Chapter 6. Bringing together fragments and constructions
In this article, I argue that Thompson’s (2002) fragment analysis of (finite) complementation in English, while insightful, cannot be plausibly duplicated for all complementation strategies in English and Spanish. Specifically, this analysis cannot account for a number of semantico-pragmatic restrictions on the NP and the XPCOMP in secondary predication in English and Spanish. Moreover, the Thompsonian analysis alone cannot capture the grammatical relations holding between clitics and matrix verbs on the one hand, and matrix verbs and their XPCOMPs on the other hand, in dialogic Spanish data involving right or left dislocation. The Goldbergian constructional analysis of secondary predication in English and Spanish presented here under the rubric of a family of subjective-transitive constructions draws on the interaction of specific, locally-bound fragments as well as more abstract constructions, thus corroborating the need to reconcile usage and morpho-syntactic facts.