What Is a Context?

The terms ‘practice’ and ‘context’ are routinely used in attempts to characterize central strands in Wittgenstein’s later discussions of language. Less often is a potential tension between the two notions observed. Whereas ‘practice’ involves the idea of repetition, of doing the same thing over and over again, sensitivity to context may be taken to involve a fragmentation of language use. So, if language is pervasively context-sensitive, then how can it exhibit the sort of continuity essential to a practice? To clarify this issue, I revisit Herman Cappelen’s and Ernie Lepore’s attempt to use the phenomenon of indirect homophonic reports to argue against what they see as radical contextualism’s hopelessly disintegrated conception of language use. I argue that Cappelen’s and Lepore’s argument fails and that it does so for very interesting reasons. In particular, I show that their criticism of so-called context-shifting arguments is misguided. An upshot of my discussion is a rejection of what I call an ‘archipelago’ conception of linguistic context. Once this conception is rejected, we can see that radical contextualism doesn’t lead to a disintegrated picture of language use but is in fact fully compatible with the sort of continuity required for language to constitute a practice.