A Study On Chinese And Western Models For Translation Quality Assessment

Translation quality assessment (TQA) plays an increasingly important role in applied translation studies. This thesis classifies, criticizes and reflects the development of TQA at home and abroad, focusing on Juliane House's and Si Xianzhu's models based on register analysis. Comparative and empirical research methods are employed in this thesis with an aim to improve the current model for TQA and to make it more objective and applicable to Chinese-English and English-Chinese translation.First of all, this thesis reviews the western and Chinese relevant studies in the field of TQA. The four approaches or methods summarized by J. House are first presented. The first three approaches do not establish a clear criterion for translation criticism and quality evaluation, nor is a model for TQA constructed successfully. As a representative of the fourth approaches, J. House successfully constructs a model for TQA.Then this thesis introduces and criticizes J. House's model. It is J. House who undertakes the first attempt to establish a model for TQA in the field of translation criticism. She bases her model mostly on systemic functional linguistics, the theory of text analysis and pragmatics, employing as an analytical tool the register analysis along a series of situational dimensions or parameters (or register variables). According to this model, functional equivalence is a vitally significant criterion and a basic foundation for TQA, which is focused on the semantic/ideational and pragmatic/interpersonal equivalence. As regards the operation of the model, the analysis of the original will be done first, followed by the comparison of the original and translation. During the comparison, genre will be taken into account, and the situational dimensions will be analyzed through lexical, syntactic and textual means. Meanwhile, mismatches of meaning or function between original and translation will be detected and described. Then social value judgment related to the mismatches is conducted. Finally, the quality of translation is stated, during which the translation types are justified. In House's model, a number of valuable concepts are distinguished, such as language function versus textual function, text typology, translation types (overt and covert transition), cultural filter, etc.As House's case studies are all English-German translation samples, its applicability to the Chinese-English translation evaluation remains to be suspicious. Thus, this thesis further has a case study to test the availability of House's model to the Chinese-English TQA. Then the model is criticized. In general, this model confines itself mainly to the linguistic equivalence without attaching sufficient importance to the social value judgment. Besides, the lexico-grammatical analysis at the micro-level does not strictly describe the transitivity, modality and information structure in systemic functional grammar, resulting in the randomness in analysis.With respect to the domestic studies, Si Xianzhu makes the greatest achievement. On the basis of the criticism of House's model, Si Xianzhu constructs a new model completely approached from systemic functional linguistics. Within a verbal behavioral framework of interaction among form, function and situation, this model reveals the ideational and interpersonal mismatches at the micro-level; then it reclassifies the mismatches and makes value judgment on them; finally a statement quality is made by taking account of the text type. As Si claims, his model employs an all-around methodology, both "top-down" and "bottom-up", both qualitative and quantitative, both descriptive and hermeneutic.In lexico-grammatical analysis, this model adopts the transitivity and modality systems, which can examine the realization of ideational and interpersonal functions respectively. Besides, this model emphasizes the reclassification of and value judgment on the mismatches by taking account of verbal communication factors, context and texture. As a result, the mismatches in the ideational and interpersonal functions will be detected more objectively and the evaluation will be more scientific than House's model. However, this model cannot be free from defects and criticisms such as the sampling and insufficient quantitative studies, etc.This thesis also makes suggestions for the western and Chinese models for TQA with an empirical study made on House's model. But no empirical study is made on Si's model because of the extreme complexity of the operation of this model, which still needs to be further revised. Finally, the future direction of the development of TQA is also discussed.