The construction of the "original position" in these sections is by reference to three moves. Firstly, Rawls discusses what he terms the "formal constraints of the concept of right" which constraints are meant to flesh out what can be properly said to belong to the concept of right and which succeeds in ruling out for consideration in the "original position" one conception that was still left in play at the conclusion of his account of the "initial situation". Secondly, the "veil of ignorance" is brought into play and its function is made clear in terms of further accentuating the role of one of the elements of the "formal constraints of the concept of right". Finally, a discussion of the rationality of the parties involved in the original position is introduced and this discussion concludes Rawls' account of what it involves to comply with the "formal constraints of right". Essentially then, the account in section 23 of the "formal constraints of right" does the majority of the work in constructing the "original position" with the veil of ignorance and the discussion of rationality acting as supplementary reinforcements of what is concluded in the discussion of these formal constraints.
The idea that there are "formal constraints" involved in the concept of right is meant to determine the shape of the types of principles that can be considered in the "original position" as reasonable ones to be entertained. These constraints do not, Rawls is careful to point out, arise as a result of a definitional approach to the concept of right but are rather intended as a means of facilitating the process of reflective equilibrium. There are five conditions that emerge as requirements for facilitating such equilibrium.
The first such condition is that principles to be considered in the "original position" have to be general in form which ensures that they cannot be formulated in a way that requires reference to "proper names" or "definite descriptions". The reason given for this generality constraint is that principles of justice should be unconditional and require no reference to contingent particulars. So generality as a constraint is given as a means of ensuring the unconditionality of the principles' scope and effects.
The second constraint on principles to be considered is that they should be universal in application and one of the ways indicated of ensuring that this constraint is met is that the principles should be generally understandable (which indicates a way of bringing the condition of universality in relation to the intuitive notion of "common sense"). Hence the principles cannot be overly complex in formulation or draw innumerable types of distinction. This criteria of universality is also used to rule out principles that are self-defeating since such defeasibility implies particularity of comprehension and application. One of the implications of such a commitment to universality is that it cannot be right that principles be selected that would have consequential import for social justice only if selectively followed. Principles that are ones of social justice cannot be regarded as fulfilled formally if they require such selective application. Rather, they are "to be chosen in view of the consequences of everyone's complying with them".
Generality and universality as the first two conditions have some special importance as will become clear when we turn to looking at the "veil of ignorance". However, they also need to be distinguished since generality touches on different criteria from universality. It is, according to Rawls, possible to universalise the conception of egoism as a first-person dictatorship as all could act in accordance with this principle. However, it is not a principle that is general since it requires reference to first-person contingencies. Similarly a principle can be general without being universal and principles that are weighted to specified groups are general in form without being universal in application. So generality, whilst important, and more important as we will see than Rawls first suggests, is in itself an insufficient constraint since it does not rule out principles that appeal to the conditions of specific groups.
The third condition, and the one that is of most interest from a Kantian viewpoint, is publicity. "The parties assume that they are choosing principles for a public conception of justice." In referring to this criteria Rawls explicitly refers to the discussion of publicity in Perpetual Peace that was treated at some length in previous postings on this blog. See for example, my earlier postings on annexation and the affirmative principle of publicity. Rawls cites not only Perpetual Peace here but also the description of "public right" from the Doctrine of Right where Kant speaks of public right as those laws "which require to be made universally public". Additionally Rawls refers to Kant's remarks ruling out "secret reservations" with regard to constitutions in the essay on theory and practice, a remark intended there to bolster Kant's case against the view that there is a "right" to rebellion.
The discussion of publicity is importantly related by Rawls to the notion of "stability" of conceptions of justice as the general awareness of universal acceptance of principles of justice should "support the stability of social cooperation". If principles are known to be acknowledged and taken as an appropriate guide by others then social cooperation is given a basis it otherwise lacks and this point about stability is effectively used as Rawls' justification of reference to publicity. Interestingly, this support for publicity is also part of Kant's case for an affirmative rather than merely negative conception of publicity. (Equally interestingly, Rawls does not here specify if his conception of publicity is positive or negative.)
The final interesting connection between Rawls and Kant when it comes to the constraint of publicity is that Rawls reads Kant's reference to the notion of the "universal law of nature" in the typic of the categorical imperative as involving an implicit publicity condition in its application. Rawls presents the typic as also a way of reading Kant's references to the "kingdom of ends" stating that this kingdom is "an ethical commonwealth, as it were, which has such moral principles for its public charter". This reference suggests Rawls was referring to Kant's account of the "ethical community" in Religion Within the Limits of Reason Alone where it is formulated as "a people under divine commands, i.e. as a people of God, and indeed in accordance with the laws of virtue" (Ak. 6: 99). Within the account of the kingdom of ends in the Groundwork, by contrast, no such ethical community is so clearly brought into view (though it is far from odd to see it as part of Kant's point there).
The fourth "constraint" is that there should be an order imposed on conflicting claims. This constraint is meant, clearly, to include in the construction of the original position a means of forestalling intuitionist scepticism. The ordering is meant as transitive and clearly the form of ordering has to match the earlier constraints of being general, universal and publicly acceptable.
The final "constraint" Rawls specifies is that principles selected within the "original position" should be final so that if reasoning is squarely and consistently based on these principles that should end the question of the acceptability of the reasoning in question. After stating this finality condition Rawls summarises the account given of the five formal constraints of the concept of right: "a conception of right is a set of principles, general in form and universal in application, that is to be publicly recognized as a final court of appeal for ordering the conflicting claims of moral persons".
An implication of the conjunction of the five formal constraints of right is the egoistic principles that were still left in play in the "initial situation" are ruled out of consideration within the "original position" constructed as the latter is by means of these formal constraints. Generality prevents appeal to first-person dictatorship and to free-riding claims though it is insufficient to exclude general egoism as the latter clearly is a general conception. The latter should be capable of being ruled out on the basis of universality but Rawls appeals not to that but rather to the need for ordering amongst principles since the universal application of general egoism would allow for no priority rules with regard to the separate claims. This is the same as saying that the adoption of general egoism is equivalent to a state of nature or lack of justice. In making this move Rawls effectively makes his case against intuitionist scepticism a social one since acceptance of the theory in question would prevent conditions of justice having any possibility of prevailing. If the formal constraints that define the construction of the "original position" are sufficient to rule out egoism by themselves then the construction of the "original position" has already scored a direct advance upon the account of the "initial situation".
Having constructed the "original position" through the description of the formal constraints of right Rawls proceeds next to accentuate its characterisation by reference to the "veil of ignorance". The point of the "original position" is to use "the notion of pure procedural justice as a basis of theory". Through the "original position" we can annul the appeal to special circumstances that have weight in ordinary social life. However Rawls next appeals to the "veil of ignorance" and the reason for this appeal is that by means of it being brought into play the parties in the "original position" will have to decide principles "solely on the basis of general considerations".
Hence the point of the "veil of ignorance" is to bring the generality constraint "closer to intuition" as Kant would put it. The "veil of ignorance" thereby operates in a way akin to Kant's typic of the categorical imperative and this reference is explicitly made by Rawls. Rawls writes of the way Kant operates the typic, which is by appealing to the universal law of nature, that Kant "must suppose that we do not know our place within this imagined system of nature". So Rawls views the reference to the universal law of nature as a means of enforcing the "veil of ignorance" and, similarly, he takes the introduction of the "veil of ignorance" to be a means of making the generality constraint publicly accessible and easier to comprehend.
The introduction of the "veil of ignorance" prevents the parties in the "original position" from appealing to natural assets, their own conception of the good, contingent elements of their psychology and even to features of their own society. Left outside this "veil" are the "circumstances of justice" that were introduced at the conclusion of the account of the "initial situation". Other, more problematic elements left outside the veil include "the principles of economic theory" and "the laws of human psychology". (Which "theories" and which "laws"?) One of the reasons the latter is assumed to be able to be left outside this "veil" is that it would count against a conception of justice if the requirements of it were too stringent for human beings as otherwise there would be "difficulty in securing the stability of social cooperation". Overly stringent conceptions would thus violate the reference to stability that we saw was used as justification for the constraint of publicity. In an important sense overly stringent conceptions would not be publicly justifiable. Conceptions of justice should, states Rawls, generate their own support so information that would ensure stability can be included as available despite the "veil".
The "veil of ignorance" is meant as a device that enables one to reason and the way it does this is, as suggested, by making the generality constraint more vividly available to one. "It must make no difference when one takes up this viewpoint" of the original position Rawls writes and the "veil of ignorance" is a condition of meeting this generality requirement as it insures that information available is "at all times the same". However the "veil" also serves another methodological requirement, that of providing a procedure that enables not merely generality of form but generality of agreement. Kant spoke of people only consenting to a law that they were capable of giving to themselves and this is echoed in Rawls' description of the "veil" as it is intended that, under the conditions of its application, that "a unanimous agreement can be reached".
Since no one knows their position in the social structure or the potential advantages and disadvantages of natural assets they would possess they reason in such a way that they adopt an effectively general standpoint and reach conclusions that are universal in application. This is thus a way of mapping autonomous reason. But there is one element left out and this indicates the reason why Rawls, at the conclusion of his account of the "initial situation" made reference to the problem of justice between generations. What is left out is a means of neutralising the place one has in the history of generations as one is existent now and not later so the conditions of the "veil" have not, as yet, prevented this contingency from having an effect on one's reasoning as one could reason in a general and universal way, publicly affirming principles that gave advantage to one's generation over and against future ones. This is why in section 22 Rawls referred to a "motivation assumption" being built into the parties of the "original position" even before he constructed this position and this assumption is that the parties were "heads of families" or that they required the parties to agree to principles subject to the "constraint" that they wish preceding generations had also adopted these principles. The former would motivate consideration of succeeding generations but would do so at the cost of ensuring an important piece of information concerning contingent circumstances was introduced so the latter way of specifying the constraint is surely to be preferred. If we wish previous generations had adopted our principles then this is equivalent to saying that we expect future generations will approve of our principles. So whatever our temporal position we are thereby "forced to choose for all".
The "veil of ignorance" is supplemented well by this further constraint and, given it, operates to neutralise all possible advantage in consideration of principles of justice. Further it provides a simulated way of ensuring that principles chosen will be such as to be universally acceptable whilst also making the choice situation much simpler, a desideratum in terms of ensuring that the choice situation is indeed constructed by reference to criteria that are generally understandable. This point about simplicity is also adduced by Rawls as a further constraint on principles of justice that will be chosen within the "original position" as, cateris paribus, principles that are simpler in their general form are to be preferred as the public conception of justice should be "evident to everyone".
The first two stages of the construction of the "original position" were thus, initially, the specification of a set of constraints on the types of principles that could be chosen within it and, secondly, a neutralisation of information that would give weight to contingencies that would threaten to render unstable commitment to the constraints previously specified. The third and final stage in the construction of the "original position" concerns the way in which reasoning within this position should be governed.
The introduction of the reasoning process involved in the "original position" is by means of how the neutralisation involved in the "veil of ignorance" is to be balanced by reference to a means by which the parties are capable of comprehending what is good without reference to a pre-existent view of the good. In resolving this conundrum Rawls does not merely suggest, as might have been expected from his account of the priority of the right over the good, that a view of the good will emerge from attending to the formal constraints of right. Rather, he refers to the list of primary social goods as being itself an "account of the good". This account has come to be referred to in discussion of Rawls as the "thin theory of the good". The view simply is that all will, at least under "normal" circumstances, "prefer more primary social goods rather than less".
This "theory of the good" will enable deliberation to take place within the "original position" as now theories can be assessed in relation to something. After this has been given Rawls can introduce what he takes to be a standard theory of rationality by means of which reasoning will be expected to be guided within the "original position" and this involves the view that persons are generally committed to a coherent set of preferences and follow the plan that satisfies more rather than less of these. Rawls does, however, add a special assumption here which is that the parties in the "original position" are free from envy and the reason for this is that envy has a generally disadvantageous social effect. Finally, the rationality in question includes the notion that the parties engaged are each possessed of a "sense of justice" and the possession of such a sense is publicly known. Once these points have been added we can say that the "original position" allows for strict compliance with the principles that will be chosen.
The "original position" has thus been constructed by means of firstly specifying constraints on the principles of justice that can be chosen within it and then by adding a procedure that ensures obstacles to acceptance of these constraints are removed and a thin model of rationality has been sketched that includes a thin theory of the good, the latter two enabling the parties in the "original position" to be genuinely seen as human and thus governed by the "circumstances of justice". Once the position has been constructed the principles that Rawls believes would be chosen within it can be considered.