從《廣雅疏證》看王念孫的聲轉理論及其實踐

Wang Nien-sun's Theories and Practices of Sheng-chuan as Seen from the Kuang-ya shu-cheng

陳雄根(Chan Hung Kan)

《中國文化研究所學報》 20卷 (1989/01) Pp. 115-180

https://doi.org/10.29708/JCS.CUHK.198901_20.0007

並列摘要

The concept of sheng-chuan 聲轉 firstly appeared in Yang Hsiung's Fang-yen 楊雄方言 Kuo P'u 郭璞 adopted this concept and used the terms of yü chuan 語聲轉, sheng chih chuan, mou-sheng chih chuan 某聲之轉 besides chuan-yü 轉語 and yü chih chuan 語之轉. Chuan-yü arises because pronunciation varies from time and from place to place. Scholars of the Ch'ien-lung and Chia-ch'ing periods (1736 - 1820) of the Ch'ing dynasty gave detailed explaination of this concept Tai Chen 戴震 in his Preface to Twenty Chapters on, Chuan-yü put forward the idea that characters which are wei-t'ung 位同 or t'ung-wei 同位 are likely to have similar meaning. Tai in his later work Fang-yen shu-cheng also emphasized the importance of the concept of chuan-yü. The present article takes all the examples given in the Kuang-ya shu-cheng and looks at their pronunciation in the light of the phonological reconstructions proposed by contemporary scholars. It is found that characters that are said to be have sheng-chuan mostly have the same kind, if not the same, initials. As for the finals, the relations may be that of tui-chuan 對轉, p'ang-chuan 旁轉, or t'ung-chuan 通轉. There are only a few exceptions. Huang K'an 黃侃 considered that the so-called yin -chuan 音轉 put forward by Wang Nien-sun is the same as shuang-sheng hsiang-chuan 雙聲相轉, i.e. tui-chuan or p'ang chuan. This suggestion is, in the author's view, not altogether correct. Yang Tuan-chih 楊端志 in his Hsü-ku hsü eh 訓詁學said that yi-sheng chih chuan 一聲之轉is the same as shuang-sheng hsiang-chuan. This argument overlooks the fact that have the same kind of initials, i.e. p'ang-niui 旁紐 an lin-niu 鄰紐, and identical initials. Yü chih ch'ih-yen 語之侈弇, yü chih ch'ing-chung 語之輕重 and yü chih huan-chi 語之緩急, which appear frequently in the Kuang-ya shu-cheng are terms used to describe the minute differences of pronunciation between the characters that are sheng-chuan. It is important to note that Wang's ch 'ih-yen and ch'ing-chung are differentiated by teng-hu 等呼 of the final on the one hand and huan-chi and ch'ing-chung which relate to tones on the other. This use of two criteria leads to confusion. This is also a mistaken equation of he -yin 合音 with fan-ch'ieh 反切. Wang followed Tai Chen's theory of t'ung-wei hsiang-chuan 同位相轉 The term yi-wei hsiang-chuan 異位相轉 refers two characters which are chuan-yü with different initials. It goes against the theory of sheng-chuan in the Kuang-ya shu-cheng which emphasizes the identity of initials. These counter-examples, therefore, should be regarded as exceptions. Wang in his treatment of chuan-yü did not necessarily quote works of Yang Hsiung and Kuo P'u. In such cases, he used dialects of his own age, variant of characters and characters that from a compound. One further point concerning Wang's work is that hs does not seem to provide much evidence for the theory of yi-wei hsiang-chuan.