This paper studies the two polysemes yì 異 and è 惡found in the three Chinese translations of the Vimalakīrtinirdeśa and their Sanskrit equivalents. The words yì and è were used to translate Sanskrit prefixes vi- and dus-, but could also be used to translate other Sanskrit words. In fact, most Sanskrit morphemes were polysemous, as were the Chinese. One might wonder how the translators translated Sanskrit polysemous morphemes into Chinese. How did they choose appropriate Chinese words for translation since the equivalent Chinese words were also polysemous? The paper explores the various meanings of yì and è and their corresponding Sanskrit morphemes or terms. Five corresponding patterns are found for each of the two polysemous words. Moreover, the author has found that Xuanzang used what Nida and Taber (1982) called ＂formal correspondence＂ while Kumārajīva employed ＂dynamic equivalence＂ to translate the Sanskrit terms. The former would use loan translation to translate Sanskrit morpheme by morpheme into Chinese, and created some novel Chinese words which never entered the Chinese lexicon. However, the latter attempted to conform to the existing Chinese lexicon, and did not deliberately follow the Sanskrit word structures, and thus his translation would be favorably accepted by the Chinese people.
