幼儿汉语口语感知特点及神经机制

Speech perception has long been an important issue in psycholinguistic researches. Despite the fact that a number of studies have focused on speech perception in different age groups including infants and adults, only very few of these studies were carried out in preschool-age children. There is still a large gap in terms of our understanding of how proficiency in speech perception developes between infancy and adulthood. Furthermore, most of models of speech perception have been proposed based on the researches of non-tone language and are not fit well for explaining Chinese processing. Chinese is a tone language, in which lexical tone is signaled by pitch variations and associated with spectral processing. Moreover, lexical tone is lexically contrastive and can distinguish lexical meaning just as phonemes are. Here we will investigate the Mandarin speech perception in 3- to 5- year-olds preschool children by combining the method of eye tracking, event-related potential recordings and source estimation (LORETA). In the present project, we will focus on the following issues: (1) how children discriminate Mandarin segmental and suprasegmental information at pre-attentive and attentive stages; (2) What are the roles of segmental and suprasegmental information on Mandarin spoken word recognition in preschool-age children, and whether the memory traces for words will be observed in preschoolers; (3) the neural mechanisms underlying Mandarin speech perception in preschool-age children. The investigations of this project present a valuable opportunity to extend the results from infants and adults, and will provide new experimental evidences for the current models of speech perception that built on the studies of non-tone languages.

口语感知是当前心理语言学的研究热点，然而以往研究大多以婴儿和成人为被试，缺乏对幼儿口语感知的研究。此外，现有口语感知模型主要是基于非声调语言研究建立起来的，对汉语并不完全适用。汉语是一种声调语言，在语音构成上不同于非声调语言。本项目将立足于汉语口语特点，以3-5岁幼儿为研究对象，考察幼儿汉语口语感知特点及神经机制。围绕口语感知的核心问题，我们将综合使用眼动方法、ERP方法和LORETA源定位技术，探讨以下问题：（1）幼儿在前注意阶段和注意阶段的听觉语音辨别特点；（2）幼儿汉语口语词汇识别过程中音段信息和超音段信息的作用；（3）幼儿汉语口语感知的神经机制。本项目研究结果将揭示幼儿汉语口语感知特点，拓展已有的研究发现，为完善现有的口语感知模型提供新的实验证据。同时，本项目研究结果还将为言语损伤儿童、阅读障碍儿童的语音加工诊断和实践训练提供理论指导。

该项目拟以3-5岁幼儿为被试，考察幼儿汉语口语感知特点及神经机制，拓展以往主要以成人和婴儿为被试的口语感知研究，为完善现有口语感知模型提供新证据。通过一系列实验，该项目系统探讨了幼儿口语词汇识别特点、口语词汇识别的影响因素及口语加工的脑机制，基本按照原定计划完成了研究目标。截至目前，已公开发表论文12篇，分别发表在《Cognition》、《Language, Cognition, and Neuroscience》、《心理学报》、《心理科学进展》等期刊，另有5篇论文正在外审。.对幼儿口语词汇识别特点进行的研究发现：（1）声调信息不匹配时会明显抑制幼儿的口语词汇识别。（2）音段信息和超音段信息都会影响幼儿口语词汇识别，但影响程度不同。对高熟悉性的口语词汇，声母对幼儿口语识别影响最大，韵母和声调在口语词汇识别中作用相似；对低熟悉性的口语词汇，声母和声调对幼儿口语识别的影响最大，且两者在口语词汇识别中的作用相似。（3）实验结果支持TRACE模型的观点。关于口语词汇识别的影响因素，我们使用眼动方法，结合视觉情境范式，考察了语音及儿歌语境、语义一致性及实物视觉复杂程度、以及情绪图片和语义信息对幼儿听觉词汇识别的影响。研究发现：（1）幼儿在口语词汇识别的早期，可以更快地察觉到声调信息的变化。相比句子语境，儿歌语境促进了幼儿口语词汇识别。（2）视觉信息及语义信息共同作用于幼儿口语词汇的识别早期，在口语词汇识别的晚期，语义信息起到了更大的作用。（3）儿童对卡通情绪图片的注意加工好于真人面孔情绪图片，当语义信息和情绪图片信息相冲突时，4-5岁幼儿判断说话者情绪时依赖于语义信息。对幼儿口语感知的神经机制进行的研究发现，在前注意条件下，幼儿对汉语一声三声对比音节诱发了失匹配负波(MMN)，对二声三声对比音节，幼儿没有诱发出MMN，表明幼儿在前注意条件下辨别声学相似度较高的二声和三声存在困难。