How peripheral a phenomenon is it?: On the interaction of DP-internal ellipsis, degree inversion, and islands

Thesis (Ph.D.)--University of Washington, 2016-06

In this dissertation I set out to solve a series of puzzles related to the notion of a DP periphery, defined as an area around the edge of a given domain targeted by operations such as movement and agreement. In solving these puzzles, I argue for a peripheral area in the nominal domain. Early arguments for a peripheral boundary in the nominal domain are found in the justifications for a functional D as the head of the noun phrase (Abney, 1987; Szabolcsi, 1987, 1994; Horrocks & Stavrou, 1985). More recently, the nominal domain has been argued to contain a peripheral area on the basis of CP/DP parallelisms (Giusti, 1996; Aboh, 2004). Here I am minimally concerned with proposing functional projections as parallels to other domains; rather, I make use of DP-particular projections -- such as DP and NumP -- to account for the three puzzles I present. I present the first puzzle in chapter 2: the licensing conditions for nominal ellipsis and nominal gapping in English and in French (Lobeck, 1995; Sleeman, 1996; Jackendoff, 1971; Yoshida, Wang, & Potter, 2012). I argue for a unified analysis of these phenomena, the mechanics of which serve as the functional basis for the remaining arguments in this dissertation. I propose a system of feature agreement, which I call Mutual Agree, in which nominal ellipsis is licensed when a higher, licensing head agrees with a lower, ellipsis head. Mutual Agree not only accounts for all instances of nominal ellipsis and gapping in English and French, but it also restricts the ellipsis mechanism to only license ellipsis in appropriate contexts. The second puzzle I consider (chapter 3) is the functional structure of phrases that undergo degree inversion (Matushansky, 2002; Den Dikken, 2006; Troseth, 2009). The main goals of the chapter are twofold. First I argue for an analysis of degree inversions that derives the interpretive parallels between simple degree inversion constructions and those with complex post-degree structure. I follow Troseth's (2009) proposal that degree inversions are derived via predicate inversion and then propose that the degree phrase predicates of these inversion constructions are complex, relative-like clauses. Second I argue that the variable ordering patterns found in degree inversions can be given a simple syntactic account if we assume that some degree words can take a null complement as opposed to an overt adjectival complement. The features that are involved in Mutual Agree also play a role in degree inversion; I specifically argue that the feature of contrastive focus is what causes the movement of a predicative degree phrase out of the small clause configuration and to a prenominal position. The final puzzle I present (chapter 4) explores the possibility of island `repair' under nominal ellipsis (for discussions of repair in the clausal domain, see Ross, 1967, and Merchant, 2001). I first show that the system of Mutual Agree accounts for why certain elements, but not others, may escape nominal ellipsis in order to be candidates for island repair. I demonstrate that apparent repair of DP-internal island violations occurs in certain nominal ellipsis scenarios. I argue that this occasional repair is due to the fact that repair is only an illusion (Barros, Elliot, & Thoms, 2014, 2015). Repair effects are the result of the ellipsis site potentially containing a non-island-violating construction; this ambiguity results in acceptance of the ellipsis-containing structure. I conclude in chapter 5 by returning to what the three preceding chapters reveal about the function of projections in DP. I argue that the combination of (i) evidence from ellipses, (ii) evidence from ellipsis escape, and (iii) evidence from non-elliptical movement reveal what appears to be an articulated, albeit small, periphery of sorts.