1 Cf. Rawls, John A Theory of Justice (Cambridge, Mass.: Harward, 1971),Google Scholar pp. 17, 21,577-87.

2 Rawls, A Theory of Justice, p. 20.

3 Rawls, A Theory of Justice, p. 141

4 Cf. Hare, R. M. “Rawls’ Theory of Justice,” Philosophical Quarterly 23 (1973), pp. 144-55,CrossRefGoogle Scholar 241-51; reprinted in Daniels, Norman ed., Reading Rawls, (New York: Basic, 1975),Google Scholar esp. pp. 82ff. Hare cites (p. 84) the Rawls passages quoted above (cf. n. 2 and n. 3 supra).

5 We are not bound to principles to which we did not actually agree. Nor, merely because we would have agreed to them had we agreed to make the contract (they would have been in our antecedent interest) must we now accept them (they may not be in our actual interst). Cf. Dworkin, Ronald “The Original Position,“ University of Chicago Law Review 40 (1973). pp. 500–533;CrossRefGoogle Scholar reprinted in Daniels, Reading Rawls, p. 16-53.

6 Dwor kin, “The Original Position” in Reading Rawls, p. 51.

7 Scanlon, T. M. “Rawls’ Theory of Justice,” University of Pennsylvania Law Review 121 (1973), pp. 1010-69;CrossRefGoogle Scholar reprinted in part in Daniels, Reading Rawls, pp. 169-205, cf. p. 178; Scanlon's view has important similarities to mine.

8 Cf. David Lyons's intersting argument in “Nature and Soundness of Contract and Coherence Arguments,” in Daniels, Reading Rawls, pp. 141-67, esp. 159-60.

9 Rawls, John “Reply to Alexander and Musgrave,” Quarterly Journal of Economics 88 (1974), p. 637.CrossRefGoogle Scholar

10 E.g., the person is calm and has adequate information about the cases being Judged. Considered moral Judgments may be of any level of generality, a shift from earlier characterizations. Cf. Rawls, “The Independence of Moral Theory,“ Proceedings and Addresses of the American Philosophical Association 48 (1974-75), p. 8.Google Scholar

11 The distinction between narrow and wide reflective equilibria is implicit in A Theory of Justice, p. 49, and explicit in “The Independence of Moral Theory,“ p. 8. Rawls has compared the search for moral principles in reflective equilibrium to the search for a grammar in descriptive syntactics: each captures a relevant ‘sense’ or competence. But the analogy at best holds for narrow, not wide equilibria; cf. my “Some Methods of Ethics and Linguistics” PhilosophicalStudies, 37 (1980). pp. 21-36. More importantly, narrow equilibrium is particularly ill-suited as a basis for a Justificational argument: its elementary coherence constraints provide inadequate pressure to revise considered moral Judgments, which have no special epistemological status and are open to many charges about bias, historical accident, and ideology. Cf. Rawls, A Theory of Justice, p. 49.

12 My formulation is not adequate as it stands since there will even be trivial truthfunctional counterexamples to it unless some specification of “interesting” and “non-trivial” is given. I also say nothing about how to measure the scope of a theory. The problem is a standing one in philosophy of science (cf. Michael Friedman's attempt to handle the related question of unifying theories in “Explanation and Scientific Understanding,” Journal of Philosophy 71 (1974), pp.5-14, esp. 15ff.). I assume this difficulty can be overcome, though doing so might require dropping the loose talk about theories. I am indebted to George Smith for helpful discussion of this point.

13 General social theory, which is used to test the feasibility of Level II principles (from behind the veil of ignorance) and Level III theories (from outside the veil), is presumably independent of Level I considered moral Judgments. How much the feasibility test narrows the field is an open question.

14 This point only suggests, but fails to prove, a violation of the independence constraint. Little hangs on the conclusion in what follows.

15 “The aim of the description of the original position is to put together in one conception the idea of fairness with the formal conditions expressed by the notion of a well-ordered society, and then to use this conception to help us select between alternative principles of Justice.” (Rawls, “Reply to Alexander and Musgrave,” p. 638, italics added). Many features of the well-ordered society are not Just formal, since they draw directly for support on the theory of the person.

16 All conditions are taken from “Reply to Alexander and Musgrave,” pp. 634-5.

17 Definitions, for example, would still have to be assessed in light of a well worked out theory. A weak publicity condition might drop (3), for example.

18 Cf. A Theory of Justice, pp. 177-79. Similarly, a strong publicity condition imposes broad limits on the allowable complexity of principles and on the amount of information needed to apply them. But Rawls does not argue for the strong publicity condition because of its connection with these effects nor because these effects might favor his principles over others. Cf. “Independence of Moral Theory,” p. 14.

19 Rawls, A Theory of Justice, p. 133.

20 Rawls, “Independence of Moral Theory,” p. 13.

21 The situation is more complex. See my “Moral Theory and the Plasticity of Persons,“ Monist 62:3 (1979), pp. 265-287.

22 Rawls, “A Kantian Conception of Equality,” Cambridge Review (February, 1975), p. 94. The wording also appears in condition (4) of well-ordered societies; see above.

23 Rawls, “A Kantian Conception of Equality,” p. 94.

24 Cf. Rawls, “A Kantian Conception of Equality,” p. 95; also, Rawls, John “The Basic Structure as Subject,” American Philosophical Quarterly 14, (1977), p. 160.Google Scholar

25 Rawls, “A Kantian Conception of Equality,” p. 94.

26 Cf. Rawls, “A Kantian Conception of Equality,” p. 94.

27 Cf. Rawls, A Theory of Justice, section 25.

28 Cf. Rawls, A Theory of Justice, pp. 256ff; also, Rawls, “Kantian Conception of Equality,” p. 98.

29 Rawls treats some of these as straightforward empirical claims about humans. Thus he thinks that (5) is true except for pathological cases, and he therefore treats ‘having a capacity to form a conception of the good and to develop a sense of Justice’ as a natural (range) property. Since this range property is equal in all persons, he infers it is a relevant basis on which to view persons as equal. Cf. A Theory of Justice, p. 505-8.

30 I am indebted to Josh Cohen for discussion of this point.

31 “A Kantian Conception of Equality,” p. 94.

32 Rawls thinks philosophy of mind and general social theory underdetermine even such a central feature of persons as the criterion of personal identity: its selection depends on coherence of the theory of the person with the rest of moral theory in wide equilibrium. This view raises many of the same questions about the independence constraint. In any case, Rawls's view of the ‘plasticity' of the person suggests he treats the theory of the person as an ideal, not an idealization. Cf. my “Moral Theory and the Plasticity of Persons,” section V.

33 There is only one reference to rights in Level Ill arguments in A Theory of Justice: In section 77 Rawls discusses the equality of moral personhood as a basis for determining to whom Justice is owed, and he remarks in a note that “This fact can be used to interpret the concept of natural rights” (p. 505, n.30). But here the talk of rights can be viewed as (misleading) shorthand for the underlying theory of the person. References to Level Ill rights enter Rawls's writings more freely after Ronald Dworkin's “The Original Position” appeared (e.g. “The Kantian Conception of Equality,” p. 94); moreover, Rawls cites Dworkin in a relevant context, while discussing the role of the well-ordered society (cf. “Reply to Alexander and Musgrave,” p. 634, n. 1.) The real issue is not, of course, whether Rawls accepts Dworkin's view (I think he does not), but whether we are compelled to accept it.

34 Cf. Dworkin's more detailed argument, “The Original Position,” pp. 42-53.

35 Cf. Rawls, “Reply to Alexander and Musgrave,” p. 638.

36 I am indebted to Miles Morgan for discussion of this point.

37 Rawls, A Theory of Justice, pp. 140,564-65.

38 Rawls, A Theory of Justice, pp. 141-42.

39 Rawls, A Theory of Justice, p. 264.

40 Rawls, A Theory of Justice, p. 257.

41 Rawls, A Theory of Justice, p. 257.

42 Cf. Rawls, A Theory of Justice, pp. 564-65.

43 It should be remembered I am nowhere assessing the soundness, only the structure, of Rawls's Justificatory argument.

44 Nozick, Robert Anarchy, State, and Utopia (New York: Basic Books, 1974) pp. 204ff.Google Scholar

45 Rawls, “The Basic Structure as Subject,” p. 159.

46 Rawls, “The Basic Structure as Subject.'’ p. 160.

47 Rawls, “The Basic Structure as Subject,” p. 161.

48 Cf. Hare's criticism in “Rawls’ Theory of Justice,” pp. 89-94, 101-7.

49 What still needs explanation are Rawls's remarks suggestive of such rigging. The key lies in the heuristic role of the contract idea. Imagine asking, “can we work backwards from the contractarian task of choosing principles to figure out what conditions on choice will yield the preferred principles?” We can, and we may suppose Rawls did. But after such heuristics, we still have to show why the conditions are acceptable.

50 Rawls, “Reply to Alexander and Musgrave:’ p. 637; cf. G. E. Pence, “Fair Contracts and Beautiful Intuitions:’ in Nielsen, Kai and Shiner, Roger eds., CanadianJournal of Philosophy, Supplementary vol. 3 Guelph: (Canadian Association for Publishing in Philosophy; 1977), p. 143.Google Scholar

51 Cf. Parfit, Derek “Later Selves and Moral Principles,” in Montefiore, Alan ed., Philosophy and Personal Relations, (London: Routledge and Kegan Paul, 1973), esp. pp. 149-60;Google Scholar and Rawls, “Independence of Moral Theory;’ pp. 17ff; and my “Moral Theory and the Plasticity of Persons.“

52 These and some of the following points are pursued further in my “Wide Re· flective Equilibrium and Theory Acceptance in Ethics,” Journal of Philosophy 76 (1979), pp. 255-82.

53 Confusion on this point leads some to construe reflective equilibrium as a form of intuitionism. Cf. Hare, “Rawls’ Theory of Justice,” p. 83, and Singer, Peter “Sidgwick and Reflective Equilibrium,” Monist 58 (1974), pp. 493-4.CrossRefGoogle Scholar

54 Points (2) and (3) are clearly stated in Rawls's “Independence of Moral Theory.“ Cf. n. 32 above.

55 Cf. my ‘Wide Reflective Equilibrium ' Theory Acceptance in Ethics,” sections IV and V.

56 The National Endowment for the Humanities Fellowship program funded this and related research. I am grateful to Josh Cohen, Miles Morgan, John Rawls, and George Smith for helpful discussion.