Abstract

In 1967, Alvin Goldman prominently claimed that the traditional JTB analysis is adequate for non-empirical knowledge. Since then, this claim has remained widely unchallenged. In this paper, I show that this claim is false. I provide two examples in which a true belief is a priori justified but epistemically defective such that it does not constitute knowledge. Finally, I submit a novel analysis of a priori knowledge that avoids the Gettier problem. What is particularly important and distinctive about my analysis is that I neither need to make the justification condition so strong that only infallible justification is allowed, nor do I need to explicitly introduce a truth condition.