考虑词汇替换的汉语古音构拟检测

Evaluation of Ancient Chinese Reconstructions Considering Lexical Substitutions

以高本汉、王力、李方桂、董同龢、周法高、郑张尚芳、潘悟云、白一平—沙加尔等八家的上古拟音、中古拟音为对象,以经过平衡的60个现代汉语方言点为参照,运用相似性自动判断程序,进行语档相似度、相似度变化速率等参数的计算,并通过数据统计考察词汇替换的影响。结果显示,综合各家拟音系统内部上古与中古的相似度、拟音与现代汉语方言的相似度、相似度变化速率、相似度变化速率比,以及八家拟音在词汇替换前后两种情况下的具体表现来看,郑张尚芳的拟音在八家中是自洽性和合理性最高的,潘悟云和白一平—沙加尔两家的拟音综合表现也相对较好。考虑词汇替换后,各家拟音与方言语档的相似度均更小,词汇替换会对语言演变速率的计算产生影响。

This paper collects 40 Swadish core words from 8 reconstruction systems for Old Chinese and Middle Chinese raised by Bernhard Karlgren,Wang Li,Li Fangkuei,Tung T’ungho,Chow Fakau,Zhengzhang Shangfang,Pan Wuyun and Baxter-Sagart,as well as 60 balanced modern Chinese dialects,transcribes them to form doculects,and uses the Automatic Similarity Judgment Program to carry out the similarity and the similarity change rate,and investigate the influence of lexical substitution by means of data statistics.The results show that,considering the similarity between Old Chinese and Middle Chinese,the similarity between reconstructions and Modern Chinese dialects,the change rate of similarity,the change rate ratio of similarity,and the specific performance of the eight reconstructions before and after Lexical substitution,Zhengzhang Shangfang’s reconstructions is the most consistent and reasonable among the eight,Pan Wuyun and Baxter-Sagart also have relatively good comprehensive performance.After considering lexical substitution,the similarity between reconstructions and dialects is smaller,and lexical substitution has an impact on the calculation of language evolution rate.