Crossing boundaries: Using French Belgian Sign Language (LSFB) and multimodal French corpora for contrastive, translation and interpreting studies

Abstract The French Belgian Sign Language (LSFB) corpus is the cornerstone of a unique multilingual data system that includes four distinct corpora. The first is a reference corpus containing dialogical LSFB data produced by deaf signers, which is also translated into written French, and the second one is a comparable multimodal corpus of Belgian French containing dialogical data produced by hearing native speakers. The other two corpora are made up of interpreted data, namely a parallel bidirectional corpus of LSFB > French data produced by hearing bimodal interpreters and another unidirectional parallel corpus of French > LSFB co-interpreted data produced by hearing and deaf interpreters working in tandem. This paper aims to describe these four corpora and to provide an overview of previous contrastive research drawing on their data and applications which have been developed so far. A new study contrasting reformulation structures in semi-spontaneous LSFB dialogical data and co-interpreted LSFB data is presented in order to exemplify how these corpora can be further compared to shed light on unknown issues to date such as the specificities of co-interpretation.