韵律模式和语素位置概率对汉语学习者切分歧义词的影响

Effects of prosodic patterns and the morpheme position probability on word segmentation and recognition in overlapping ambiguous strings by learners of Chinese

在中文文本阅读过程中,读者如何对词语进行分词和识别,基本加工单位是什么,一直存在着争论。本文采用线上词汇命名任务,以汉语母语者和高级水平的二语者为被试,研究韵律模式和语素位置概率对汉语词语切分和识别的影响,并在此基础上探讨基本加工单位的大小问题。研究发现:韵律因素和语素位置概率显著影响词语的切分和识别;汉语母语者和二语者最大的差异在于二语者加工单位的长度小于母语者,这可能是由于二语者的知觉广度范围较窄导致的。汉语的基本加工单位大小基本遵循“长词优先”原则,同时受韵律特征、语素特征、语言水平等因素的影响。

There is an ongoing debate on how readers segment and recognize words during Chinese text reading and what the basic processing unit is.This study explores the impacts of prosodic pattern and morpheme position probability on readers􀆳word segmentation and recognition,and discusses the size of basic processing units.Using online word naming tasks,ambiguous three-character strings ABC and four-character strings ABCD are briefly shown to native Chinese speakers and learners of Chinese.We conclude that prosodic factors and the morpheme position probability significantly affect Chinese word segmentation and recognition.Importantly,the biggest difference between native Chinese speakers and learners of Chinese is that the basic processing unit of learners Chinese is shorter than that of native speakers,which is probably due to the narrower perceptual span of Chinese learners.Moreover,the size of basic processing unit in Chinese reading generally follows the principle that long words get priority,and prosody features,morpheme features and language ability will affect the size of the basic processing unit as well.