试论量名不匹配构式在语料库中的低频出现率及内在原因

The low frequency of the classifier-noun mismatch configuration in corpus and its underlying reasons

前置量词有利于构建汉语宾语关系从句的前提之一是"量词-名词"不匹配构式的使用足够高频。本文借助新闻语料库ChineseTreebank5.0,考察该前提假设是否得到支持。分析显示,量词在宾语关系从句中更倾向于置后;置前的量词仅有10例,而且没有一例属于真正意义上的"不匹配-匹配"构式。本文试图阐释这种不均衡分布的内在原因,并指出汉语使用者在生成句子结构时,可运用其他语言手段,成功规避因毗邻成分不匹配所造成的潜在的语义理解困难。

By using data from the Chinese Treebank 5.0 Corpus,this paper investigates whether the classifier-noun mismatch configuration occurs frequently enough,a premise assumed in the psycholinguistic research on the facilitative role of pre-RC classifier in object-extracted relative clause(RC) processing.The corpus analyses of object-extracted RCs show that classifiers tend to occur in the post-RC position,but seldom occur in the pre-RC position,totaling 10 tokens only,none of which actually conforms to the mismatch-match pattern.This offers accounts for the asymmetrical distribution and suggests that natural language users mobilize other linguistic tools to avoid processing difficulties induced by a pre-RC classifier that is incongruent with the immediately following noun.