Recent epistemology divides theories of knowledge according to their diagnoses of cases of failed knowledge, Gettier cases. Two rival camps have emerged: naturalism and justificationism. Naturalism attributes the failure of knowledge in these cases to the cognizer's failure to stand in a strong natural position vis-à-vis the proposition believed. Justificationism traces the failure to the cognizer's failure to be strongly justified in his belief. My aim is to reconcile these camps by offering a version of naturalism, a reliability theory of (...) knowledge, that conforms to the central justificationist tenets. I argue that proposed reliability theories of knowledge, reliable indication theories, offer no prospect of a reconciliation because they misdiagnose failed knowledge in such a way as to violate a basic justificationist tenet. Proposed versions of justificationism, it turns out, fare no better with this tenet. I offer an alternative reliability theory of knowledge, a reliable process theory, that conforms to the justificationist tenet. (shrink)