In the first section of this paper I present a question that needs to be
addressed by proponents of interaction oriented conceptions of language, but
which is seldom acknowledged (let alone answered). The question is this: How
should the boundary between linguistic and non-linguistic behavior be
delineated? In the second part of the paper I appeal to Davidson's notion of
triangulation (as applied in his account of the underpinnings of language) in
order to answer this question. In particular, I argue that the notion of
correspondence between language and world (even without facts to which true
sentences correspond) can play a part in the solution of this problem.