Investigating corpus-assisted translation teaching: A pilot study

This contribution investigates how and the extent to which parallel corpora affect students' translation quality. The study presented is based on a large-scale web-based Chinese-English parallel corpus, that is, the Hong Kong Parallel cum Comparable Corpus (HKPCC), which comprises three text types (legal documents, news texts, parliament proceedings). The corpus has more than one hundred million Chinese characters with corresponding parallel English texts aligned at the sentence level. Since the study was set up as an empirical comparative study, a control group and an experimental group were used to test how students would perform translating in a corpus-assisted environment. Both groups were given the same piece of Chinese news text, which they were required to translate into English. The control group used conventional resources (i.e. monolingual and bilingual dictionaries, thesauri) for the translation assignment and the experimental group used a parallel corpus (i.e. an HKPCC excerpt). Data analysis of the student translations shows that the experimental group performed better in a number of aspects, including collocations, phraseology, spelling, terminology and word choice. Holistic scores given by two independent examiners also show that the experimental group produced better translation output than the control group. Based on the findings of the study, it is argued that parallel corpora are a useful resource for both translation teachers and translation students. In addition, it is argued that the corpus-assisted approach can be an innovative and effective means to complement the traditional translation teaching/learning approaches