Processing NP trace and reflexives in Chinese.
Item
-
Title
-
Processing NP trace and reflexives in Chinese.
-
Identifier
-
AAI9720153
-
identifier
-
9720153
-
Creator
-
Xie, Jiaye.
-
Contributor
-
Adviser: Janet Dean Fodor
-
Date
-
1997
-
Language
-
English
-
Publisher
-
City University of New York.
-
Subject
-
Language, Linguistics
-
Abstract
-
This dissertation reports a research project on the on-line processing of five constructions in Chinese: "bei" passive, "shi" passive, copular construction, topicalization, long distance reflexives. The goal was to investigate the processing of binding relationships between dependent elements such as NP-trace, WH-trace and overt anaphors such as "ziji", and their antecedents. The primary purpose was to explore the psychological reality of NP-trace.;The existence of phonologically empty categories, such as NP-trace and WH-trace, in natural language has been a matter of debate in linguistic studies. Government-Binding theory postulates four types of empty categories, while nontransformational grammars such as Head Driven Phrase Structure Grammar, which takes maximum advantage of lexical structures, rejects empty categories. This is a controversy which is difficult to resolve by theoretical studies. Therefore, experimental data on sentence processing can be helpful. Taking the transparency hypothesis as a working assumption, psycholinguists have recently made efforts to address this issue.;In previous experiments on English, stronger evidence was found for WH-trace than for NP-trace. Nicol (1989) attributed the lack of strong on-line effects for NP-trace to the temporary structural ambiguity in English passives, which obscures the binding relationship. Because the passive verb is ambiguous and the antecedent NP is in an A(rgument) position (surface subject), the parser cannot tell that it was moved to that position, so it cannot predict the upcoming trace, as it can in the case of WH-trace.;Chinese offers an unambiguous passive structure on which this hypothesis can be tested. In the "bei" passives, the passive preposition "bei" (by) precedes the agentive NP, the verb and the following trace, and unambiguously cues the parser that there will be a trace after the verb. Two cross-modal experiments, one using a naming task and the other a lexical decision task, were conducted on the processing of "bei" passives and the other four Chinese structures listed above.;Neither experiment showed any evidence for NP-trace. In Experiment I, all results were as predicted except for the "bei" passives. The data exhibited a significant antecedent-activation effect for "ziji", but no effect for the copular sentence, which contains no trace, or the "shi" passive, which is ambiguous. Experiment II found, as predicted, no effect of the WH-trace in topicalization, which is also an ambiguous structure in Chinese. The data showed no priming in the "bei" passive, but a reverse effect for the overt anaphor "ziji". Some possible explanations for the latter finding are proposed. Though the overall pattern of results is not fully decisive, the consistent absence of priming in the "bei" passives challenges the existence of NP-trace.
-
Type
-
dissertation
-
Source
-
PQT Legacy CUNY.xlsx
-
degree
-
Ph.D.
