Use of Discourse Cues During Garden-Path Resolution is Modulated by Verb Argument Structure

Loading...
Thumbnail Image

Embargo Date

Degree type

Discipline

Subject

Funder

Grant number

License

Copyright date

Distributor

Related resources

Contributor

Abstract

Studies on garden-path sentences such as While the man hunted the deer ran into the woods have shown that comprehenders face processing difficulties due to the locally ambiguous noun phrase “the deer”. This critical noun phrase tends to be initially interpreted as the object of the preceding verb, but it must ultimately be interpreted as the subject of the following clause. This grammatical role ambiguity is particularly of interest because in English (and other languages) discourse information tends to be packaged in such a way that objects are typically indefinite, new information, and subjects are most often previously-mentioned, definite information (e.g., Comrie 1989, Prince 1992). We hypothesized that, if discourse information is at play, noun phrases that are more subject-like might facilitate garden-path resolution relative to more object-like noun phrases. However, in order to better understand when the discourse level is engaged in processing these constructions, we examined garden-paths with two verb types: reflexive absolute verbs (RATs, e.g., “wash”) and optionally transitive verbs (OPTs, e.g., “hunt”). Due to their reflexive nature, RATs were expected to operate mainly through a structural route (syntax-only). On the other hand, OPT verbs can introduce implicit arguments and are more likely to engage in operations beyond the domain of syntax (e.g., syntax+discourse). In this study, we discuss data from a self-paced reading experiment (see Besserman and Kaiser 2016) to shed light on the syntax/discourse division of labor: we found effects of information status related to subject-/object-hood in the processing of garden-paths with OPT, but not RAT verbs. These findings suggest that engagement of discourse representations was modulated by the verb’s argument structure.

Advisor

Date Range for Data Collection (Start Date)

Date Range for Data Collection (End Date)

Digital Object Identifier

Series name and number

Publication date

2017-01-01

Volume number

Issue number

Publisher

Publisher DOI

Comments

Recommended citation

Collection