Abstract

Lewis' concept of admissibility was introduced as an integral part of his famous Principal Principle; the principle that initial rational/reasonable belief should conform to objective chance unless there is evidence to the contrary. At that time Lewis offered only the rough and ready characterisation that evidence not to the contrary of such dependence is admissible. This, together with some sufficiency conditions, served well enough until it became clear that admissibility was central to debates on the viability of Humean Supervenience and the analysis of objective chance. In response, Thau and Lewis refined the concept of admissibility in various ways. Since the mid 90's those who have employed the concept have, with minor variations and additions, followed the Thau/Lewis line. Yet, in the 30 years since its introduction what has been all too conspicuous by its absence is a full formal definition of admissibility and its degrees. Herein a family of definitions—all in terms of screening off by chance—that capture much that has been agreed about admissibility are proposed and evaluated; one of which is ultimately found to be serviceable as a definition schema for relative admissibility and its degrees.