| Aside from their intrinsic value for the purposes of language description, corpora can be used in language courses under the broad umbrella term of "data-driven learning", or DDL (Johns 1990). They can be used as a teaching aid or learning tool, but also as a reference resource, particularly for writing and revision / error-correction, with an increasing body of empirical evidence (cf. Boulton in press) largely concerning students needing English for academic purposes. Corpora can also be used as a reference resource in translation, and here the DDL-related research focuses largely on translation trainees (e.g. Beeby et al. 2009). However, as Zanettin (2009) points out, translation is also a staple activity in many undergraduate language programmes where learners may also benefit from corpus consultation, and not just to help with the immediate translation assignment. The techniques involved build on many cognitive and metacognitive skills (O'Sullivan 2007), and once mastered, can be applied to other language courses and a variety of future language needs. The main research question in this paper is thus: Can general language corpora be used by non-translation students for translation purposes? The students in this study are enrolled in a translation course in the third year of a degree in English at the distance education centre of a French university. They have no prior experience of corpus use and are expected only to work only with freely available on-line monolingual corpora of contemporary English. The simple interface provided by Mark Davies (http://corpus.byu.edu/) to large corpora of British and American English (100 million and 400+ million words respectively) is widely used for such purposes, with over two thirds of all users declaring their primary interest as language learning, teaching or translation. The sites are suitable for novice users to navigate and are accompanied by tutorials and help features; this is highly desirable in the distance teaching context where face-to-face input is not an option, and means that the course itself can keep the introduction to the basic concepts and techniques to a minimum. After that, the constraints of the distance context play to the strengths of constructivism as the students explore the corpora on their own; though email contact with the teacher is possible, and there are discussion forums to facilitate peer-to-peer collaboration, these are generally under-used. Following an earlier pilot study, the results presented here are based on on-line examinations where the methodology component requires the students to choose sections of a previously unseen text and demonstrate and explain how they use corpora to solve the problems encountered in context. Comparing data from two sessions shows how the students come to grips with corpora for translation, and allows a qualitative analysis of individual performance on the various techniques used with greater or lesser success. These data are backed up by questionnaires submitted after the examinations to gain feedback from both successful and less successful students. Particular attention is accorded to corpus use beyond the usual concordance lines, including frequencies, register distributions, collocates lists, word comparisons, and so on, which allow the learner to ask not just 'can I say this?', but 'is this appropriate in this translation context?'. |
Récupérer au format
