A Causal Exclusion Problem for Knowledge-First Epistemology

According to knowledge-first epistemology, knowledge is irreducible to true belief plus something else, and yet knowledge entails belief. Moreover, knowledge plays an ineliminable role in causal explanation of action. This paper develops and defends an entirely new line of reasoning against this view: in conjunction with two independently plausible principles about explanatory realism and causal exclusion, this view leads to the conclusion that beliefs are epiphenomenal in cases of knowledge, which is when and only when action is deemed epistemically rational. Even worse, if a causal criterion of ontological commitment is accepted, it follows that postulating the existence of beliefs in cases of knowledge is redundant. Several specific objections are raised, followed by replies. While the paper does not pretend to exhaust all lines of resistance, the upshot is that the burden is now squarely on friends of knowledge-first epistemology to pursue one or more of these further.