汉语语义记忆提取的初步研究

PRELIMINARY STUDY OF SEMANTIC RETRIEVAL FOR CHINESE LANGUAGE

对汉语语义记忆的提取进行了两个实验。实验1是自由联想实验,从十三个句子中选出二十三个词作为刺激词。结果表明,以两词间所保持的常识关系来定义的语义距离,对预测自由联想的作业是有效的。实验2采用反应时技术,对真假命题判断及信息提取的时间进行了考查。实验结果与A.M.Collins等人的结果一致,说明了存储于网络不同水平上的句子其提取时间不同这一事实。结果还表明,在信息提取阶段上,汉字字形的作用不是一个重要的因素。

Two experiments were conducted to investigate the semantic retrieval for Chinese language. In experiment 1, twenty-three two-character words selected from thirteen sentences were used as stimuli in a free association test. The results indicated that semantic distance in terms of commonsense relations between words is useful to predict the performance of the free association test. In experiment 2, the true/false judgement reaction time technique was used. Thirty-eight (20 true and 18 false) sentences were used as stimuli in a true/false judgement test. The findings coincided with the results obtained by Collins and Quillian. The difference of retrieval time between levels in which the sentences were stored was verified. The results indicated that in the stage of information retrieval, the morphological effect of Chinese characters is not an important factor.