三國支謙譯經研究

本論文的目的在於考察三國支謙譯經中的文獻問題，主要從四個角度展開，即：支謙譯經的經錄記載、支謙譯經的譯者歸屬、支謙譯經的原典語言以及支謙譯經對前人譯經的改譯現象。研究方法上，主要使用了史料來源考察、多文本對勘、梵漢對音、計量統計等方法。研究材料上，為了確保文獻分析的可靠，本文使用了多樣化的材料來源。比如在探討語言問題時，既使用了敦煌寫卷、日本古寫經、不同朝代的刻本等多種版本，也對比了梵語、巴利語、犍陀羅語等不同語言的記載。在探討譯者問題時，既使用了經錄、史傳等外部材料，也使用了大量的內部材料，比如譯經中的詞、短語乃至無法形成獨立意義的片段。本文的主要結論可以分為以下幾點。在經錄記載方面，《出三藏記集》對支謙譯經的記載，依靠了一部分無法判斷可靠性的材料來源，因此我們根據《出三藏記集》來研究支謙譯經時，應該更加謹慎。此外，隋法經等撰《眾經目錄》與《出三藏記集》的記載有所區別，但是這些區別並非憑空產生，很多可以在《歷代三寶紀》所引經錄中得到證實。這也就意味著，如果要更為全面地了解支謙譯經，應該保持更為開放的態度來審查那些可能為支謙所譯的經典。在譯者歸屬方面，T6《般泥洹經》、T362《阿彌陀三耶三佛薩樓佛檀過度人道經》、T474《維摩詰經》以及T708《了本生死經》乃至T225A《大明度經》第一品，都應該歸為支謙所譯。在原典語言方面，支謙譯經中殘留了很多印度俗語的痕跡，支謙譯經的原典語言相比起古期印度雅利安語出現了語音的弱化。我們可以推測，這種語音的弱化，很可能即是支謙譯經原典語言的面貌。在改譯現象方面，從《法句經》、《太子瑞應本起經》對《中本起經》的改譯情況，我們可以得到一個重要的結論，即支謙會採取不一樣的標準來改譯前人譯經。學者多以支謙《大明度經》對支婁迦讖《道行般若經》改譯情況來確定支謙的譯經偏好，認為支謙偏好文雅，偏好四言格，偏好簡略，從而在判斷某些譯經的譯者歸屬時，將不符合這個條件的譯經視為非支謙所譯，如認為T362《阿彌陀三耶三佛薩樓佛檀過度人道經》非支謙所譯。事實上，這種偏好只是支謙部分經典所體現的風格而已，並不是每部支謙作品都如此。《法句經》、《太子瑞應本起經》對《中本起經》的改譯情況，讓我們知道支謙改譯的時候也會嘗試著貼近印度語系文本，或者嘗試使用零散多變的句式。因此在探討支謙譯經風格以及探討支謙譯經譯者歸屬的時候，需要考慮到更多不同的可能性。

關鍵字

經錄；音譯詞；原語；改譯；《大明度經》；《太子瑞應本起經》

並列摘要

The purpose of this thesis is to study Zhi Qian’s translation of Buddhist texts in the Three Kingdoms period, from the four perspectives: the Chinese Buddhist catalogue, attribution of the translation, the source language, and the phenomenon of Zhi Qian’s revision of the previous translators’ work. In terms of research methods, it mainly uses historical source analysis, comparative study of texts in different languages, Sanskrit-Chinese transcription research, measurement statistics. In terms of research materials, this study uses a variety of material sources. For example, when discussing language issues, it uses different editions, such as Dunhuang manuscript, ancient Japanese manuscript and carved editions from different dynasties. What is more, it also compars the materials in Sanskrit, Pāli, Gāndhārī, and other languages. When it comes to the attribution of the translation, not only external materials such as Chinese Buddhist catalogue and Biographies are used, but also a large number of internal materials, such as single words, short phrases, and even fragments which cannot form independent meaning. The conclusions of this research are as follows: In terms of Chinese Buddhist catalogue, the most reliable catalogue Chusanzang jiji出三藏記集 records numerous Zhi Qian's translations, but this is based on the materials which in some cases may not be reliable. Therefore, we should be more careful when we study Zhi Qian's translations based on Chusanzang jiji. In addition, there are some differences between the records of the first completed Chinese Buddhist catalogue, Chusanzang jiji, and the second one, Zhongjing mulu, which was compiled by Fajing in the Sui dynasty. Many of them can be confirmed by the citations from Lidai sanbao ji which was complied by Fei Zhangfang in the Tang dynasty. This means that if we want to have a more comprehensive understanding of Zhi Qian's translations, we should keep a more open attitude to examine the texts that are suspected to be translated by Zhi Qian. In terms of the attribution of the translations, the Bannihuan jing 般泥洹經(T6), Amituo sanyesanfo saloufotan guodu rendao jing 阿彌陀三耶三佛薩樓佛檀過度人道經 (T362), Weimojie jing 維摩詰經(T474) , Liaoben shengsi jing 了本生死經 (T708), and T225A the first chapter of Damingdu jing 大明度經 should all be treated as Zhi Qian’s translations. In terms of the of the source language, there are numerous transcriptions in Zhi Qian's translations, about 300 of which are not found in the works by previous translators. These transcriptions are important materials for the study of Chinese phonetics. By comparing the pronunciation of these transcriptions in the Eastern Han dynasty corresponding Sanskrit words, one can see that there is a big gap between them. It can be explained according to the general evolution law of Indian phonetics. Zhi Qian's translations are influenced by Indian Prakrit. It is particularly noteworthy that the source language of Zhi Qian's translations weakened its pronunciation compared with the Old Indian-Aryan language. In terms of the phenomenon of revision, we can draw an important conclusion that Zhi Qian adopted different measures to retranslate the texts. Scholars usually use the case of the Zhi Qian’s revision of Damingdu Jing to infer his preferences in translation. They thought that Zhi Qian prefers elegant translation style, prefers the form of four-character sentences and concise expression. Therefore, when determining the attribution of the translation of some texts, they assume that those texts are not translated by Zhi Qian because they do not meet this condition . For example, they thought T362 was not translated by Zhi Qian. In fact, this kind of preference can be only be found in some of the Zhi Qian’s translations. The revisions in Faju jing and Taizi ruiying benqi jing show that in some of his translations, Zhi Qian tried to make the Chinese texts closer to the Indian language, and tried to use scattered and changeable sentence patterns. Therefore, more possibilities need to be taken into account when discussing the style and the attribution of the Chinese Buddhist translations.