基于类比语料库的汉文训读词汇特征研究

Comparable corpus-based study on lexical features of Kanbun-kundoku

本文基于自建汉文训读语料库与同时期和文语料库,类比考察汉文训读文本区别于和文文本的词汇使用特征,以及不同文本类型对训读结果的干扰。研究结果显示,佛经类训读与汉籍类训读的词汇特征无显著性差异,表明训读语言的词汇特征并未因文本类型的变化而受到明显干扰;与和文文本相比,汉文训读文本的词汇呈现出总体一致的复杂化趋势,具体表现为信息密度较高、词汇范围较广、概括性较强、汉语词较多而和语词较少。本研究认为,训读规则下的双语激活、目的语语体规范以及特定社会语境因素的限制和参与是汉文训读语言呈现出复杂化趋势的潜在原因。

Based on newly-built Japanese Kanbun-kundoku Corpus(including Buddhist scriptures and Chinese classical literature)and the contemporaneous Wabun Corpus(original Japanese),this study investigates the lexical features distinguishing the Kanbun-kundoku texts from the Wabun texts,and the interference of different text types on the lexical features of their outputs.No significant difference was observed in the lexical features between the outputs of Buddhist scriptures and Chinese classical literature,indicating that the lexical patterns of Kanbun-kundoku texts are not significantly influenced or interfered with by the variation in text type.When compared to Wabun texts,the lexical features of Kanbun-kundoku texts have a highly consistent tendency of complexity in terms of higher informational density,wider lexical range,higher generality,more Chinese-derived words,and fewer native Japanese words.We argue that the tendency of complexity in the Kanbun-kundoku texts can be attributed to the constraint and involvement of the factors including the bilingual activation under the kundoku's pattern,the stylistic norms of the target language,and the specific social context.