The syntax of wh- and polar questions in Marshallese

In this article, we are concerned with the syntactic structure of questions in Marshallese. We explore both whand polar questions in turn: Section 2 is dedicated to wh-questions, where, applying diagnostics suggested by Adger and Ramchand (2005, henceforth AR), we conclude that a base generation analysis is more empirically adequate for Marshallese than a movement account. In Section 3 we discuss polar questions with the particle ke, first considering an analysis offered by Willson (2005), and concluding that, despite accounting for the grammatical polar questions, the analysis overgenerates and thus is not descriptively adequate. We then propose our own analysis of ke-questions which builds on crucial insights from Willson’s work while also blocking ungrammatical derivations. Overall we find that polar questions are more syntactically complex than wh-questions, which in turn are more syntactically complex than, but built on a similar structure as, clefts; forming, in some sense, a hierarchy of syntactic complexity1. Finally, in Section 4 we briefly discuss an additional puzzle pertaining to the wh-adjuncts ia and ewi.