Why indefinites can escape scope islands

Abstract

One of the big questions about indefinites is why they can escape scope islands (Fodor and Sag, in Linguist Philos 5:355–398, 1982). In the recent approach of Brasoveanu and Farkas (Linguist Philos 34(1):1–55, 2011) scopal relations with syntactically dominating quantifiers are hard wired into the semantic definition of the existential quantifier, which immediately explains why the semantic scope of indefinites may exceed their syntactic scope. In this paper, I argue for the revival of an alternative approach which places the explanatory burden on the idea that indefinites are essentially referential expressions, similar to definites, and not plain existential quantifiers. I propose one fully explicit variant of such theories and argue that it comes with a number of conceptual and empirical advantages over competing theories.