The third of Parfit's 2002 Tanner Lectures concentrates specifically on contractualism and eventually becomes Chapters 15 and 16 of On What Matters. The discussion of contractualism in the 2002 lecture is effectively divided, however, into 2 distinct parts, the first of which first derives a reason for discussing Rawls' conception of contractualism and then presents a criticism of the view that it can provide us with a basis for determining the wrongness of acts. The second part of the lecture, by contrast, presents a case instead for what Parfit terms "Kant's contractualist formula", a notion the derivation of which at the close of the second lecture I traced in a previous posting. However the account of Rawls turns out itself to be divided into 2 parts with the first part considering Rawls' conception of rationality and the second his account of the veil of ignorance. In this posting I am going to concentrate only on the account of rationality and in subsequent postings I'll look at the discussion of the veil of ignorance and the account of "Kant's contractualist formula".
Parfit opens the third lecture by defining the basic notion of contractualism as offering an agreement on which moral principles we can accept. As Parfit puts it, this appears to imply what he terms the "Rational Agreement Formula" defined as stating that we ought to act on the principles to whose acceptance it would be rational for everyone to agree. However, notably Parfit immediately assumes a view of rational agreement as involving choices that would be "likely to be best" for those making the choice and whilst in a way that seems alright it does also seem to imply that the process of attaining agreement is defined by axiological criteria, something decidedly not neutral between different conceptions of rationality. Not only does this appear to be the case but Parfit further assumes that the parties coming to the contract are ones that have available to them extensive knowledge concerning their present position as it refers, for example, to different possible scales of understanding of what is best for one should one be rich as opposed to being poor.
The basic point of the agreement is defined as indicative of a general understanding that the world without an agreement is, in some important way, radically deficient so that there is a basic incentive to reach agreement. Parfit indicates that some theorists see this point about the preferability of the world with the agreement above the world without it as the "essence" of the contractualist position. This assumes that the outcome is to reach a mutually advantageous bargain and thus to view just or right principles as simply those that emerge from such a bargain rightly constructed. Parfit adds to this picture a complication when he points out that, given the assumption of general information being available as to one's present condition, it follows that some have more incentive than others to reach an agreement. Roughly, those who are worst off now are likely to be more desperate for an agreement than those who are currently doing reasonably well for themselves. So the latter might hold out for principles that would secure their special status rather than place it in the melting pot. This picture of the contractualist view is defended by contemporary Hobbesian theorists such as David Gauthier and explicitly repudiated by Rawls. Rawls puts the problem with this view clearly enough when he describes it as dependent on a conception of "each according to his threat advantage".
It is after setting the discussion up in this way that Parfit turns now to assessment of Rawls' version of contractualism. Parfit opens with an account of what he takes Rawls' assumptions concerning rationality to involve. Parfit indicates that Rawls accepts the Deliberative Theory of rationality according to which we ought rationally to do whatever would best achieve what we most want after informed deliberation. However Rawls is clear that he disentangles this conception from any reference to a Self-Interest Theory of rationality. In describing Rawls' view in this way Parfit is drawing on section 86 of A Theory of Justice (part of the final Chapter IX) where Rawls describes "rational choice" as involving the notion of deliberative rationality and as including the sense that it is better to act from the standpoint of justice above all else and even that it is rational to do so. When Rawls brings this point in to his discussion it is part of his description of acquiring a sense of justice that is "truly final and effective" so that justice has precedence over all else and is publicly so understood. However Parfit assumes that the theory of rationality in question with which Rawls is working is one that is, as Parfit puts it, "desire based" as followed also from his earlier account of deliberative rationality as concerned with we "most want" after informed deliberation. This view of Rawls' conception of rationality as based on desires is, I think, quite mistaken.
Rawls refers not to "desires" but to a conception of "primary goods" and he understands the latter to include, for example, such goods as respect and self-esteem, not terms that fit with standard empiricist models of "desire" to which a "desire based" theory belongs. It is not simply a case for Rawls of determining what we "most want" but rather an account of how "primary goods" that everyone is assumed to require can be savoured by all in a setting that guarantees security and justice for each. Rawls draws on a conception of persons as "free and equal rational beings" and devises a contractualist view that is meant to capture and further this sense of what persons are like. The reason why Parfit makes the mistake that he does here appears when he responds to the discussion of the agreement in section 86 of Theory. Parfit quotes this section to support the view that Rawls' theory of rationality is "desire based" as in this section Parfit takes Rawls to make an important admission. The admission, according to Parfit, is that there may be some people who would benefit from injustice and not care about morality and to such people we cannot honestly recommend justice as a virtue.
Now the passage to which Parfit points is one in which Rawls does not state simply the view that Parfit attributes to him. Rather, what Rawls does is mention the conception attributed to him by Parfit as arising from Philippa Foot, not as a conception that he endorses. The people in question that would deny caring about morality would be operating only according to a "thin" theory of the good and Rawls responds to Foot's point by asking whether it is correct to say that in the original position it is only apt to recommend people to act in terms that can be justified by the thin theory of the good alone. Rawls denies that the parties to the agreement are confined by such a condition and thus denies that the parties to the agreement are ones that find those who who don't think morality is a benefit are people who they should accommodate but this does not mean that reasons do not publicly exist as to why such people should nonetheless comply with the agreement that surpasses the thin theory of the good. The people who are denying that they take morality seriously are operating only with a general conception of egoism which Parfit has already acknowledged is not part of deliberative rationality for Rawls. The principles of right and justice are collectively rational and because they are it is in the general interest of all that all should comply with just arrangements. So if there are some who state that they don't find that it is in their nature to act morally then to them Rawls states one can answer both that this does not mean all the same that requirements of deliberative rationality don't apply to them (which provides a rational answer to them) and that "their nature is their misfortune" and not a ground for something being different.
So Rawls simply does not accept that there are some who have reasonable outstanding grounds for refusing to accept the constraints of right and justice and hence he does not rest his view of deliberative rationality on requirements motivated by an empiricist conception of desires. It also follows that there are ends that rationally people should accept for their own sake on Rawls' view, the ends that are expressed in primary goods and which are secured and furthered by recognition of their rational nature. Parfit makes similar mistakes when he assumes that Rawls' recognition of the "impartial spectator" model of rationality is part of a "desire based" view of reasons when, to the contrary, it is part of a problem Rawls has with the way such a model does not include a sufficiently robust view of motivation which is why Rawls writes in section 30 of Theory that: "The impartial spectator definition makes no assumptions from which the principles of right and justice may be derived" and, in this respect, it is quite unlike the contract doctrine that Rawls is defending. Parfit's criticisms of Rawls' general model of rationality are thus quite wrong-headed and seem to turn on the requirement that a conception of rationality that is not "desire based" has to be one that is consequentialist. Rawls' conception, however, is neither desire based nor consequentialist and Parfit has mis-represented it.
Parfit appears to partially recognise that he has not got Rawls' model right when he later argues that Rawls appears at times to make assumptions about well-being, something that Rawls does not do and which appears to be a confused recognition of the theory of primary goods. Parfit next discusses what he terms "Rawls' Formula" which is the formula of the veil of ignorance and the rest of Parfit's discussion of Rawls turns into an account of the veil of ignorance. I'll turn to this discussion in a subsequent posting.