Antecedent Contained Deletion in a Variable-Free Semantics

Abstract

No abstract.

Free and Bound Pro-Verbs: A Unified Treatment of Anaphora

“…This is in decided contrast to pro-form accounts (e.g. Hardt (1993), Jacobson (1992a) Schwarz (2000, Szabolcsi (1992)) in which the elided constituent is not explicitly instantiated at any level of syntactic representation.…”

“…10 Following Schwarz (2000), Jacobson (1992a), Szabolcsi (1992, and Hardt (1993), I propose a pro-form account of VE in the Combinatory Categorial Grammar (CCG) tradition-cf. Steedman (2000), Jacobson (1999), Barker (2005), and many others.…”

MaxElide and the Re-binding Problem

“…In order to differentiate Re-binding from other cases, we have made crucial use of variables and variable names. However, variables and variable names do not play any role in variable-free semantics (Jacobson 1992(Jacobson , 1999. Specifically, a pronoun does not make any contribution to the meaning of an elided VP in sloppy identity construction such as (36).…”

“…The prediction that Re-binding is never possible fo llows from the simple parallelism condition introduced above (together with the assumption mentioned in endnote 1, or (26}), and Sag and Williams support it by a variety of empirical observations. Counter-evidence, however, has been accumulated over the years (Evans 1988, Fiengo and May 1994, Jacobson 1992, Merchant to appear, Schuyler 2001, among others). We argue that the empirical discrepancy in the past literature results from the fact that Re-binding is allowed, but is constrained.…”

Processing of ACD gives no evidence for QR

“…An alternative is the Transitive Verb (Phrase) (TVP) ellipsis analysis, which is often embedded in a Categorial Grammar (see Cormack 1984, Jacobson 1992a, 1992b, 2003. By way of background, consider a non-ellipsis case like (2): (2) Sarah will read every newspaper that Katie will read.…”