Deriving null functional heads: a study on variation of functional structure

The aim of this dissertation is to develop a framework of functional structure that ensures that the representation of any phrase is only as large as necessary to capture the syntactic relations relevant to it. I argue that the success of this project requires the elimination of null functional heads from the lexicon. Rather, I propose that null functional heads and their projections are dynamically created during the derivation as an extension of the projection of lexical items (i.e. lexical heads and overt functional heads). To this end, I make two proposals. Firstly, I argue that the featural specifications of lexical items are more extensive and have a more complex structure than previously thought. Secondly, I refine Giorgi and Pianesi’s (1997) Feature Scattering operation so that it applies to entire segments of a featural specification, instead of individual features. One beneficial implication of this formulation is that it reduces head movement to the incidental scattering of the phonological features of a head due to independent syntactic factors. Hence, I present an analysis of a number of cases of head movement in support of the proposed framework of functional structure. Amongst other things, I address V-to-v movement (as in the case of English main verbs), V-to-T movement (as in the case of Romance verbs and English auxiliaries) and V-to-C movement (as in the case of Germanic V2 and English residual V2). Additionally, I extend the analysis to cases of head movement to an initial position, including the movement of the verb in verb-initial clauses and the movement of the noun in noun-initial nominal phrases in Semitic and Celtic languages, which have previously received little attention in the strand of research that adopts Feature Scattering or other similar re-projective mechanisms. Beyond head movement, I develop a uniform analysis of various subject/non-subject asymmetries, including Subject Auxiliary Inversion and do-support in English wh-questions and the that-trace effect in English embedded clauses involving wh-extraction, with the intention to bring the relevant phenomena to bear on the overarching hypothesis that functional structure is variable