Abstract

It is all but universally accepted in epistemology that knowledge is factive: S knows that p only if p. The purpose of this thesis is to present an argument against the factivity of knowledge and in doing so develop a non-factive approach to the analysis of knowledge. The argument against factivity presented here rests largely on empirical evidence, especially extant research into visuomotor noise, which suggests that the beliefs that guide everyday motor action are not strictly true. However, as we still want to attribute knowledge on the basis of successful motor action, I argue that the best option is to replace factivity with a weaker constraint on knowledge, one on which certain false beliefs might still be known. In defence of this point, I develop the non-factive analysis of knowledge, which demonstrates that a non-factive constraint might do the same theoretical work as factivity.