The sentential divide in language and cognition

Some linguists have argued that sentences should not be studied in isolation. They argue, rather, that the structure of sentences is largely the result of constraints imposed upon them by the discourses they are embedded in. I want to argue that this approach is misguided and that sentence-level syntax and discourse structure constitute distinct domains of study, at least in part because grammar is underdetermined by function. Moreover, I argue that discourse and sentence structures illustrate two types of cognition, dynamic vs. static, and thai these necessarily involve different theoretical constructs for their explanation. That is, they constitute distinct epistemological domains. An important conclu-sion of this study is that principles of sentence grammar cannot be discovered by studies of discourse, but only by careful, theory-guided examination of individual sentences. Because it draws on multiple cognitive domains, discourse itself is inherently resistant to analysis by any single theoretical framework or discipline. Sentence grammar, on the other hand, is amenable to study by a single discipline, i.e., linguistics. These ideas are explored partially here via a review of the book Pragmatics of Word Order Flexibility. It is argued that the studies in this collection have little light to shed on syntactic studies, other than the fact that discourse can exploit independently available syntactic structures.