On the Status of Binding Out of Nominals (by Daniel Plesniak)
언어학 제 99 호 (2024. 8. 30: 77-116), 사단법인 한국언어학회
* Seoul National University. Linguistics. Post-doctoral researcher (dhplesniak@gmail.com)
Abstract
A persistent challenge for c-command-based accounts of bound variable anaphora has been that elements within nominals sometimes seem to “bind out” of them, as in the case of “possessor binding” (a.k.a. “spec-binding”). While these phenomena are much discussed in the literature, certain details about them have rarely been noted, most crucially that the acceptability of bound variable readings in such cases are subject to a great deal of variation, which is conditioned on the individual in question, the particular lexical items involved, and more. Recent work suggests that this variation is not random, but implicates the effects of multiple distinct “sources” of interpretation, ranging in nature from structural to linear precedence-based to semantic/pragmatic, each of which has its own clearly detectable signature. In this paper, using the judgements of myself and two other researchers, the effects of these sources are teased apart and the implications thereof considered. The results show c-command does indeed play a role, but as one of many factors that determine bound variable reading acceptability. Crucially, binding out is unavailable when binding is based purely on c-command, but is available when based on other sources.
One Response
I feel that footnote 19 on page 104 (page 24 /40) is very important and needs to be read first before Japanese speakers try to judge the sentences in Japanese.