Designing a Multimodal Corpus of ELF in LSP: Sustainable Data Collection and Analysis

ELF and Language for Specific Purposes (LSP) appear to be intrinsically intertwined phenomena in the interpreting practice (see Albl-Mikasa 2010; 2013) and deserve to be one of the core topics in formal interpreter training. To explore and model the impact of ELF in LSP on interpreting students, we have designed a case study adapted from Bale (2013). Here, authentic video recorded LSP conferences in ELF scenarios are used as teaching material in MA level English-to-Italian SI classes. Primary teaching objectives are exposing students to real life situations and raising their awareness on quality benchmarks. In this case study, teaching material is being gathered in a multimodal corpus along with the audio recordings and transcriptions of the student’s interpreted text, and an in-depth terminology analysis of the source texts. The resulting corpus is designed as a dynamic unit and its benefits are at least threefold: It can help i) modeling the impact of ELF in LSP on the students’ performance, ii) providing new insights for developing teaching material tailored to the students’ skills and learning objectives, and iii) getting an improved understanding of LSP patterns in ELF. To triangulate all data sources in our corpus, we have integrated an ad hoc analytical workflow in the corpus design. Our workflow includes two steps, tackling text preprocessing and analytics respectively. This analytical paradigm provides for data scalability and result reproducibility, and helps reducing the bias arising from a purely human performed analysis. Some preliminary results will be presented.