Introduction

There is a long-standing tradition in moral philosophy to found a theory of distributive justice on some concept of “impartiality” as a prerequisite to the acceptance of its principles as being universal moral obligations. In this tradition there are two non-independent strands. One, found in most “social contract” theories, emphasizes the procedural justification to moral principles: The moral imperatives are “impartial” because the procedure by which they were determined is “impartial.” The second, eminently represented by the Kantian categorical imperative, emphasizes individual autonomy as a basis for moral laws, and the impartiality of these laws results from the universality of reason.

The idea of “original position” as used in recent theories of distributive justice, in particular Rawls' and Harsanyi's, is a convenient way to integrate the two strands – although differently, because in the case of Rawls the original position is thought of as an original negotiation between rational representatives, whereas in Harsanyi the original position consists in one individual playing the role of a “sympathetic but impartial” decision maker facing the equal-chance lottery of being anyone in society. The original position is always described as a hypothetical situation in which any person, as a moral observer, is supposed to forget his own characteristics. The parties in the original position are put “behind a veil of ignorance.” However, in order to choose principles of justice, essentially based on interpersonal comparisons, a given amount of common information is required.