Abstract

Any good theory of knowledge ascriptions should explain and predict our judgments about their felicity. I argue that any such explanation must take into account a distinction between three ways of using knowledge ascriptions: to suggest acceptance of the embedded proposition, to explain or predict a subject's behavior or attitudes, or to understand the relation of knowledge as such. The contextual effects on our judgments about felicity systematically differ between these three types of uses. Using such a distinction is, in principle, open to both contextualist and pragmatic invariantist accounts of knowledge ascriptions. However, there are some implications pertaining to the use of the “method of cases” in the debate about knowledge ascriptions.