中文「著」的第二語言習得：語料庫研究

Abstract

本研究檢視了中文時貌標記「著」的句法及語義特性，並針對成人二語學習者如何習得「著」進行探討。本研究使用的語料來自中央研究院平衡語料庫及國立台灣師範大學國語教學中心經由線上測驗系統而建立的中介語語料庫。首先，我們把「著」的用法分為三類：「V著」、「V1 著V2」、及「V1 著V1 著（就）V2」，再將這三類分為次類，以此分析比較中介語語料庫以及中央研究院平衡語料庫的語料。本研究發現有三：（1）高級學習者使用各種「著」的情形與母語者最相似。（2）學習者程度愈高，「V 著」的使用比率也隨之增加。（3）學習者程度愈高，「V1 著V1 著（就）V2」的使用比率亦隨之減少。本研究亦檢視了學習者使用「著」時的偏誤，此偏誤分析使我們對「著」的二語習得更加了解。本研究建議，「著」的二語教學必須顧及「著」的句法、語義以及與「著」同現的動詞間的關係。

This paper examines the semantic and syntactic properties of the aspect marker -zhe in Mandarin Chinese as well as L2 adult learners’ acquisition of -zhe using data from the Academia Sinica corpus and a learner corpus compiled by the Mandarin Training Center at National Taiwan Normal University. -Zhe sentences were categorized into three major patterns — V-zhe, V1-zhe-V2, and V1-zhe-V1-zhe-(jiu-)V2, with each further divided into sub-types. Major findings were: (a) Usage of the various types of -zhe produced by the highest proficiency level learners in the corpus was closest to that of native speakers’. (b) The rate of V-zhe occurrence increased with the learners’ proficiency level. (c) The rate of V1-zhe-V1-zhe-(jiu-)V2 occurrence decreased with the learners’ proficiency level. Error analysis revealed additional insights. These findings suggest that L2 teaching of Mandarin must take the various patterns of -zhe and the semantic and syntactic properties of the verbs occurring with -zhe into consideration.