2010 Fiscal Year Final Research Report
Experimental Study on Interaction of Syntactic, Prosodic, and Working Memory Constrains in Auditory Comprehension of Reanalysis Sentences
| Project/Area Number |
20520396
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Linguistics
|
|---|
| Research Institution | Mejiro University |
|---|
Principal Investigator |
TOKIMOTO Shingo Mejiro University, 外国語学部, 教授 (00291849)
|
|---|
| Project Period (FY) |
2008 – 2010
|
| Keywords | ガーデンパス / パラ言語 / リーディングスパンテスト / ワーキングメモリ / 統語解析 |
|---|
| Research Abstract |
This study experimentally examined the possible interactions between syntactic, prosodic, and working memory constraints, to construct a sentence processing model plausible in real time.
Several types of Japanese reanalysis sentences were auditorily presented with syntactic and prosodic properties manipulated. Their comprehension accuracy was analyzed in relation to the working memory capacity of the participants estimated by Japanese Reading Span Test. Our results indicated that (1) prosodic constraints were not so strong as to cancel the effect of syntactic structure, that (2) prosodic effect was stronger in reanalyses with greater processing load, and that (3) the effect of working memory constraints was more salient in less costly reranalses. Further, a speaker with a high score in the Reading Span Test comprehended the sentences with costly reanalyses more accurately and spent longer time than one with a low score. This suggests that the efficiency generally assumed for the speakers with great (Japanese) Reading Span Scores does not necessarily imply the rapidity in reanalyses. It is possible that mental resources are assigned differently depending on the working memory capacity of a speaker.
|
