Bringing the learner back into the picture: Corpora and elicitation in transfer studies
(2008) 26th AESLA (Asociación Española de Lingüística Aplicada) Conference — Location: Almería (3.April.2008)
Files
No attached file found for this publication.
Details
- Authors
- Abstract
- The use of corpora, i.e. large collections of authentic text, has made it possible to cast a new light on the phenomenon of transfer over the last few years. On the one hand, learner corpora have enabled researchers to make claims about what problems learners actually encounter when using a foreign language. On the other hand, bilingual corpora have made it possible to carry out careful contrastive analyses and thus avoid hasty conclusions about transfer (see e.g. Schachter [1974] and Zhao’s [1989] and Kamimoto et al.’s [1992] reinterpretation of her study on the basis of contrastive corpus data). The Integrated Contrastive Model (cf. Granger 1996 and Gilquin 2000/2001) has been specifically designed to investigate transfer on the basis of corpus data. It relies on the combination of two types of analysis, namely contrastive analysis and contrastive interlanguage analysis. While the former consists in the comparison of two languages (in the form of original languages, or source and translated languages), the latter compares varieties of language (native and non-native, or several non-native varieties). The assumption underlying this model is that differences between the learner’s mother tongue (L1) and the target language (L2) are likely to lead to problems in the interlanguage, the learner transferring the L1 pattern into the L2. The model, however, takes little account of what happens in the learner’s mind. In particular, it establishes equivalences between L1 and L2 by means of texts produced by expert (native) writers and translators, and neglects the (sometimes incorrect) equivalences found in the learner’s mental lexicon. In this paper, I will plead for a more central position of the learner in the Integrated Contrastive Model – and in transfer studies in general. Following Kellerman (1977, 1979), it will be claimed that the learner’s perception of the target language is an important issue to consider, and it will be shown how the Integrated Contrastive Model could benefit from the combined use of corpus data, which reflect the learner’s performance, and elicitation data, which give a more direct access to his/her competence. Thus, instead of relying on bilingual corpus data to compare L1 and L2, the method proposed here uses elicitation tests which aim to identify the equivalent of an L1 item in the learner’s L2 mental lexicon. The use of the expression in fact by French-speaking learners will serve as an illustration. It will be analysed by means of the original Integrated Contrastive Model, as well as the revised, cognitively-oriented model, in an attempt to detect and explain possible traces of transfer. The comparison of the two analyses will show that the combined use of corpus data and elicitation data may explain cases of transfer which would be unaccountable on the basis of corpus data alone.
- Affiliations
Citations
Gilquin, G. (2008). Bringing the learner back into the picture: Corpora and elicitation in transfer studies. 26th AESLA (Asociación Española de Lingüística Aplicada) Conference, Almería. https://hdl.handle.net/2078.5/225205