从核心词替换微观个案看词阶法

A Study of Rank Analysis From a Micro Case of Core Word Replacement

词阶法作为语源关系判断的一种方法,其理论基础可以通过观察词汇替换微观个案予以补充和检验。在由语言接触引发核心词替换的骑马坝傣语个案中,200核心词替换过程严格符合高阶少于低阶的分布模式,符合词阶法的理论假设;hihi-lolo词表和基于雅加达-莱比锡词表的改编词表等替换过程出现了较多例外,可能不宜应用到词阶法中。通过对观测数据进行曲线拟合,我们认为语言接触引发的核心词替换过程并不是匀速、渐变的,而是非匀速、突变的。拟合曲线和语言使用情况在对比验证中呈现出较强的关联性,整体上反映了语言接触情形的变化。我们通过模拟的方法将200核心词表与hihi-lolo等词表进行比较,发现200核心词表更能反映接触信息,更适合被应用到词阶法中。

The method of rank analysis,as a way to determine language relationships,can be supplemented and tested by observing micro cases of vocabulary replacement.In the case of the Qimaba Dai language,where core word replacement is triggered by language contact,the replacement of 200 core words strictly conforms to the distribution pattern of high rank being less than low rank,which is consistent with the theoretical assumption of rank analysis.However,in cases such as hihi-lolo and Jakarta-Leipzig adapted word lists,the number of high rank core word replacement is not always fewer than that of low rank replacement,indicating the crucial role of word list selection in the application of the rank analysis method.Through data fitting,we found that exponential or sigmoid curves can achieve higher goodness of fit than linear fitting.The data also exhibits a significant jump around 1990,indicating that the core word replacement process in the Qimaba Dai case is not uniform or gradual,but rather abrupt and non-uniform.This can also be explained by the contact mechanism,where contact suddenly occurs or intensifies within a certain time period,leading to abrupt replacements of vocabulary,similar to waves crashing.The comparison between the ftted curve and the language usage patterns shows a strong correlation,reflecting the overall changes in language contact situations.Based on the fitting of available data,it is predicted that the intersection point of high and low rank substitutions may occur for participants born between 2031 and 2033,indicating a reversal of high rank substitutions surpassing low rank substitutions.Additionally,Mandarin Chinese may completely replace Dai language as the first proficient language within language communities at this time.These predictions made by the fitting model require further tracking and survey validation.The simulation shows that the 200 core word list reflects language contact information better than the hihi-lolo list and the Leipzig-Jakarta adapted word list.When controlled for the same size of word lists,the simulation demonstrates that using the 200 word list reduces the possibility of errors or invalidity in implementing the rank analysis method compared to the other two word lists.Different core word schemes have a significant impact on whether there is a difference in the number of high and low rank replacements.Comparatively,the 200 word list is more effective in creating a difference in the number of high and low rank replacements in the cases analyzed in this study.