【作者】 张亚蓉；

【导师】 张玉来；

【作者基本信息】 苏州大学 ， 语言学及应用语言学， 2008， 博士

【摘要】 上古音在汉语史乃至整个汉藏语系的研究中都占有非常重要的地位。上古音研究开始的相当早,早在先秦就有学者关注,历代都有研究,至今已有两千多年的历史。到了清代,上古音研究则盛极一时。但是,清儒研究古音,精力主要集中在先秦韵语的整理上,成就也多局限于古韵部的划分。上古声母方面的研究成就远在韵部之下。直至20世纪西方历史比较语言学的理论和方法传入我国,上古音研究才有了重大突破。学者们开始运用历史语言学的方法,从汉语史的角度,构拟上古音系,不断对前人的拟音系统做出改订,提出新的构拟系统。国内学者主要分为两派:一派以王力、李新魁、何九盈、郑张尚芳等为代表,主张阴阳入三分,阴声韵不带塞音韵尾说;另一派以李方桂、周法高、余迺永等为代表,主张阴声韵带浊塞尾说,或者改浊塞音尾为浊擦音流音尾。随着前人不断地深入探索,上古音的研究材料也日益丰富,诸如先秦韵文、重文、异读、《说文》谐声、汉语方言、域外对音等;研究方法也日臻完善,包括韵脚归纳法、谐声推演法、域外对音法、历史比较法等。本文主要运用历史比较语言学的方法,从汉语语音史的角度,研究上古汉语的语音系统和变化规律。论文以《说文解字》中的谐声字为主要材料,从分析原始谐声系列的基础工作做起,对每个系列的谐声字都进行了中古音韵地位的分析,然后用概率公式计算各声类、韵类、调类发生谐声关系的几率,并以此推断它们间的疏密关系,最后根据统计出的数据信息来探讨上古汉语的声、韵、调系统。具体步骤如下:首先,结合大徐本、小徐本《说文解字》、段玉裁《说文解字注》、朱骏声《说文通训定声》、桂馥《说文解字义证》、王筠《说文解字句读》、《古文字诂林》、《说文解字诂林》对《说文解字》的各个版本进行校检,整理出字形无误的形声字,最终形成《说文》谐声表。其次,对《说文》谐声表中的每一个字标注中古音韵地位,进行古音上的系联。最后,采用统计学中的概率统计法计算出《说文》中形声字声、韵、调的实际相逢数及几遇数,对实际相逢数超过几遇数的谐声系列进行音理上的讨论与分析。全文共分七章,第一章主要回顾了人们对谐声字音读关系的认识历程,并对各家总结的谐声原则做了详细的介绍。第二章主要根据不同的《说文》版本以及《古文字诂林》等材料对《说文解字》中形声字的收字数量以及形声字的误析现象做了详细校检。第三章对《说文解字》中谐声字的基本声符做了仔细认定。第四章详细介绍了本文所用材料的计算机处理方式以及数理统计的计算方式。第五章、第六章、第七章分别结合《说文解字》中的谐声数据对上古的声、韵、调系统做了深入的探讨。几率统计法是本文最重要的研究手段。全文所有数据均来自声母、韵部及调类谐声频率统计表和谐声几遇表。在这些数据的基础上,再根据实际相逢数和几遇数的大小来确定某两类字之间是否有音理关系可循。将谐声关系几遇率很高的那些声类归纳之后,可以看出《说文解字》谐声反映出的上古汉语单声母系统比较整齐,至少包含以下五类:(1)唇音组[p]、[ph]、[b]、[m]。(2)舌音组[t]、[th]、[d]、[n]、[l]。(3)齿音组[ts]、[tsh]、[dz]、[s]、[z]。(4)牙音组[k]、[kh]、[g]、[η]。(5)喉音组[?]、[x]、[γ]异常谐声则表现为语音性质的近似性或者复辅音的遗留。《说文》谐声反映出的上古汉语韵部系统,主要是以30部自谐为主,即谐声的原则还是主要元音相同或相近,韵尾相同。即使有个别表现为阴入通谐的异部谐声现象,也非主流。《说文》谐声反映出的调类对立不是非常鲜明,平入突出是其反映出的最主要的调类特征。我们赞成辅音韵尾的脱落在声调形成过程中的作用,韵尾辅音说较之其他学说有其合理性,它有助于解释许多特殊的音变现象,如破读就可以解释为有无-s尾的转化。辅音韵尾说对某些特殊的谐声现象也具有较强的解释力。但是,这些韵尾是如何脱落的,脱落后又如何转变成音高特征,目前的研究仍有许多值得探索之处。声调的形成因素应该比较复杂,可能包含有音高、韵尾、清浊等多方面原因。因此,对于声调最初状态的研究还需继续。更多还原

【Abstract】 The Old Chinese phonology plays an important role in the study of the history of Chinese language, also in that of Sino-tibetan phylum. Although it started early, the study of the old Chinese phonology reached its prosperous period during Qing dynasty. However, most Qing scholars concentrated only on the rhyme categories, the achievement of initial consonants was far behind the study of the former one. The great progress had not been made until the introduction of the language theory of historical comparison in the 20^th century. Scholars began to reconstruct the system of old Chinese phonology with scientific phonetic symbol. Some scholars, e.g., Wang Li, Li Xinkui, He Jiuying, Zhengzhang Shangfang etc., consider that the rhyme category should be sorted as y（?）nsh（?）ng words, y（?）ngsh（?）ng words and r（?）sh（?）ng words, and there is no final consonant at the end of y（?）nsh（?）ng words. Others like Prof. Li Fangkui, Prof. Chou Fagao and Prof. Yu Naiyong etc., argue that there are final consonants or glide consonants in the end of y（?） nsh（?）ng words.On the basis of the previous research, the materials on the old Chinese phonology become more and more enriched, such as the verses in Qin dynasty, the xi（?）sh（?）ng characters in shuowenjiezi, etc.. Correspondingly, we have accumulated much more research methodologies, e.g., summarizing the rhyme in shijing, summarizing the relations of xi（?）sh（?）ng characters in Shuowenjiezi, the historical comparison, and the comparison of two relative languages, etc.. Based on the Phonogram characters in Shuowenjiezi, in the present dissertation, we use statistical method to discuss the relations among xi（?）sh（?）ng characters, and further explore the initial consonant system, the rhyme categories and the tone system in the old Chinese phonology.We discuss the above mentioned materials in this thesis in seven parts. In Part 1 we briefly introduce the essence of xi（?）sh（?）ng characters and the principles of xi（?）sh（?）ng characters. In Part 2 we mainly study how to recense the xi（?）sh（?）ng characters in shuowenjiezi. In Part 3, the cognizance of the basic phonology symbol of xi（?）sh（?）ng characters is introduced. Part 4 introduces the mathematical statistics and uses it to deal with the xi（?）sh（?）ng characters via computer technique. Finally, the last three parts explore the system of initial consonant system, the rhyme categories system, and the tone system according to the statistics from the xi（?）sh（?）ng characters in shuowenjiezi.In summary, the high relations in xi（?）sh（?）ng characters, we get a regular consonants system from the nomal xi（?）sh（?）ng relation. It includes five groups:（1） labial initials phonology group: [p], [ph] ,[b],[m].（2） dentals phonology group: [t],[th],[d],[n],[l],（3） dentals sibilants phonology group: [ts],[tsh],[dz],[s],[z].（4） velars phonology group: [k],[kh],[g],[η].（5） laryngeals phonology group: [?],[x],[γ].We can explain the abnormal xi（?）sh（?）ng relations by a theory of comparability in phonology and some complex consonants. We can also have a kind of rhyme categories like traditional 30 groups from xi（?）sh（?）ng characters.The difference between four tones does not stand out in xi（?）sh（?）ng system, just p（?）ng and r（?） stand facing each other in xi（?）sh（?）ng system. It is possible that the form of the tone includes other complex reasons, and the final form of p（?）ng sh（?）ng q（?） r（?） experiences a long process in phonology development.更多还原