A Corpus-based Approach to the Modelling of the Explicitation Process in English-Chinese Translation

Features of translation products provide evidence of translation processes. Apart from think-aloud protocols (TAPs) and the Translog approach that have been used for data elicitation and collection in process research, translation products have been shown to contain systematic features or traces that may as well shed some light on what happens during translation (Bell 1991; Mason 2002). This paper reports on a corpus-based study of the process of explicitation, which has been hypothesized as one of the features that distinguishes translated text (TT) from its source text (ST) as well as from non-translation (NT) in the target language (TL). A model of explicitation is proposed, focusing on its realization in relation to both ST and NT. While process research is usually associated with the investigation of what goes on in ST-TT conversion, explicitation through the subconscious use of optional linguistic elements (e.g. reporting that in English and 會[possibly] in Chinese) can only be observed in a comparable corpus that consists of TT and NT in the same language. The factors that motivate explicitation in the translation process are mainly inferred from a wide variety of genuine translations contained in a 1.8-million-word English-Chinese parallel corpus of popular science texts. As each ST in the corpus is accompanied by two Chinese TTs produced in Taiwan and China, respectively, it is relatively easy to determine whether a particular instance of explicitation should be treated as obligatory or optional-two main categories distinguished in the current proposed model. Variation in the use of the explicitation strategy is addressed in terms of three aspects: individual, institutional, and setting and readership.