Empirical content and rational constraint

It is often thought that epistemic relations between experience and belief make it possible for our beliefs to be about or "directed towards" the empirical world. I focus on an influential attempt by John McDowell to defend a view along these lines. According to McDowell, unless experiences are the sorts of things that can be our reasons for holding beliefs, our beliefs would not be "answerable" to the facts they purportedly represent, and so would lack all empirical content. I argue that there is no intelligible conception of what it is for beliefs to be answerable to the facts that supports McDowell's claim that our empirical beliefs must be justified by experience.