The GeWiss corpus

Research on academic language has flourished in recent years, including academic German. The corpus resources available for larger, empirically based research projects remain, however, limited, even with regard to written academic language, and they are practically non-extant for spoken academic language. A detailed, empirical analysis of linguistic conventions and formulaic language used in (oral) academic communication is, however, all the more important in a day and age where our academic landscapes are becoming ever more internationalised. GeWiss aims to lay a foundation for such research: With GeWiss we are currently constructing a parallel corpus consisting of spoken academic language data from German, English, and Polish. Our contribution outlines the corpus design and the methodological procedures used for the construction of GeWiss. The project focuses, at least initially, on two key genres: Academic papers / student presentations and oral examinations. These are recorded, transcribed and stored in a searchable database. The corpus will comprise native speaker data from Polish, English, and German academics and students, as well as German as a Foreign Language (GFL) data of nonnative speakers of German. These are recorded at all three partner institutions. GeWiss will therefore be the first corpus comprising GFL learner data. The paper focuses on corpus design as well as data collection and transcription. It also discusses selected research questions for which GeWiss can serve as an empirical basis.