Parallel texts and corpus-based contrastive analysis

In this chapter we explore a corpus-based approach to contrastive analysis, in which the tools and concepts of corpus linguistics are applied to contrastive studies (Altenberg & Granger 2002: 7). Following this approach, we start with few assumptions and examine the frequency data generated by the application of a parallel concordancer to a translation corpus. We use this data to guide our investigations into similarities in the paradigms of go and aller, focusing on collocations and idiomatic uses. We show how collocate frequency information is used to determine equivalent collocations in two languages and to distinguish different usage patterns.