In Part II of the book we looked at “problems for everyone.” That is, we looked at some of the perennial problems in epistemology – problems that any adequate theory of knowledge should say something about. The argument developed there was that the present account of epistemic normativity helps us to make progress on those problems. In Part III of the book we turn to “problems for reliabilism.” These are problems that have persistently dogged reliabilist theories of knowledge despite the obvious advantages of those views. Here I will argue that, once again, the present account of epistemic normativity can help. Hence this third part of the book develops one of the major themes from Chapter 1: that a virtue-theoretic account of epistemic normativity is tailor-made for reliabilism.

THE PROBLEM OF STRANGE AND FLEETING PROCESSES

Reliabilist theories have long been plagued by counterexamples ­involving strange and fleeting cognitive processes. The idea is that if a cognitive pro­cess is either strange enough or fleeting enough, then it will not give rise to knowledge even if it is reliable. For an example of a strange cognitive pro­cess, consider the case of the Serendipitous Brain Lesion. Suppose that S has a rare brain lesion, one effect of which is to reliably cause the true belief that one has a brain lesion. Even if the process is perfectly ­reliable, it seems wrong that one can come to have knowledge that one has a brain lesion on this basis.