In joint work (Giannakidou 2004, Etxeberria 2005, Etxeberria and Giannakidou 2010), we argue that one of the functions a D head can perform in language is to provide domain restriction. For this, we propose a new mode of composition for D--DDR. DDR is a type-shifting of D from its "regular" individual yielding function (iota) to a modifier function supplying the context set variable C. In Greek, St'át'imcets Salish, and Basque we find DDR applying to Q, but in St'át'imcets, the D-head shifts to DDR also when it combines with the NP under Q. Our analysis relies on Salish D being a definiteness head, since C is a discourse salient or familiar property; we thus challenge Matthewson's (1998, 2001) treatment of St'át'imcets Salish D as indefinite. Our goal here is to consider the St'át'imcets data and show that (a) the arguments for indefiniteness of D are not conclusive, and (b) the St'át'imcets facts are fully consistent with an analysis of DDR as a manifestation of definiteness. The St'át'imcets D, we will suggest, is best analyzed as a deictic, demonstrative-like D head, consistent with its morphology.