A Study Of Chinese-English Translation Errors Based On The Translation Corpora Of Intermediate And Advanced English Learners

Translation competence is the target to the cultivation of translation talents.Translation errors,as an inevitable problem in learners'translation learning,can directly reflect learners' translation competence.This study takes the BFSU-GSTI corpus and PACCEL-W as research resources,exhaustively annotates the chosen data from the two corpora with error-tags under the classification of translation errors of functionalism and make comparative analysis between advanced learners and intermediate learners in terms of the distribution of translation errors,the characteristics of translation errors,and reasons that may cause errors from the perspective of text genre,four dimensions(language error,pragmatic errors,cultural errors,strategic errors)and two levels(binary and non-binary).At the same time,the long-standing translation errors in both corpora with high-frequency are discussed.The study found that(1)the translation error rate of advanced English learners in political and economic texts is significantly higher than that of other texts,and the translation error rate of intermediate English learners in literary texts is significantly higher than that of advanced English learners.(2)As the learner enters a more advanced learning stage,the distribution of translation errors in types has shifted from a linguistic error dominated distribution mode to an even distribution of linguistic,pragmatic,and strategic errors,and(3)the proportion of non-binary errors has increased significantly making the non-binarism the main error feature.(4)For both intermediate and advanced English learners,lexical deviation errors are the dominant error type with the highest frequency.According to this,this thesis puts forward some thoughts on how to construct a constructive translation teaching model aiming at cultivating translators' selection ability,and measures can be taken to clearly classify translation errors,to distinguish binary errors from non-binary errors,and to strengthen the context-based memory of low imageable words.