New Insights on the Nature of D-linking

Abstract

Two important challenges in the study of weak islands are the lack of an adequate definition of d-linking, and the still poorly understood nature of the principles responsible for the higher acceptability of d-linked constructions. In the present work, I address both of these issues. In particular, I argue that the notion of d-linking is better captured by assuming it to be composed of two primitives, givenness and restrictiveness. To this end, I provide the results of a linguistic acceptability questionnaire that support the soundness of such a conclusion. Following Kluender (1992, 1998), Hofmeister (2007) and Hofmeister and Sag (2010), I also explore the possibility that the higher acceptability of d-linked structures does not ensue from some peculiar syntactic property associated with d-linked constituents, but from general processing mechanisms. I therefore present the results of two additional experiments which aim to test the validity and the limits of this proposal. As these two latter experiments point out that the d-linking of constituents other than left-peripheral elements has a very local effect, I conclude by suggesting that the characteristic properties exhibited by d-linked constituents are dependent on their collocation at the left edge of the clause, and in particular on the high processing load associated with clause-initial positions.