A minimalist account of optional wh-movement

Abstract

Wh-movement has been assumed to be a parametrized fact about language, and, thus, whether a language has overt wh-movement or not has been assumed to be invariant within a language. Also, Chomsky's recent Minimalist Program only allows overt movement that is motivated by the presence of a strong feature. Languages with weak wh-features do not have overt wh-movement. Under this version of the theory as well, there should be no instances of optional wh-movement within a language. In this dissertation, however, I show that an Athabaskan language, Babine-Witsuwit'en, has two methods of wh-question formation--leaving the wh-word in situ or fronting it. I argue that this kind of optionality can be successfully explained using Minimalist mechanisms.Languages that have been claimed to exhibit optional wh-movement, or superficially look as if they do, are shown to not have true wh-movement (movement to Spec of CP). Rather, I argue that the fronting is motivated by features other than wh-features and that the wh-phrases move to projections other than CP. I then argue that Babine-Witsuwit'en (and perhaps Ancash Quechua and Malay) do exhibit optional wh-movement by demonstrating how the positions of their wh-phrases differ from those of non-wh-phrases; by ruling out topicalization, focus, and clefting; and by illustrating that the wh-phrases in these languages can violate island configurations. An explanation for this optionality is provided by assuming that, in these languages, C can be optionally selected from the numeration. Interrogative Cs in Babine-Witsuwit'en are shown to motivate movement, but not affect interpretation. A separate typing phrase is shown to carry features which type the clause as interrogative or declarative and to mark scope. I then show how this proposal fits into a broader view of wh-movement by examining languages with wh-movement and languages without, as well as languages with multiple wh-movement. As a consequence of the analysis of optional wh-movement proposed here, my examination of these diverse language types without optional wh-movement provides new insight into the role of wh-movement in general in a Minimalist theory of syntax.