Levelling Out In Translational Literary Chinese:A Corpus-based Study Across Multiple Language Pairs

The investigation of translation universals has been an important area of inquiry in corpus-based translation studies,with one prominent universal being the phenomenon of levelling out,also referred to as convergence.The levelling out hypothesis posits that sets of translated texts in a corpus display less variance than comparable sets of non-translated texts.This hypothesis has yet to be adequately examined in translation studies,and still less is relevant research on translational Chinese.The present study conducts an examination of the levelling out hypothesis in the context of contemporary Chinese translations based on a self-constructed corpus of 43,600,000 characters,which equates to approximately 30,000,000 tokens.This corpus is comprised of seven comparable sets of contemporary novels,including a reference corpus of native Mandarin Chinese novels and six subcorpora of novels translated into Chinese from various languages.Each subcorpus includes 30 novels,most of which were published between 1980 and 2021,with the original texts produced within the same time frame.The reason for choosing works in these six languages during this period is due to the fact that after year 1978 when the Third Plenary Session of the 11th Central Committee of the Communist Party of China was held,China ushered in a new era of reform and opening up,and cultural exchanges with other countries or regions have also become more frequent and inclusive,thus triggering a new wave of translation.Changes in the usage of vernacular Chinese were witnessed,standards of selecting works to be translated were improved,the number of translations of foreign literary works in China was increased,and the genres and themes have also become more diverse.This is particularly the case with the translations of literary works in these six languages.Accordingly,exploring the linguistic features of the Chinese translated language as a foreign literary medium would help capture and understand the nature and characteristics of translation activities.To analyze the corpus,25 linguistic parameters,including but not limited to STTR,the ratio of content words to functional words,the proportion of conjunctions,the ratio of adjectives to verbs,as well as the length and standardized frequency of syntactic structure,were adopted for data mining and interpretation.Besides ANOVA test for between-groups difference,principal component analysis and cluster analysis were applied to the two datasets to investigate the cross-linguistic levelling out.The research questions are as follows:(1)Does the levelling out phenomenon distribute consistently across varying levels of linguistic features of literary Chinese translated from different languages?(2)To what degree does levelling out occur in translational Chinese as compared with non-translational Chinese?The results broadly support the levelling out hypothesis,particularly when considering nine parameters,such as lexical diversity,informational load,and connectivity.On the other hand,the remaining lexical and syntactic features,including the length-frequency distribution related to the "principle of least effort" typically found in non-translational languages,show that texts translated from different languages are homogeneous,yet they do not differ from the native ones to a statistically significant extent.When considering the results from the multivariate analysis in conjunction,it was confirmed that translational Chinese as a whole tends to cluster as a relatively compact group distinctive from nontranslational Chinese.This phenomenon indicates that translational languages follow certain patterns governed probably more by translation norms in the target culture than the typology of source languages.Meanwhile,it should be noted that the Chinese texts translated from Japanese are found significantly deviant from other translational Chinese in such aspects as the sentence length and the frequency of adverbs,which suggests that the underlying influence of source languages is worth further investigation.As regards the research significance,on the theoretical level,this study focuses on linguistic features and generalizes the regularities and characteristics of contemporary literature translation in Chinese."With corpus-based research ideas and methods,it verifies the cross-language levelling-out hypothesis as a translation universality,and provides reference information for describing the characteristics of Chinese literary translation in the new era.In terms of research methodology and application,this study has established a medium-scale corpus(tens of millions of words)and taken into account the random sampling criteria,representativeness,and balance of the corpus to ensure the quality and size of the data.The constructed corpus can be used for exploring language features,literary translation strategies,translation and social norms,and other related areas of investigation.In terms of teaching,the systematic differences between translated Chinese and original Chinese revealed in this research can provide reference information for translation theory teaching and application.It would help translation students to further understand the cross-cultural characteristics and social aspects of translation behavior,and encourage a dialectical view of the phenomenon of translationese.