Download Full PDF

郑博心

北京外国语大学，中国

 

摘要

本研究基于大规模语料库，采用多重对应分析方法，对汉语本族语者及低、高水平汉语学习者在“使”字句与“把”字句选用上的原型使用特征及其习得路径进行了量化考察。研究发现：（ 1）在汉语本族语中，两类构式原型使用特征呈显著的对立分布模式。“使”字句原型使用特征为无生致事、事件致事、状动形结果谓语、事件使事、无生使事以及动词使事，“把”字句原型使用特征为有生致事、具体物致事、动补形结果谓语、有生使事、具体物使事以及代词使事；（ 2）汉语水平学习者与本族语者高度依赖致事相关特征选用构式，但低水平学习者的语言知识框架较为简化，对第二维度原型使用特征敏感度不足 ；进入高水平阶段则开始浮现出多维特征，汉语学习者的语言知识框架向本族语基准靠拢。据此，本研究认为应采用“核心—边缘”层级教学模式，建议两类构式教学中应优先教授致事语义类、致事有生性等核心维度特征，进而逐步引入使事、结果谓语等补充维度特征，以帮助学习者构建多因素权衡的构式知识系统。研究在一定程度明晰两类构式的多维特征差异，也为帮助汉语学习者构建多因素权衡的语言知识框架提供了实证依据与教学路径。

 

关键词

第二语言习得，多重对应分析，多变量统计建模

 

Multivariate Statistical Modeling of Prototype Usage Features in ba and shi among Learners of Chinese as a Second Language

 

Boxin Zheng

Beijing Foreign Studies University, China

 

Abstract

Drawing on a large-scale corpus, this study employs Multiple Correspondence Analysis (MCA) to quantitatively examine prototype usage features and acquisition trajectories in the choice between shi and ba among native speakers of Mandarin Chinese and learners of Chinese as a second language(CSL) at lower and higher proficiency levels. The results show that: (1) prototype usage features among native speakers of Chinese show significant contrasting distribution pattern. The prototype usage features of shi include inanimate causer, eventive causer, adverbialverbal predicate, eventive causee, inanimate causee, and verbal causee; whereas the prototype usage features of ba include animate causer, concrete causer, verb-complement predicate, animate causee, concrete causee, and pronominal causee. (2) Both CSL learners and native speakers highly depend on causer-related features for construction selection. However, lower proficiency CSL learners operate with a simplified linguistic knowledge schema and show insufficient sensitivity to prototype usage features on the second dimension. At higher proficiency levels, multidimensional prototype usage features gradually emerge, and learners’ linguistic knowledge system converge toward the native-speaker baseline. Based on these findings, the study proposes a “core–periphery” hierarchical teaching strategy: instruction should prioritize core features such as CauserSem and CauserAni and then progressively introduce additional dimensions (e.g., the prototype usage feature related causee or predicate) to support CSL learners in building a constructional knowledge system that integrates and weights multiple interacting factors. This study clarifies the multidimensional feature differences between the two constructions and provides empirical evidence and pedagogical strategies to support CSL learners in building a linguistic knowledge schema that integrates and weights multiple interacting factors.

 

Keywords

Second language acquisition, multiple correspondence analysis, multivariate statistical modeling