漢譯できない語をどうするか --唐の玄奘に託された「五不翻」說の再檢討--

Abstract

This article is a re-evaluation of what is traditionally called “Xuanzangʼs theory of the five kinds of untranslatable Sanskrit words, ” widely known in East Asian Buddhism. These “five kinds” are (1) words whose real meaning should not be known to the public, such as spells (dhāraṇī); (2) words that have multiple meanings such as bhagavat (i. e. “World-Honored One”); (3) words whose referent does not exist in China, such as “Jambu-tree; “(4) words that have traditionally been expressed as Chinese phonetic transcriptions for such as “anuttarā samyaksambodhi; “ and (5) words which benefit Buddhist believers when they remain in Sanskrit transcription, such as “prajñā.” After a critical examination of the issue, focusing on whether or not these five types of untranslatable words are Xuanzangʼs original idea, this paper argues that: First, the “Five Kinds of Untranslatable Words” sometimes contradicts how Xuanzang actually selected words to use in his translations into Chinese. Second, there is a clear gap between the date of Xuanzangʼs death and the first reference to the “Five Kinds of Untranslatable Words.” The earliest reference to this idea is found in a commentarial work composed by Chaowu 超悟in 789 CE, suggesting that the “Five Kinds of Untranslatable Word” may have appeared as late as 130 years after Xuanzangʼs death in 664. Third, the contents of the “Five Kinds of Untranslatable Words” are also found, albeit partially, in the work of Chinese exegetes that pre-date Xuanzang. In conclusion, this paper calls into question the traditional attribution of the “Five Kinds of the Untranslatable Words” to Xuanzang.