习语表征混合观研究二十年: 回顾与展望

A Hybrid View on Idiom Representation over Two Decades:Review and Prospect

习语是整体表征还是可以拆分?心理语言学界对此各执一词,迄今未有定论。习语表征混合观(hybrid view)尝试综合考量习语的结构双重性,为解决整体观与可分观的矛盾提供了新思路。本文首先回顾了习语混合表征理论近二十年的发展历程,然后着重梳理介绍最具影响力的超词条(superlemma)混合模型及其在一语和二语研究领域内的多维度应用,最后通过对比分析相关研究,探讨其优势、不足以及未来发展前景。

Psycholinguistic researchers have not reached any consensus on whether idioms are compositional or represented holistically in the mental lexicon.In an effort to embrace the duality of idiom structure,the hybrid view on idiom representation proposes that at different levels of processing,idioms are both holistic units containing single lexical concepts,and compositional units consisting of independently occurring words.Such a view reorganizes the internal framework of idiom representation system and sheds fresh light on the“holistic or compositional”dilemma.This paper first presents an overview of the development of idiom hybrid theories over the past two decades.Then,it focuses on introducing the most influential hybrid model(i.e.,the superlemma model)and its multidimensional application in L1 and L2 research.By comparison and contrast of relevant studies,the paper points out some major advantages of the superlemma model in terms of its valuable theoretical insights,solid foundation for model construction,and potentially broad applicability.Nevertheless,it also argues that both the theoretical and applied dimensions of the hybrid view are yet insufficiently developed.Hence,the paper provides suggestions for future studies,including further enriching the contents of hybrid theories,stimulating discussion on compositionality and syntactic constraints of idioms,as well as extending current research into the fields of typology and second language acquisition.