Abstract

This paper contains a development of the consequences of a form of skepticism closely akin to traditional fallibilism. It is contended that fallibilism properly understood is compatible with actual possession of knowledge and with rationally continuing claims to such possession. In order to justify this contention, the notion of a K—game as a species of game played in accordance with rules is developed, first leaning on Brians Skyrms' notion of a rational dialectic and then on independent grounds. The notion of knowledge which emerges from this is compatible with some but not all of the analyses of knowledge which have recently been put forward.