建構中的經典及其意義──《成實論》的漢譯、弘傳及其梁代義疏

本論文從《成實論》在中國的翻譯、弘傳及義解等三個面向，探討漢譯佛典的文獻性與詮釋歷史互相滲透的交織關係。《成實論》在鳩摩羅什手執胡本，口自傳譯202品「原譯本」之後，迭經曇影重譯並以「五聚」結構重整為「曇影本」，復有僧嵩細分「念」與「憶」譯語改定的「僧嵩本」。至梁代智藏《成實論大義記》稱曇影舊本及僧嵩新本「兩本俱行」時，此二本已同時流通近百年。梁代成實師更將兩本整合為「整合本」。臺灣國家圖書館藏敦煌寫卷《成實論義記卷中》、乃至《大正藏》所見《成實論》底本皆已是「整合本」。
《成實論》本身先前即有部派與大乘中觀思想混融的疑慮，漢譯羅什原譯本經「曇影本」、「僧嵩本」至「整合本」的整理，至此已難見其原貌。此外，〈發聚〉究為羅什原譯的內容，抑或曇影據四諦聚整理的導讀，雖無法確切判定，但其中所載部派議題的十論，或有為法三相、三假、二諦、念與憶、七實法等議題，都與梁陳時期成實師關注的議題高度一致。而這些現象也說明佛典的文獻性與詮釋歷史受傳譯者意識形態介入的影響。要求返回原初佛典固然難以達成，亦毋須太早放棄重建經典與追求真實的可能。經典結構的封閉性與詮釋循環的開放性，使它們恆處於建構與解構的一體兩面之中。

This article aims to demonstrate the intertwined relationship between the documentary and interpretive histories of Chinese translations of Buddhist texts from three perspectives: translation, transmission, and interpretation of *Tattvasiddhi in China. After the original translation of 202 chapters, the *Tattvasiddhi was then collated by Kumārajīva’s disciple T’anying曇影into the“T’anying text”with the structure of the“five aggregates”, and then the translation of“manaskāra”and“smṛti”was subdivided by Sinsung僧嵩, and the text was changed to the“Sinsung text”. By the time Zhizang’s Cheng shi lun da yi ji”of the Liang Dynasty stated that“both texts were in circulation”, these two texts had been in circulation for nearly a hundred years. In the Liang Dynasty, Master Chengshi combined the two texts into the“Consolidated Text”. The original text of the Chengshi treatise, as found in the Dunhuang manuscripturs of the Cheng shi lun yi ji (Taipei 131), was already a“consolidated text”that merged the T’anying text with the“Sinsung text”.
*Tattvasiddhi itself had previously been suspected of mixing Abhidharma Buddhism philosophy and Mahayana Mādhyamika philosophy, it is diffi cult to identify the original Chinese translation of the first Chinese translation of Kumārajīva. Although it is not possible to determine whether the“Introduction”chapter was originally translated by Kumārajīva or whether it was compiled by T’anying which based on the Four Noble Truths, the ten treatises on sectarian issues contained therein, such as the three phases of the Dharma, the three prajñapti, the two Noble Truths, smṛti and manaskāra, and the seven realities, are all highly consistent with the issues of concern to the Interpreters of *Tattvasiddhi in the Liang-Chen period. These phenomena also show that the documentary and interpretive history of Buddhist texts is influenced by the ideological intervention of the translators.
While it is difficult to return to the original canon, it is not necessary to give up too soon the possibility of reconstructing the canon and pursuing the truth. The closed nature of the canonical structure and the openness of the interpretive cycle keep them in a constant state of construction and deconstruction.