国内外近十年三语反向迁移研究综述

A Decade of Research on L3 Backward Transfer: A Review of Domestic and International Studies

摘要: 随着多语现象的日益普及，跨语言影响研究已逐渐打破单向局限，将视野扩展至第三语言对先习得语言的反向迁移领域。本文对近十年国内外核心数据库知网(CNKI)与Web of Science收录的21篇三语反向迁移文献进行了系统梳理，从研究方法与研究内容两方面系统整理该领域的发展现状。文献回顾表明，现有研究在一定程度上呈现出以下局限：一是在研究维度上存在一定失衡；二是在反向迁移的稳定性及效应性质等核心理论问题上仍存较大争议；三是形态句法等领域的考察倾向于依赖离线测试与书面语料，导致研究的生态效度受限。基于此，本文将对当前研究中存在的不足进行述评并对其未来研究方向提出相关的思考。

Abstract: With the increasing prevalence of multilingualism, research on cross-linguistic influence has gradually broken through unilateral limitations, expanding its scope to the field of backward transfer (or reverse/regressive transfer) from a third language (L3) to previously acquired languages. This paper systematically reviews 21 articles on L3 regressive transfer indexed in core domestic and international databases (CNKI and Web of Science) over the past decade, outlining the current developmental status of this field from the perspectives of research methodology and research content. The literature review indicates that existing research still faces several limitations: first, there is an imbalance in research dimensions; second, empirical findings regarding theoretical verification across different levels remain controversial; third, investigations in domains such as morphosyntax rely heavily on offline testing and written corpora, which limits the ecological validity of the studies. Based on these findings, this paper provides a critical review of the shortcomings in current research and offers insights into future research directions.