词频分布参数可以细分汉语新闻语体吗?  

Can Parameters in Word Frequency Distribution Distinguish Chinese News Genres?

作　　者：黄伟[1] 刘海涛[2,3] Huang Wei Liu Haitao

机构地区：[1]北京语言大学对外汉语研究中心,100083 [2]浙江大学外语学院,310058 [3]广东外语外贸大学外国语言学及应用语言学研究中心,510420

出　　处：《语言教学与研究》2017年第4期79-86,共8页Language Teaching and Linguistic Studies

基　　金：国家社会科学基金(13CYY022);中国博士后科学基金(2016M590529);北京语言大学青年英才培养项目资助

摘　　要：本文统计分析了口语体和书面语体的现代汉语新闻报道与新闻评论文本中的词频分布规律,发现:现代汉语新闻语体文本的词频分布符合普适的词频分布规律(齐普夫-曼德博定律);书面语体新闻文本的词频齐普夫-曼德博分布的参数a显著高于口语体新闻文本词频分布参数a;参数a还受文本长度影响,但在报道体与评论体中不存在显著差异;齐普夫-曼德博定律的另一个参数b在口语体与书面语体的新闻文本间不存在显著差异。对文本的词频分布规律在不同语体中体现出的普遍性与多样性的研究,有助于加强语体描写与语言学研究的客观性、精确性与科学化。To ascertain the relation between word frequency distribution and text genres, a quantitative investigation is conducted into word frequency distribution in modern Chinese texts of oral and written news （and news comments）, and yields the following find- ings. First of all, although the Zipf-Mandelbrot law fits the word frequency distribution well in Chinese texts of both oral news and written news, the parameter a of this law is signifi- cantly larger in texts of written news than that in texts transcribed from oral news. In addi- tion, although the parameter a seems influenced by the length of text, no significant differ- ence is found between texts of news and texts of news comments. Finally, as to the parame- ter b in Zipf-Mandelbrot law, there seems to be no significant difference in it between oral texts and written texts. In brief, word frequency distributions in different text genres seem to present both homogeneity and heterogeneity, which may be exploited to make for objec- tive, accurate, and scientific linguistic descriptions of and inquiries into text genres.

关 键 词：新闻语体 词频分布 语体特征 计量语言学 齐普夫定律 

分 类 号：H15[语言文字—汉语]