敦煌石室《心經》音寫抄本校釋序說

敦煌遺書裡有五件《心經》音寫抄本，其中斯2464號的抄本即是當年《大正藏．般若部》（No.256，冊8，頁851a-852a）據以錄文而題名為《唐梵飜對字音般若波羅蜜多心經》的本子。關於此一音寫抄本的譯者歸屬問題，學者之間雖也有長田徹澄針對該一抄本題記裡所云「出自奘師之手」的說法提出質疑（參見《密教文化》第56號昭和10年，頁44-5），但多數學者似乎仍然接受斯2464號抄本題記中的記述，而認定該一文本乃出自玄奘所譯。在這個問題上，陳寅恪先生或許是華夏學圈裡首先發難質疑題記之說的學人，他曾懷疑該一石室文本或許並非出自玄奘而是來自不空所譯（參見氏著《金明館叢稿二編》，上海古籍出版社1980年初版第一刷，頁175）。此後方廣錩教授也提出了類似陳寅恪先生的觀點，並主張該一文本或為「玄奘原譯」，而由「不空潤色」。但陳、方兩人也都不曾提出任何積極有效的證據以實其說。（方氏之說參見氏編《般若心經譯注集成》，上海古籍出版社1994年初版第一刷，「前言」頁12）及至1985年，福井文雅教授以〈新出「不空譯」梵本写本般若心經〉為題，藉由文獻學的比對與若干經錄的載記，學界之間似乎才有學者真正面對了問題而嘗試證成此一說法（福井之文參見《仏教學論集──中村瑞隆博士古稀記念論集》，頁229-246）。此後福井教授繼雲居寺《房山石經》的出土與刊佈，一連串發表了不少有關《心經》音寫本的論文。然而，在有關譯者的歸屬問題上，福井教授的立場則一如他所使用的論證方式，始終都未曾改變，仍然堅持敦煌石室抄本與《房山石經》裡兩方《心經》音寫刻石之一，是出自不空三藏之手。我們雖然並沒有在譯者歸屬的認定上提出任何新穎的說法，但是在這個問題的探究上，卻也不打算再度採取福井教授那種純以外部證據來進行論證的方式。在本文對《心經》音寫抄本的校釋過程裡，我們將嘗試讓這些文本自己訴說它們的身世。換言之，我們準備從這些文本自身的「口供」與「筆錄」著手，而試圖透過其內部字裡行間所反映出來的語音特徵的描寫，以及其間音寫用字的選造分析，來找出更為直接而有力的證據，從而顯示我們在譯者歸屬問題上的論斷也更具可靠性。並且，我們認為這種藉由語段的音韻分析以及字詞選造的風格學分析（stylistic analysis），其間所能得出來的證據將會比福井教授所舉出來的那些外部證據更為直接而有力。蓋諸如《房山石經》之類的出土文獻，這一類外部證據的獲致往往都是既偶然而又意外的，但文獻的語音特徵乃至其字詞選造的語言風格之類的成素，則明顯的具有其存有論上的旨趣（ontological significance），因為它們都是內在於文獻之中並隨文獻的存在而存在。

關鍵字

《心經》；音寫；音譯；梵漢對音；佛經譯音；二合；引；中古音；隋唐音韻；長安方言；河西方言；風格學分析

並列摘要

Among the texts rediscovered at Dunhuang, there are five manuscripts which are phonetic transcriptions of the Heart Sutra. Of these, Stein No. 2464 was the version used as the basis for T. 256 (Tang Fan fandui ziyin bore boluomiduo xin jing 唐梵飜對字音般若波羅蜜多心經; T8.851a-852a). Although Nagata Tetsusumi doubts the traditional claim that Xuan Zang is the translator (see Mikkyo bunka 密教文化 56, Showa 10 (1935), pp. 44-45), most scholars accept the description in the colophon of Stein No. 2464 that the work was translated by Xuan Zang. However, in Chinese academic circles, Chen Yinke appears to be the one to have initiated the claim questioning the validitity of this colophon. He suspected that this Dunhuang version was not translated by Xuan Zang, but rather by Amoghavajra (see his Jin Ming cong gao er pian 金明館叢稿二編, Shanghai: Shanghai guji chubanshe, 1980, p. 175). Subsequently, Prof. Fang Guangchang proposed a similar idea, that the text may have been ＂originally translated by Xuan Zang,＂ but ＂edited and embellished by Amoghavajra＂ (see Fang Guangchang, ed., Bore xin jing yizhu jicheng 般若心經譯注集成, Shanghai: Shanghai guji chubanshe, 1994, p. 12). However, neither Chen nor Fang have set forth solid, valid evidence proving their theories. Prof. Fumimasa Fukui’s article of 1985, ＂Shinde 'Fuku’ Bonbon shabon honnya shingyo＂新出「不空譯」梵本写本般若心經, addresses this issue more fully, attempting to prove this theory using philological comparisons and information from numerous catalogs (see 仏教學論集──中村瑞隆博士古稀記念論集, pp. 229-246). After this, Prof. Fukui worked on Yunju si's excavation and publication of the Fangshan stone tablet sutras. He also wrote a series of articles on the phonetic transcription manuscripts of the Heart Sutra. Prof. Fukui maintains that Amoghavajra is the translator of two of these manuscripts: one from Dunhuang and one from Fangshan. In the current article, although I do not propose any new conclusions regarding the identity of the translator, I avoid arguing on the basis of purely external evidence in the manner of Prof. Fukui. Rather, I will attempt to allow these manuscripts to tell their own stories through the process of their annotation and explication. Specifically, using ＂oral source＂ and ＂dictation thereof＂ as reference points, I will describe the phonetic features implicit within the text, and analyze the choice of characters used in the phonetic transcription process. This uncovers direct and powerful evidence proving, with a higher degree of reliability, that our conclusions about the identity of the translator are correct. Additionally, I believe that this type of segment phonetic analysis, combined with a stylistic analysis of word choice can provide evidence that is more direct and more powerful than the external evidence presented by Prof. Fukui. Generally speaking, with regards to excavated materials such as those of Fangshan, the acquisition of external evidence is often haphazard and unintended, while the elements of phonetic features, character choice, and style clearly possess ontological significance because they are all internal to the text, and exist by reason of the existence of the text.

並列關鍵字

Heart Sutra ； transliteration ； transcription ； Sanskrit-Chinese phonetic comparison ； Middle Chinese pronunciation ； Sui, Tang phonology ； Chang'an dialect ； dialect for the region west of the Yellow River ； stylistic analysis