期刊文献+

基于汉文体计量特征的《日本书纪》各卷分类研究

A Study on the Classification of Various Volumes of the Nihon Shoki Based on the Quantitative Characteristics of the Classical Chinese Writing Style
原文传递
导出
摘要 《日本书纪》是日本现存最早用汉字和汉文编纂的史书。从语言的角度来看,该书各卷所使用的汉文体呈现出了不同的特点。前人根据万叶假名的汉字及音韵系统、天皇即位和定都的叙事风格、天文天象的记载情况、语言表达的语法问题等尝试对该书各卷进行划分,但是这种分类不能够系统反映各卷的汉文体特点,也无法对第1卷、第2卷、第30卷进行划分。常用字词的使用和分布可反映文章的行文特点,本文利用无指导机器学习技术对这些汉文体语言特征进行聚类分析并和《隋书》进行对比发现,《日本书纪》30卷的汉文体可以分成4类,部分结果和前人研究结果一致。新的发现是,《日本书纪》第1、2、3卷的汉文体和《隋书》一致;第14卷至30卷(第23卷除外)和《隋书》相似;第5卷至13卷以及第23卷和《隋书》差距较大;第4卷的行文不但和《隋书》不同,也和《日本书纪》其他各卷有很大差距。通过对《日本书纪》和《古事记》汉文体的研究也证明,前沿的语言智能技术可以在古文献研究中发挥作用。 The Nihon Shoki,an ancient book of classical Japanese history compiled in the Classical Chinese writing style,exhibits various linguistic characteristics.Previous scholars attempted to divide the volumes of the Nihon Shoki based on the Chinese characters and phonetic system of the Man'yogana,the narrative style of the emperor's ascension and relocation of the capital,the recording of astronomical phenomena,and grammatical issues in language expression.However,this classification cannot systematically reflect the characteristics of the Classical Chinese used in each volume,nor can it divide the first,second,and thirtieth volumes.The use and distribution of common words and phrases can reflect the characteristics of the text.This study uses unsupervised machine learning technology to cluster these Classical Chinese language features and compares them with the Sui Shu.It was found that the Classical Chinese used in the 30 volumes of the Nihon Shoki can be divided into four categories,and some of the results are consistent with previous research.The new discovery is that the Classical Chinese used in the first,second,and third volumes of the Nihon Shoki is consistent with the Sui Shu,the fourteenth to thirtieth volumes(except the twenty-third volume)are relatively similar to the Sui Shu,and the fifth to thirteenth volumes and the twenty-third volume are quite different from the Sui Shu;the text of the fourth volume is not only different from the Sui Shu,but also greatly different from the other volumes of the Nihon Shoki.The study of the Classical Chinese used in the Nihon Shoki and the Kogiki also proves that cutting-edge language intelligence technology can play a role in the study of ancient documents.
作者 施建军 王大钊 Shi Jianjun;Wang Dazhao(Shanghai International Studies University,China)
出处 《日语学习与研究》 2024年第4期12-25,共14页 Journal of Japanese Language Study and Research
基金 2014年度国家社科基金重大项目“日本汉文古写本整理与研究”(项目批准号:14ZDB085)的阶段研究成果,项目主持人:王晓平。
关键词 日本书纪 隋书 汉文体 文本聚类 Nihon Shoki Sui Shu Classical Chinese text clustering
  • 相关文献

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部