梵汉对音与中古汉语声调的构拟——再论不空音系的声调和梵语的重音规则

Reconstructing Middle Chinese Tones through Sanskrit-Chinese Transcriptions:The Tones of Amoghavajra and Sanskrit Accentual System Revisited

一般认为,可以在梵汉对音中通过吠陀梵语的乐调重音构拟古汉语声调的调值,如不空最常用上声对译高调(udaita),因而其调值也是最高的。但实际上,不空一派对音普遍倾向于多用上声,与梵语的重音位置无关;所以既不能直接证明上声读高调,也不能说明吠陀梵语的重音规则还保持不变。虽然当时的梵语文献已经不再标注重音,但通过近现代口语的读音、中古印度语和梵语发生音变的条件,以及古代印度文献的记载三方面材料的相互印证,可以看出梵语的重音规则确实发生了变化。因此,除非有其他确凿的证据,不宜直接使用吠陀梵语的乐调重音规则处理梵汉对音材料。至于不空音系四声的音长,进一步印证了去声最长,入声最短;提出上声可能略长于平声,但差别并不显著。

It is generally accepted to reconstruct ancient Chinese tones through Vedic accents in Sanskrit-Chinese transcriptions.For example,Amoghavajra most frequently used the rising tone(shangsheng)to transcribe Vedic accented(udatta)syllables,suggesting that its pitch is the highest.In fact,however,it is a common preference to apply the rising tone,regardless whether it corresponds with udatta or not.Thus,it neither implies the pitch of the tones,nor proves the Vedic accentual system was still preserved.Although accent was no longer marked in Sanskrit texts,the accentual system is believed to be changed,according to(i)modern pronunciation of Sanskrit and New Indo-Aryan languages,(i)the conditions of sound changes in Middle Indic and Sanskrit,and(ii)philological evidence from Ancient India.Consequently,it is not appropriate to apply Vedic accentual system on Sanskrit-Chinese transcriptions,unless there are any strong reasons.As for the duration of tones,the departing tone(qusheng)is proved to be the longest,and the entering tone(rusheng)the shortest,while the rising tone is likely to be slightly longer than the level tone(pingsheng).