The grammar of Q : Q-particles and the nature of Wh-fronting, as revealed by the Wh-questions of Tlingit

Abstract

The central claim of this thesis is that the agent responsible for a variety of phenomena surrounding wh-operators is not those operators themselves, but rather a distinct element that we label a 'Q(uestion)-particle'. In many languages, the Q-particle is phonologically empty, and so its role in various phenomena has not yet been recognized. Most importantly, careful study of these Q-particles reveals that the phenomenon known as 'pied-piping' does not exist, and that all putative examples of it are actually instances of normal phrasal movement of the Q-particle. This thesis starts from the demonstration that wh-fronting in Tlingit (Na-Dene; Alaska, British Columbia, Yukon) does not involve a syntactic relationship between the interrogative C and the wh-word. Rather, it involves a probe/Agree relation between C and an overt Q-particle c-commanding the wh-word. Fronting of the wh-word in Tlingit wh-questions is a mere by-product of fronting the projection of the Q-particle. From this core observation, a syntax and semantics for Tlingit wh-questions is developed. Given the strong similarity between the wh-constructions of Tlingit and those of more widely studied languages, the analysis developed for Tlingit is then applied to a range of other languages. It is found that such a 'Q-based' theory of wh-constructions holds a variety of analytic consequences. Regarding so-called 'pied-piping structures', the Q-based theory provides an analysis of such structures where the very concept of 'pied-piping' is eliminated from the theory of grammar. Furthermore, the Q-based theory provides a semantics for wh-questions that correctly interprets pied-piping structures without recourse to any mechanisms beyond those needed for wh-questions without pied-piping.

(cont.) Finally, the Q-based theory accounts for various constraints on pied-piping, and correctly predicts the scope and limits of its variation across languages. Beyond its treatment of pied-piping, the Q-based theory also provides a novel syntax and semantics for multiple wh-questions, which successfully ties the presence of Superiority Effects to the absence of Intervention Effects, and which correctly predicts a previously unnoticed Intervention Effect in English. Moreover, it provides a novel, unified account of the ill-formedness of left branch extractions, as well as of preposition stranding.