论基于记忆资源的依存局域理论及潜在问题——来自汉语关系从句的挑战

Issues on Gibson's dependency locality theory:Challenges from Mandarin relative clauses

本文旨在探讨依存局域理论在应用到汉语关系从句加工时所遇到的问题,即它所预测的加工困难与实证研究的结果并不一致。通过具体剖析汉语主语、宾语关系从句加工难易的分歧、主语关系从句中是否蕴含前置限定词的难易对比、关系从句中是否蕴含"被"的难易对比等情况,本文试图揭示该理论尚不全面,仍有改进余地。

This paper aims to examine questions arising from Gibson's dependency locality theory (DLT) when it is applied to relative clause (RC) processing in Mandarin,specifically,the inconsistencies between its predictions and the data from empirical research.Three issues are discussed:(a) the controversy on the processing asymmetry between subject-extracted and object-extracted RCs in Mandarin;(b) the loci of processing difficulty in subject-extracted RCs that contain pre-RC classifiers;(c) the loci of processing difficulty in object-extracted RCs that contain a passive marker bei (被).It shows that Gibon's DLT as a resource-based complexity metric measured by linear distance and incomplete heads cannot fully explain the Chinese data.