检验自由押韵的统计方法

Statistical Methods in Testing Free Rhyming

押韵材料是研究汉语历史音韵的重要材料.然而,合韵现象的存在给传统的研究带来了困难.通过引入“自由押韵”的概念,我们可以使用统计方法解决这一问题.本文将介绍三种判断两组字是否自由押韵的统计方法,分别见于白一平(Baxter 1986)、朱晓农(1989)和白一平(Baxter 1992).通过评估与比较,我们认为仅有第一种方法是可靠的,其他两种方法不可靠的原因主要在于对“自由押韵”这一概念的误解.因此目前通过后两种统计方法得到的结论还应当谨慎对待.使用统计方法观察押韵材料,不仅可以更准确地区分自由押韵与非自由押韵,还可以发现更多问题,推动汉语历史音韵的研究.

The material of rhyme sequences is very important in studying the historical Chinese phonology.However,the phenomenon of mixed rhyming has brought many difficulties to traditional scholars.Through introducing the notion Free Rhyming,we can use statistical methods to resolve the problems.In this paper,we will present three main statistical methods to test whether two groups of words rhyme with each other freely.They are Baxter(1986),Zhu(1989)and Baxter(1992).With comparison and evaluation,we think only the first one is reliable and the core problem in the other two methods is caused by a misunderstanding of the concept Free Rhyming.Using statistical method to analyze the materials of rhyming can not only distinguish Free Rhyming and Non-Free Rhyming more accurately,but also can propose new questions and promote the research of historical Chinese phonology.