The debate concerning epistemic contextualism represents a kind of linguistic turn in epistemology, where the focus has shifted from theorising about knowledge to theorising about knowledge attributions. Such a shift may well prove valuable, but only if we are clear on what the relationship is between a semantic analysis of knowledge attributions and a philosophical analysis of knowledge. One plausible approach is to claim that the semantic analysis entails and is entailed by the philosophical analysis. Yet this view - referred to here as the default view - has been explicitly adopted by few in the contextualism debate. This paper considers a form of argument in favour of the default view, and then considers the challenges that arise from either accepting or rejecting the default view.