In common with many other "virtue epistemologists," Abrol Fairweather and Carlos Montemayor contend that in order to count as knowledge, a mental state must be the product of truth-apt dispositions. I question their theoretical motivations. First, I note that unlike virtue ethics, affect is irrelevant to knowledge. A generous act is arguably better if it is performed warm-heartedly, but a belief is no more creditable if it is performed with the right affect. Second, I argue that non-discursive skills are better cases for virtue epistemologists than cognitive reliability—but because it is hard to know how skills work in individual instances of belief-formation, they cannot figure in the accreditation of knowledge. Finally, I argue that the o...