The aim of this paper is to explore the concept of comparability in corpus-based Interpreting Studies and, more in particular, the risk of using comparable components of parallel interpreting corpora. Quite a number of studies based on such an approach have yielded inconclusive results, which could be due, it is argued, to the fact that the comparable components are drawn from plenary sessions of the European Parliament (EP), in which the groups generating the data share a working environment and could therefore influence one another.
To investigate the potential linguistic convergence that is likely to result from this mutual influence, both a theoretical and an empirical approach are taken. The theoretical approach seeks to determine whether EP interpreters and members of the EP could be analysed as constituting one single discourse community, according to the criteria put forward by Swales (1990). The empirical study of three discourse markers, based on data from a parallel corpus of EP interpreting and a comparable corpus of British parliamentary debates, aims to yield evidence of linguistic conversion.
