| Corpus-based Translation Studies (CTS) have made great achievements since Baker (1993) initiated the use of corpora in empirical studies of authentic translated texts. CTS incorporates both quantitative analysis and qualitative interpretation through which the theoretical proposals in translation studies are better supported with empirical evidence. With the development of CTS. Translation Universals have become an important topic of research. As defined by Baker (1993:243), Translation Universals are the typical or universal linguistic features in translated texts rather than in native languages which are independent of translation pairs. In line with Baker, many other scholars hold similar views toward TUs. A growing number of those universal features of translational languages have emerged and been discussed by researchers, such as Simplification, Explicitation, Source Language interference, Leveling-out and so on. Nevertheless, TUs-oriented studies so far are largely confined to English translations and a number of closely related European languages. As is known to all, English and Chinese are two "genetically" distant languages which belong to two distant language families, thus evidence from distant language pairs like English and Chinese will arguably be more necessary and convincing in the discussion of Translation Universals. In addition, previous research on Translation Universals hypotheses has been directed predominantly to the features at the overall level in a particular translational language. Little attention has been paid to the commonalities and differences of translational language in specific registers and genres, which has already led to many contradictory conclusions in the discussion of the universal features. As one of the major registers or genres in modern times, news reports have become more important in linguistic and translation studies. The TUs hypotheses should also hold in this register as defined. In the thesis, the writer will try to testify those hypotheses in translational Chinese and translational English news.This thesis will make use of four comparable corpora, namely LCMC (The Lancaster Corpus of Mandarin Chinese) vs. ZCTC (The ZJU Corpus of Translational Chinese), FLOB (The Freiburg-LOB corpus of British English) vs. COTE (The corpus of Translational English) in the register of news to carry out both comparable and cross-lingual contrastive analyses. With the help of corpus linguistic tools and methodologies, we will first analyze the distinctive linguistic features of translated Chinese news in contrast to non-translated Chinese news, and pinpoint the distinctive features of translational English in relation to non-translational English news based on the above mentioned comparable corpora in Chinese and English. Second, we will make cross-lingual contrastive investigations to the universal features of translational Chinese and translational English news, with the aim of testifying the hypothetical universal features in translational variations of Chinese and English.This thesis investigated a range of specific linguistic parameters such as STTR (Standardized Type-Token Ratio), lexical density, ratio between content words and function words, mean sentence length, mean paragraph length, mean sentence segment length, bei passives and so on. Investigations show that linguistic features observed in the register of news share some commonalities with the previous hypothetical universal features in translational variations of Chinese and English. In terms of ratio of content words, both translational Chinese news (LL=62.40, P=0.000) and translational English news (LL=0.82, P=0.364) show lower ratio compared with native Chinese and English news, which can be regarded as evidence in support of Simplification hypothesis. As for function words, the ratio of function words in both translational Chinese (LL=50.34, P=0.000) and translational English news (LL=585.45, P=0.000) are higher than that in native Chinese and English news, which can be explained as a factor supporting Explicitation hypothesis. And with respect to lexical density, LD in both native Chinese (LL=62.40, P=0.000) news and native English (LL=411.79, P=0.000) news are higher than that in translational Chinese and English news, which helps to explicate Simplification hypothesis in TUs.Nevertheless, there are still some differences that do not appear to support the TU hypotheses in the present study:the sentence length in translational English news is longer than that in native English news, which goes against the observation of Laviosa’s (2002:77). who points out in his own study based on comparable English corpus that the mean sentence length of translational English news is shorter than that in native English news. Passive constructions in the thesis are in favor of hypothesis of Source Language interference. But the proportion of bei passives in native Chinese news are fewer than that in translational Chinese news, on the contrary, the proportion of bei passives in native English news are higher than that in translational English news, suggesting that Source Language interference may be influenced by specific languages. |
PDF Full Text Request