What's Hard in German? WHiG: a British learner corpus of German

Abstract

This short paper reports on the construction of a freely available learner corpus of advanced British English undergraduate learners of German, which was developed at Bangor University (United Kingdom) and the Humboldt-Universität zu Berlin (Germany). The corpus, What's Hard in German? (WHiG), can be compared to its sibling learner corpus, Falko-L2, and to the native speaker corpus, Falko-L1, using a creative commons user licence or a specifically designed online corpus platform interface, Annis. The main aim of the WHiG project was to collect typed-up essays from participants who achieved a CEFR proficiency of level B2 or higher in the metadata obtained, and whose essays are subsequently POS-tagged, lemmatised and error-tagged. A multi-layer annotation offers researchers the opportunity to work with the collected data and even to provide their own layers of annotation, which in turn can be made available online.