In TJ §40, Rawls claims that his theory of justice can be given a Kantian interpretation. This formulation refers more to the justification than to the content of the theory. In an earlier but quite clearly separate discussion (TJ 29), Rawls points out that the content of his theory, the argument for the two principles of justice, expresses a version of Kant’s principle that persons are to be treated not as mere means, but also as ends in themselves. Utilitarianism, Rawls argues, violates this Kantian principle by allowing the happiness of a minority to be traded off for the greater happiness of a greater number. The principle of utility could not be chosen by the parties in the original position, since they cannot rule out that they will occupy the position of the minority. Here the substantively Kantian, anti-utilitarian content of the theory is dictated by the specific constraints of the original position. How these constraints themselves are to be justified is a separate and deeper question, and it is this question that the Kantian interpretation is meant to address.

According to TJ §40, the original position can be understood as a procedural interpretation of Kant’s idea of autonomy. Rawls notes that the central idea of Kant’s moral philosophy is that “moral principles are the object of rational choice” (TJ 221). The parties in the original position are of course understood as making a rational choice of the two principles of justice under conditions of uncertainty. In an early critical article, Oliver Johnson objected to Rawls’s appeal to a Kantian interpretation, arguing that because parties in the original position are clearly trying to advance their own interests, their choices are, in Kant’s terms, heteronomous rather than autonomous (Johnson 1974).