A Fuzzy TQA Model Specific To Legal Texts

These years when translation services are gradually industrialized and translation quality is strictly required translation quality assessment (TQA) has attracted more and more attention and a lot of researches have been diverted to the TQA methods which can not only provide a way to evaluate the quality of translation, but also serve as a guidance for the improvement thereof. Seeing that errors in legal translation has led to offences against law, this thesis is to build a fuzzy TQA model which can be used specifically to assess the quality of legal text translation and thus help to make legal rules understood and observed exactly and accurately in a different language environment.Williams (2004) has classified the approaches to TQA into two categories:quantitative and non-quantitative. In application the former may reduce the subjectivity and help to reach a higher degree of objectivity, which is the very kernel requirement of TQA. Since legal texts put more emphases on objectivity, a quantitative approach is more appropriate for their evaluation. Before the construction of a systematic TQA model a justified theoretical basis should be established. The previous researches show that the goal of legal translation is to produce texts which express the meaning accurately and achieves the legal effects intended by the author. This suggests two criteria for the translation of legal texts from the pragmatic perspectives. Therefore, the parameters to be assessed may be prepared respectively at two levels:propositional content and legal effect. A lot of researches also have been done by many influential scholars as to the equivalence of information and legal speech act. They form a significant contribution to the assessment of the parameters. According to Williams, almost all approaches to translation quality assessment are characterized by error categorization and analysis, the model built by the present study is no exception.In order to eliminate the dilemma arising in grading the parameters, fuzzy mathematics is introduced into the model. It saves the pains in making a decision on the precise score of each parameter and may achieve a relatively objectified and convincing final result. In the field of fuzzy TQA model construction, Fan Shouyi (1987), Feng Zhijie (1998), Mu Lei (1991) and other scholars in China have done many researches. Based on their study, especially those on membership function, grade set and weight assignment, a fuzzy operation applicable to the present study is designed. By grading each parameter at either level and put their score into the fuzzy operation an objective result will be obtained.Finally a sample evaluation is done to test the feasibility and the operation of the model. This test shows that the model set up by the present study is not only systematic, but also can serve as a practical guidance for the translation quality assessment of all other legal texts.