現代漢語選擇問句法的來源

Tue Origin of the Disjunctive Question in Modern Chinese

梅祖麟(Tsu-Lin Mai)

《中央研究院歷史語言研究所集刊》 49本1分 (1978/03) Pp. 15-36

並列摘要

The disjunctive question (DQ) in Modern Chinese is characterized by three features. (1) The forms是shih and 還是 hai. shih serve as the DQ-marker. The copula shih is identical with one of these markers, and is a constituent of the other. (2) A Modern DQ may use a single marker or a pair of markers; it can also dispense with it altogether. In contrast, the DQ-markers of Old Chinese and Han Chinese cannot occur twice in a disjunctive question. (3) The DQ of Old Chinese and Han Chinese almost always ends in an interrogative particle, e. g.,乎 hu, 邪 yeh, or與yü. The Modern DQ does not follow this rule. The interrogative particle 嗎, 麽 .me does not occur in a Modern DQ, and the forms which do occur, such as 啊 a or 呢 ne, are not interrogative particles. The present paper traces the origin of the Modern DQ, as characterized above, to the 5th century A. D. The main sources used consist of the Buddhist texts of Early Middle Chinese (4th to 6th C.), the Tun-huang 敦煌 manuscripts, the Tsu-t'ang chi 祖堂集(952 A. D., the earliest comprehensive history of Zen Buddhism) and the Chu-tzu yü-lei 朱子語類(1170-1200).
Disjunctive questions without a sentence-final interrogative particle already occurred during the 5th century. This accounts for feature (3).
Also in the 5th C., the particle 爲wei began to be used as a DQ-mar. ker, either singly or in a pair, yielding a new type of DQ with the form
N1 VP1 wei (N2) VP2
or N1 wei VP1 wei (N2) VP2.
The word wet meant' to act, to serve as' in QC. But in Early Middle Chinese, it began to be used as a copula, When wet was replaced by another copula shih 是, the word shih also became a DQ-marker.
The Modern DQ-marker hai. shih has a more complicated history. (1) In Early Middle Chinese, there occurred a general process of di-syllabification. One favorite device was to suffix a syllable-filler當tang, shih, or是fu to an already existing monosyllabic form. As a result of di-syllabification, the DQ-marker wei became爲當wet-tang,爲是wei-shih, or爲復wei-fu. Disjunctive questions containing these new di-syllabic DQ-markers occurred frequently in Middle Chinese. (2) Lexical substitution of wet by hai還in the context of disjunctive questions yielded a new DQ-marker hai during the 9th C. With the suffixing of是shih to hai, Early Modern Chinese acquired the DQ-marker hai. shih. This DQ-marker has two other sources, one involving the semantic shift of hai. shih from 'still' to 'whether, or', and the other resulting from the substitution of hal f or wet in the context of爲是wei-shlh.
The entire set of Modern DQ forms can be found in the Chu-tzu yülei. It is also shown that the Cantonese DQ-marker定dihng first occurred in the 8th C., in two poems by Tu Fu, and that during the 12th C. it was already a dialect word used only in South China.
Two methodological remarks conclude this paper. (1) In studying historical syntax, one should make a distinction between the frame and the filler. The structural frame of the Modern DQ already existed in the 5th C. The specific Modern DQ-markers shih and hai. shih, which serve as fillers, did not occur in that capacity until the 12th C. They came about through a complicated process of di-syllabification and lexical replacement. But throughout that process, the structural frame remained intact. (2) In the Appendix the author points out that Han Yd's韓愈prose contains several post-Han grammatical constructions, including Middle Chinese forms for the disjunctive question. He draws the conclusion that literary Chinese, which includes Pre-Ts'in prose as well as the writing of Han Yü, is too heterogenous a corpus to be studied scientifically.