ISSN 1671-3710
CN 11-4766/R
主办:中国科学院心理研究所
出版:科学出版社

• •    

二语词汇表征从“寄生”到独立的发展机制:基于联结主义框架的模型对比分析

李子怡, 卢天伊, 药盼盼   

  1. 北京语言大学北京语言大学心理与认知科学学院, 北京 100080 中国
  • 收稿日期:2025-10-24 修回日期:2026-06-29 接受日期:2026-07-10
  • 基金资助:
    北京语言大学校级项目资助(中央高校基本科研业务费专项资金)(24QN42)

The Developmental Mechanism of L2 Lexical Representation from “Parasitic” to Independent: A Comparative Analysis of Connectionist Models

Li Ziyi, Lu Tianyi, Yao Panpan   

  1. School of Psychology and Cognitive Science, Beijing Language and Culture University 100080, China
  • Received:2025-10-24 Revised:2026-06-29 Accepted:2026-07-10

摘要: 在第二语言习得领域,一个核心的问题是二语(L2)词汇表征如何从依赖母语(L1)的“寄生”状态发展为相对独立的表征系统。为了深入探讨这一“寄生→独立”发展过程的认知机制,本文总结了L2词汇习得过程中的经典现象和影响因素,并系统对比分析了基于联结主义框架的六个关于二语词汇习得的模型对这些现象的解释及其局限性。具体来讲,本文从词汇竞争识别(双语交互激活模型)、发展维度(词汇固化理论、模糊词汇表征假说、词汇个体发展模型)、计算建模(发展性词库模型-II)、以及神经基础(互补学习理论)这四个角度,由表及里地探讨二语词汇表征从“寄生”到独立这一发展过程中各经典现象背后可能的认知机制。同时,鉴于中文与拼音文字在语言文字系统上的巨大差异,本文呼吁未来在汉英/英汉双语领域开展更多实证研究,并发展针对性的理论模型,以检验和拓展现有模型的跨语言适用性。此外,未来研究还需要考察语言之外的认知因素、融入发展框架、使用计算机建模等技术以期深化对二语词汇习得认知机制的理解。

关键词: 二语词汇发展机制, “寄生”表征, 词汇表征独立性, 联结主义框架

Abstract: In the field of second language acquisition, a central question concerns how L2 lexical representations develop from an initial “parasitic” state dependent on the first language (L1) into a relatively autonomous representational system. To further elucidate the cognitive mechanisms underlying this developmental trajectory—from parasitic to independent representations—the article summarizes classic phenomena and influencing factors in L2 lexical acquisition and provides a systematical analysis of six connectionist models that account for these phenomena, highlighting both their explanatory strengths and limitations. Specifically, focusing on the developmental shift in L2 lexical representation from parasitic to independent patterns, we examine this process from four perspectives: lexical competition and recognition (Bilingual Interactive Activation model), developmental dimensions (Lexical Entrenchment Account, Fuzzy Lexical Representations hypothesis, and Ontogenesis Model of the L2 Lexical Representation), computational modeling (Developmental Lexicon II), and neurocognitive foundations (Complementary Learning Systems). These four perspectives progress from surface-level processing to deeper underlying mechanisms. Meanwhile, given the substantial differences between Chinese and alphabetic writing systems, this review further calls for more empirical research in Chinese–English/English–Chinese bilingual contexts and the development of tailored theoretical models to test and extend the cross-linguistic applicability of existing frameworks. Furthermore, future research should also examine cognitive factors beyond language, incorporate developmental frameworks, and employ computational modeling techniques to deepen the understanding of the cognitive mechanisms underlying L2 lexical acquisition.