1 Rawls, John, A Theory of justice (Cambridge, Mass.: Belknap Press of Harvard University Press, 1971).Google Scholar All references to this work will be given in parentheses in the text. I am indebted to Peter Singer and Raziel Abelson for reading and commenting on several versions of this paper.

2 For example, Rawls says, “Recall that the original position is meant to incorporate widely shared and yet weak conditions … At the basis of the theory, one tries to assume as little as possible.” (129).

3 There is a great problem, however, about how such assumptions are to be made and where in the theory. In his review of Rawls’ book R. M. Hare points out that Rawls differs from other users of contract theory in covertly smuggling in many assumptions through intuitionist talk about “natural” and “reasonable“ assumptions, (“A Critical Study of Rawls’ A Theory of justice”, Philosophical Quarterly, Vol. 23, No. 91). In another article, Hare also disputes Rawls’ claim that his assumptions on the contract are the only possible ones by which one can get the desired principles.Hare argues that the constraints imposed by the “rational contractor theory” and his own theory are “for practical purposes“ equivalent. (“Rules of War and Moral Reasoning”, Philosophy and Public Affairs, Vol. 1, No. 2 (Winter, 1972), pp. 167-171).

4 However, it is still true that it is always logically possible for the intuitions in reflective equilibrium to require a deep change in moral principle, perhaps because of a greatly changed social environment or economic upheavals in a society.

5 The relevant arguments are in Chapter VII.

6 Rawls admits that “ … the more specific objections are correct in implying that the force of the argument depends on the features of the consensus appealed to. (581).

7 Rawls, J., “Justice as Fairness”, Philosophical Review, Vol. LXVII (1958),Google Scholar and “Outline for a Decision Procedure in Ethics”, Philosophical Review, Vol LX (1951).

8 See Dretske, Fred, “Epistemic Operators”, journal of Philosophy, Vol. 63, No. 24 (1970).Google Scholar

9 For a discussion of these problems see Peter Singer's, “Sidgwick and Reflective Equilibrium” (Monist, Vol. 58. No. 3).

10 Rawls, in two places, characterizes his method as aiming to be a “kind of moral geometry” (121, 126). He says we should aim for all the “rigor which this name connotes” (21). His argument, he says, “aims eventually to be strictly deductive“ (121). Rawls fails to see that the method of reflective equilibrium contradicts the ideal of a moral geometry. In a strict deduction, once one accepts certain premises, one cannot balk at accepting the logical conclusions. Such a balk must occur when we are unable to accept certain consequences on grounds of conflict with deep moral beliefs, and thus, when we are forced to go back and modify the “premises” of the original position or the “deduction”.

11 For example, Henry Sidgwick, a classical utilitarian, could not accept these constraints without qualification. Sidgwick wrote, “Generally, a Utilitarian in recommending, by example or precept, a deviation from an established rule of conduct desires his innovation to be generally imitated. But in some cases he may neither expect nor desire such imitation; though such cases are rare and difficult to determine.” Methods of Ethics (Dover, 1907), Section 3, Chapter V, Book IV.

12 This distinction is due to Hare, R. M. “Rules of War and Moral Reasoning”, Philosophy and Public Affairs, Vol. 1, No. 2 (Winter, 1972), pp. 167–171.Google Scholar

13 Remarks in the last two paragraphs are suggested by Dworkin, R.'s, “The Original Position”, University of Chicago Law Review, Vol. 40, No.3 (Spring, 1973), pp. 528 ff.CrossRefGoogle Scholar

14 This problem explains, I think, the last-minute addition of a “thin theory of the good” or a theory of motivation. The theory of primary goods is an obvious stopgap measure, but insufficient. The same problems reappear in it since real persons differ about how much they desire of any primary good like wealth, power, opportunity, etc. And more important, they differ as to how much other people should desire wealth, power, etc.

15 R. M. Hare, in his review of Rawls’ book (op. cit.) may misunderstand Rawls remark that, “We want to define the original position so that we get the desired solution.” The desired solution is not, as Hare says, the Rawlsian principles or justice, but instead, the satisfaction of unanimity.

16 This point has been made by several critics of Rawls, including R. M. Hare, cop. cit., (“Critical Review“); Nagel, T., “Rawls on justice”, Philosophical Review, Vol. 62, No. 2 (April 1973), p. 255;Google ScholarCare, N., “Rawls and Moral Contractualism”, Review of Metaphysics, Vol. 23 (1969).Google Scholar

17 R. M. Hare, “Critical Review”, op. cit., pp. 247-247. Hare shows that the requirement of impartiality may be insured in a variety of ways. Hare only argues, however, that the veil of ignorance is not the only way to insure impartiality. He did not see that impartiality was already insured via the unanimity requirement.