Th/Ex, Agreement, and Case in Expletive Sentences. (December 2014)

• Record Type: Journal Article
• Authors: Sobin, Nicholas
• Is Part Of: Syntax. Volume 17:Number 4(2014:Dec.)
• Page Start: 385
• Page End: 416
• Publication Date: 2014-12
• Subjects: Grammar, Comparative and general -- Syntax -- Periodicals; 415
• DOI: 10.1111/synt.12021
• Languages: English
• ISSNs: 1368-0005

Abstract

<abstract abstract-type="main" id="synt12021-abs-0001"> <title>Abstract</title> <p>Recent accounts of existential sentences (ES) within the government‐and‐binding/Minimalist Program tradition (Chomsky 2000, 2001; Deal 2009) claim that nominative Case valuation and ϕ‐agreement take place in ES (directly or indirectly) between T and the associate DP, essentially the same as between T and "subject" in non‐ES. However, the surface facts of Case and ϕ‐agreement in ES run contrary to such claims (e.g., <italic>There is only me in that picture</italic>/*<italic>There am only I in that picture</italic>;<italic> There is/?*are a boy and a girl at the door</italic>). Furthermore, such accounts leave unexplained the short‐movement phenomenon (including Milsark's [1974] "Leftmost <italic>be</italic> Condition") found in ES (e.g., <italic>There is someone being arrested</italic>/<italic>* There is being someone arrested</italic> or <italic>There is a train arriving/*There is arriving a train</italic>). Following claims of Richards & Biberauer (2005) and Deal (2009) that merger of <italic>there</italic> in ES is low, I explore the possibility that "short movement" in ES, as well as the ban on multiple instances of <italic>there</italic>, is accounted for via Extended Projection Principle features of the relevant verbalizing functional heads that are articulated into their Agree and Merge components. Furthermore, an analysis of Case and ϕ‐agreement is presented that accounts directly for the broad range of ϕ‐agreement forms (including "frozen" forms and first‐conjunct agreement), the necessarily accusative Case form of pronominal associates allowable ES, and the ban on ϕ‐agreement with pronominal associates in ES.</p> </abstract>