Abstract
In this paper we comment on modern corpus linguistic methods in the context of German as foreign language teaching, driving on the Co-Occurrence Database CCDB, or more specifically its modules SOM and CNS that are being developed at the Institute for German Language in Mannheim (IDS). The underlying research is based on the notion of language structures as entities given not a priori but emerging on the basis of empirical analysis of very large amounts of real language data instead.
One of the postulates is that the language use in parole is constitutive for both, syntactic structures and - in reference to these - denotative and connotative aspects of language entities, with words being a particular case of these. Unsurprisingly, these aspects are posited here to be central pillars of vocabulary acquisition of any foreign language.