Rawls and subject 19
Put differently, this explains our insistence on distinguishing between the person occupying a standpoint and the standpoint itself. The person occupying the standpoint accepts the standpoint’s constraints and deliberates and justifies in accordance therewith. To which we must add that the person can abstract at any point from the standpoint to come back on the sequence as a whole without taking the standpoint’s constraints with her. Certainly, the judge standpoint has fewer constraints on information and knowledge available to deliberation and justification, but it still accepts, at least in part, the results of the preceding stages, all of which the person, you or me, outside of that standpoint are free to reject.
This point recalls that which we made at the outset of the exposition of the four-stage sequence, namely that the four-stage sequence is a device of representation or an abstraction. For our benefit, Rawls reminds the reader of just this when he closes §31 in remarking:
It is essential to keep in mind that the four-stage sequence is a device for applying the principles of justice. This scheme is part of the theory of justice as fairness and not an account of how constitutional conventions and legislatures actually proceed. It sets out a series of points of view from which the different problems of justice are to be settled, each point of view inheriting the constraints adopted at the preceding stages. Thus a just constitution is one that rational delegates subject to the restrictions of the second stage would adopt for their society. And similarly just laws and policies are those that would be enacted at the legislative stage.
In other words, the four-stage sequence consists in a manner of rendering the principles of justice arrived at in the original position more concrete and readily applicable to the kinds of deliberation and justification which the person is likely to encounter outside of the decisionmaking procedure. In this way, the procedure issues in a relatively determinate political conception with which the person may then confront the society in which she finds herself and remain assured that her basis for doing so remains objective and autonomous.
This answer anticipates one question that the reader might put Rawls: what is the person to do with the political conception which she has formulated with the aid of the four-stage sequence? As suggested, she would make use of such in her deliberation and justification of different positions with regard to the society in which she lives. One may further suppose that her deliberation and justification will confront that of others who have similarly arrived at a political conception through the four-stage sequence. Yet nothing suggests that the other persons will have arrived at the same political conception, even in following the same decisionmaking procedure, i.e. the four-stage sequence. The author allows for just such a possibility, all the while turning it to his advantage.
Of course, this test is often indeterminate: it is not always clear which of several constitutions, or economic and social arrangements, would be chosen. But when this is so, justice is to that extent likewise indeterminate. Institutions within the permitted range are equally just, meaning that they could be chosen; they are compatible with all the constraints of the theory. Thus on many questions of social and economic policy we must fall back upon a notion of quasi-pure procedural justice: laws and policies are just provided that they lie within the allowed range, and the legislature, in ways authorized by a just constitution, has in fact enacted them. This indeterminacy in the theory of justice is not in itself a defect. It is what we should expect. Justice as fairness will prove a worthwhile theory if it defines the range of justice more in accordance with our considered judgments than do existing theories, and if it singles out with greater sharpness the graver wrongs a society should avoid.
On that count, the question resembles that which we asked of reflective equilibrium above and to which we answered, with Rawls, that a range of variation is admissible so long as the points on that range meet certain conditions and approximate our considered judgments. Yet, when applied to the political domain, specific decisions must be made on processes, statutes and cases. Thus, the fact remains that it may prove necessary for the person to adjudicate between rival political conceptions, for which further specification is needed concerning the forum, the varieties of deliberation, justification and reasons admissible therein, and the kind of agreement sought between conceptions. It is just such specification which Rawls seeks to provide with the three phases of justification proceeding in view of public reason and overlapping consensus.
 TJ, p. 176.