interest’ does not, I believe, affect the argument of the essay. Nonetheless, the reader should be alert to potential problems here. (Ned McClennen has expressed worries to me on this point.)

19 MA, pp.113–114. See David Hume, A Treatise of Human Nature, Book III, Part II, Sections II, and An Enquiry Concerning the Principles of Morals, Section III, Part I; Hart, H.L.A., The Concept of Law (Oxford: Clarendon Press, 1961), pp.189–195;Google ScholarRawls, , A Theory of Justice, pp.126–128.Google Scholar

20 But in fact the Sophists grasped, for the first time in human thought, the standpoint of a person who does stand outside social life, not in her capacities, not in being able to live without society, but in her motivations, in being able to view society as purely instrumental to goals that do not require social life for their formulation (MA, p.312). This line of argument will turn out to be problematic for Gauthier, for one of the main rationales for SI will turn on the matter of the origins of preferences. As I shall argue, preferences that have come about in certain ways may be ignored by rational choice reconstructions of justice in order to provide what I shall call a “fundamental evaluation” of morality.

21 I owe this suggestion to Jean Hampton, who points out to me how much like Hobbes Gauthier would be in his desire to answer such an individual.

22 “Morals by agreement have a non-tuistic rationale. … But it does not follow that [they] bind only non-tuists” (MA, pp.328, 329).

23 See The Groundwork of the Metaphysics of Morals, especially the Preface. Kantian conceptions of necessity and contingency are foreign to the enterprise of morals by agreement, as Gauthier notes:

Rawls supposes that he follows Kant in treating the principles of justice in ‘independence from the contingencies of nature’. In our view human individuality cannot be separated from these contingencies, and moral principles must not deny but reflect them (MA, p.237).

24 I discuss issues having to do with moral standing in section XII. At least it should be clear that morals by agreement will not accord moral standing to, e.g., rational individuals who throw bombs into crowded restaurants and shops. For a further discussion of these issues, see my “Punishment and Moral Standing” (manuscript).

25 I owe this point to L.W. Sumner.

26 It might be argued that SI is a simplification necessary for determinate conclusions. I have no doubt that this may be a motivation for assuming SI, especially in economics. However, it is not a rationale or justification for the assumption.

27 Friedman, Milton, Essays in Positive Economics (Chicago: University of Chicago Press, 1953), pp.3–43.Google Scholar

28 “… which means whether it yields sufficiently accurate predictions.” Friedman Essays, p.15.

29 One major problem is that Friedman's account conflates explanation with prediction. See Ernest Nagel, “Assumptions in Economic Theory”, reprinted in The Philosophy of Social Explanation, Ryan, A., ed. (Oxford: Oxford University Press, 1973), pp.130–138;Google ScholarRosenberg, Alexander, Microeconomic Lams (London: University of Pittsburgh Press, 1976), pp.155–170.Google Scholar

30 This may be too strong in several ways. First, intuitive moral judgments may have heuristic value for rational choice moral theory, pointing in the direction that contractarian theory might investigate. Also, insofar as moral conventions actually exist, for whatever reason, there may be good rational choice reasons for adopting (some of) these rather than seeking to establish new ones. The moral capital of past generations, however established, may be too valuable to do without. Lastly, insofar as morals by agreement is explanatory, positive fit between theory and ordinary moral judgments is confirmatory.

31 Hobbes, Thomas, Leviathan, ed. Macpherson, C.B. (Harmondsworth, Middlesex: Penguin, 1968), Chapter 15, p.203.Google Scholar

32 In “Foundationalism in Ethics”, Ethics: Foundations, Problems, and Applications, Morscher, E. & Stranziger, R., eds. (Vienna: Holder-Pichler-Temsky, 1981), pp.134–136Google Scholar, and “Foundationalism and Contractarian Ethics.”

33 Rawls, , A Theory of Justice, p.580Google Scholar. See also his “Justice as Fairness: Political not Metaphysical”, Philosophy & Public Affairs, Vol. 14 (Summer 1985), pp.223–251.Google Scholar

34 Rawls, , A Theory of Justice, p.129.Google Scholar

35 … we do not suppose that actual moral feelings represent the outcome of a prior valuing of participation and an awareness that voluntary participation requires the acceptance of moral constraints. Rather our argument is that, if we are to consider our moral affections to be more than dysfunctional feelings of which we should be well rid, we must be able to show how they would arise from such a valuing and awareness (MA, p.339).

36 I owe this suggestion to L.W. Summer.

37 Gauthier does not mention the significant result that the Pareto-efficient equilibria of perfectly competitive markets are in “the core” – that is, there are no other outcomes where some subset (or “coalition”) of individuals could improve each member's position independently of that of nonmembers. The standard source is Arrow, Kenneth & Hahn, Frank, General Competitive Analysis (San Francisco: Holden-Day, 1971).Google Scholar For an accessible presentation of the two fundamental theorems of welfare economics, see Feldman, Allan, Welfare Economics and Social Choice Theory (Dordrecht: Kluwer-Nijhoff, 1980),CrossRefGoogle Scholar Chapters 3–4. These results may have interesting implications for determining the membership of the “cooperative ventures for mutual advantage” that constitute contractarian societies.

38 It should be noted that:

Against the market background of mutual unconcern, particular human relationships of trust and affection may flourish on a voluntary basis. Those who hanker after the close-knit relationships of other and earlier forms of human society are in effect seeking to flee from the freedom to choose the persons in whose interests they will take an interest (MAx p.102).

39 I put many economists of Chicago and UCLA persuasions, as well as some heads of government in the US, Britain, and France, in the camp of those who unduly neglect the dangers to which Gauthier refers. Amongst his colleagues, James Buchanan's work is notable for the emphasis he puts on the nonmarket conditions of markets.

40 This, it should be noted, is Wicksteed's nontuism, which is weaker than other versions of SI. Given that agents, for the purposes of economic theory, can be households or other collectivities, it is this weaker assumption which is required. (Of course, it may not be possible to interpret the behavior of, e.g., households, as maximizing a single function, even those the members of which take an interest in each other's interests.)

41 For the assumptions of the standard welfare arguments, see, in addition to the texts mentioned in note 37, Winch, D.M., Analytical Welfare Economics (Harmondsworth, Middlesex: Penguin Education, 1971), pp.34–35,Google Scholar 89. See also the discussion (pp.H7ff.) relaxing the conditions of perfect competition and introducing externalities.

42 “An ‘externality’ occurs if you care about my choice or my choice affects you.” Schelling, Thomas, “Hocky Helmets, Daylight Saving, and Other Binary Choices”, in Micromotives and Macrobehavior (New York: W.W. Norton, 1978), p.213.Google Scholar

43 The analogy to the possibility, or rather, existence proofs of general equilibrium theory I owe to Alan Nelson, who notes “the similarity between proving the existence of economic equilibrium and deriving a state from individual moral principles.” See his “Explanation and Justification in Political Philosophy”, Ethics, Vol. 97 (October 1986), pp.170ff.Google Scholar

44 This is fortunate since we would want to claim that individuals can and (often) should have (or develop) interests in one another's interests. Further, it is theoretically possible to achieve Pareto-efficient outcomes in noncompetitive, socialist economies under certain conditions. See Winch, , Welfare Economics, p.94.Google Scholar

45 Smullyan, Raymond, This Book Needs No Title (Englewood Cliffs, NJ: Prentice-Hall, 1980), p.56.Google Scholar Cited by Elster, Jon, “The Market and the Forum”, Elster, & Hylland, A., eds., Foundations of Social Choice (Cambridge: Cambridge University Press, 1986), p.115.Google Scholar

46 On several occasions in conversation, David Gauthier has given the worry about double-counting as a reason for SI.

47 See Nozick, Robert, Anarchy, State, and Utopia (New York: Basic Books, 1974) pp.6–9;Google Scholar Laver, Private Desires, pp.13ff; and Nelson, “Explanation,’ pp.157ff. See also Jean Hampton, “The Social Contract Explanation of the State” (manuscript, UCLA, 1985), as well as her Hobbes and the Social Contract Tradition (Cambridge: Cambridge University Press, 1986), Chapter 9.Google Scholar

48 Marxist critics, among others, would surely be skeptical as to the requisite independence. We might think of fundamental evaluations or justifications with false principles or norms as “principle-defective” potential evaluations after Nozick's analogous notion of explanation. Insofar as Marxists attempt justifications of states, these may be nonfundamental as they would appeal to principles internal to that which is being evaluated, e.g., capitalism by references to bourgeois principles of justice.

49 The suspicion may remain that without foundationalism, fundamental evaluations lose much of their interest.

50 Rawls, , A Theory of Justice, p.7.Google Scholar

51 As perhaps are also the doubts of many contemporary economists when they suggest, as they so often do, that behind lofty appeals to moral virtue lies mere subjective utility maximization.

52 See MA, Chapter II.

53 Rawls, , A Theory of Justice, p.4.Google Scholar

54 Economists often argue that appeals to any other standard, including that of considered preference, would be “paternalistic” (in the economists’, and not the philosophers’ sense) and thus inappropriate.

55 Recall the old debate on “consumer sovereignty.” The fact mat consumers freely buy the products they are offered did not impress critics who claimed that producers created much of that demand. The latter were implicitly calling for a fundamental evaluation. (On these matters, proponents of Say's Law may not find much interest in fundamental explanations.)

56 A point made essentially by Taylor, Michael in Anarchy, Community, and Liberty (Cambridge: Cambridge University Press, 1982).CrossRefGoogle Scholar

57 I do not claim originality in requiring that our protagonists be stripped of their social attributes. I am following Rousseau here in his criticism of Hobbes’ account of human beings in the state of nature. Rousseau's alleged “noble savage” should not be understood as some ideal man to be contrasted with our fallen, modern selves. Rather, the human being in the early stages of Rousseau's hypothetical state of nature is merely the product of stripping ordinary individuals of all of their social attributes. Since Rousseau believed, contra Hobbes, that essential human attributes are largely a product of socialization, stripping individuals of their social attributes turns them into the naked scarecrows that we find in the Second Discourse.

58 We may thus eliminate preferences that have been formed in certain ways that threaten a fundamental evaluation. It may not be completely true, then, “that in characterizing a being as asocial, we are concerned, not with her origins, but with her motivations and values” (MA, p.310).

59 Suppose that humans are genetically predisposed, for instance, to have concerns for their relative standing. See Frank, Robert, Finding the Right Pond (New York: Oxford University Press, 1985), Chapter 2.Google Scholar We might then want a fundamental evaluation of our actual utility functions. See also Frank, ‘If Homo Economicus could choose his own Utility Function, would he want one with a Conscience?“ American Economic Review (forthcoming).

60 “The assurance that interaction is subject to the constraints of justice [which are determined by self-regarding rational choice] meets the concern emphasized in feminist thought, that sociability not be a basis for exploitation“ (MA, p.351).

61 Recall the distinctions I drew earlier in note 13 between asociability and egoism, as well as Laver's notion of asociality. For an even more fundamental evaluation of our practices and sentiments, then, we may impose the further condition of asociability. A more radical position yet may be achieved by assuming asocial motivation or asociality. From this position we may evaluate (fundamentally) our moral and non-moral tuistic sentiments, as well as our characteristic sociability.

62 Rawls, , A Theory of Justice, pp.3, 7.Google Scholar

63 Hume, , Enquiry Concerning the Principles of Morals, Section III, Part I.Google Scholar

64 It is perhaps this feature of justice that makes many social theorists long for a society “beyond justice” or claim that the virtue is inappropriate for realms, e.g., the family, where people do take interests in others. The first position is often attributed to Karl Marx. Hegel finds the

ethical aspect of marriage [to consist] in the parties' consciousness of this unity as their substantive aim, and so in their love, trust, and common sharing of their entire existence as individuals …

[Thus] marriage, so far as its essential basis is concerned, is not a contractual relation. On the contrary, though marriage begins in contract, it is precisely a contract to transcend the standpoint of contract, the standpoint from which persons are regarded in their individuality as self-subsistent unity.

The Philosophy of Right, trans. Knox, T.M. (Oxford: Oxford University Press, 1967), paragraph 163.Google Scholar Aristotle is well-known for his view that “when men are friends they have no need of Justice”, Nicomachean Ethics, trans. Ross, W.D., Book VIII, 1155a.Google Scholar

65 I owe basis for this objection to SI to G.A. Cohen and L.W. Sumner. In a different context, that of the prebargaining situation, Gauthier claims that: it would be irrational for an individual to dispose herself voluntarily to make unproductive transfers to others. An unproductive transfer brings no new goods into being and involves no exchange of existing goods. … Thus it involves a utility cost for which no benefit is received, and a utility gain for which no service is provided (MA, p.197).

66 This case is explicitly considered by Hume:

Were there a species of creatures intermingled with men, which, though rational, were possessed of such inferior strength, both of body and mind, that they were incapable of all resistance, and could never, upon the highest provocation, make us feel the effects of their resentment; the necessary consequence, I think, is that we should not, properly speaking, lie under any restraint of justice with regard to them, nor could they possess any right or property. … Enquiry Concerning the Principles of Morals, Section I, Part I.

67 This principle governs the distribution of the benefits of cooperation and requires that these be distributed so that the greatest relative concession required of any individual be as small as possible. See MA, Chapter 5.

68 In discussion David Gauthier suggested that a troublesome variant of this case is where Charles's tuistic preferences are such that in the hypothetical bargaining situation, he would insist that Becassine be given moral standing.

Wayne Sumner has argued that the range of beings accorded moral standing by contractarian theory is determined in part by the range of the concerns of the agents. I failed to understand his point at the time, but it now seems correct to me. See Sumner, “Subjectivity and Moral Standing,” Morris, “Value Subjectivism, Individualism, and Moral Standing: Reply to Sumner,’ and Sumner, ”A Response to Morris,“ Sumner, Wayne, Callen, Donald, & Attig, Thomas, eds., Values and Moral Standing, Bowling Green Studies in Applied Philosophy VIII (Bowling Green, OH: Bowling Green State University, 1987), pp.1–15, 16–21, 22–23.Google Scholar

69 James Buchanan and Gilbert Harman, perhaps for different reasons, will both urge that determinate conclusions were never to be expected from this enterprise.

70 In the event that we drop SI, we are faced with Gregory Kavka's case of the Inequality Glutton, an individual whose preferences are, in varying ways, for more than others. See the review mentioned in note 17. Kavka's purpose is to show how counterintuitive Gauthier's principle of minimax relative concession will be given certain preference profiles, a standard by which Gauthier does not intend morals by agreement to be judged. I suspect that others will not find it beneficial to allow such Gluttons to join in their cooperative ventures for mutual advantage. People with certain sorts of concerns for relative standing may not be attractive partners for cooperative endeavors until they forgo those concerns.

71 Philippa Foot suggests that we “ask awkward questions about who is supposed to have the end which morality is supposed to be in aid of. … Perhaps no such shared end appears in the foundations of ethics, where we may rather find individual ends and rational compromises between those who have them.” “Utilitarianism and the Virtues,” Mind, Vol. XCIV (April 1985), p.209.Google Scholar

72 Interpersonal comparisons need not be made here.

73 More precisely, Annabel maximizes a weighted sum of Pa and Pb:

U _a = w _aP + y _aP _b where −1 < w _a, < 1 and −1 < y _a < 1.

When y _a = 0 (and w _a # 0), Annabel is a pure egoist. When w _a = 0 (and y _a # 0), Annabel is a pure altruist. (When y _a < 0, Annabel's “altruism” is negative, i.e., she suffers from envy.)