A Relevance-Theoretic Account Of Mistranslation

This thesis aims to analyze mistranslation from cognitive pragmatic perspective within the framework of Relevance Theory. For long there exist disputes over the interpretation of mistranslation and only few of the claims go beyond the limits of empirical deliberations. Thus, it is necessary to propose original solutions and move on the theoretical plane.Relevance theory is primarily a principle that governs human communication and cognition, by which translation is seen as an activity of communication in the framework of Relevance Theory and the translator plays a double role and functions as the midway transferring point which is restrained in optimizing relevance. Aside from the code model of translation, this thesis brings out two basic models of inferential translation in the light that there exist two types of translational constituents: univalent and bivalent ones.Another point is that translation is seen as interlingual interpretive use, for its intention is to restate in one language what is said or written in another language. In relevance theory, the core of communication is the seeking for optimal relevance, which is applicable in the process of translation as regarded part of communication.As an interlingual interpretive use of language, translation is also constrained by the notion of faithfulness and the translator should make the translation resemble the original closely enough in relevant aspects. But the relevant aspects are not clearly defined. Interpretive resemblance itself provides no answer, which is proved by example of Chinese poem.However, the answer could be found in the relevance theoretic account of faithfulness in the interpretive use of language. The translation should offer adequate contextual and cognitive effects in a manner that yields the intended interpretation without putting the audience to unnecessary processing effort. The translator should make assessment of the IE (intended effect) in the original language and then try to convey that to the audience. But the AE (achieved effect) not necessarily equals IE because extra effect is achieved by extra processing effort. Thus, the El (intended effort) and the E2 (actually consumed effort) not necessarily or possibly equal each other for the two variables are comparative more than quantitative. The author of the thesis makes comparison of the two variables in translating literary works and technical worksrespectively.Meanwhile, formulas are devised relying on the relationship between El and E2, and the relationship may be quantitative if questionnaires are specially designed for the rational common audience (readers) and on a fairly large scale. Mistranslation of each condition naturally falls into one of the categorizations owing to the continuity of the division.The analysis will be of illuminating value to translators in several perspectives. Optimal relevance of the original texts and that of the translated texts is the ultimate pursuit of translating behavior. The dilemma of whether and to what degree should the translator explicates the implicatures vanishes, and the disputes among the different strategies as foreignization and domestication are compromised. The translator can make self-elevation firstly by identifying the audience's cognitive environment while aiming at optimal relevance. Meanwhile, the thesis evokes new perspective on relevance theory and takes logic and quantity into analysis, which would be illuminative to software programming and machine translation.Finally, new strategies are available and suggestions on self-evaluation of translators are provided.