Corpus Linguistics meets Contact Linguistics and Sociolinguistics: multimodal German–Polish digital linguistic corpus.

The objective of this paper is to present procedures of building, tool construction and annotation schemes of bilingual corpus LangGener, which is an innovative compilation of data influenced by contact linguistics and sociolinguistics. The multimodality of the corpus is related to the possibility of viewing the transcription and grammatical and sociolinguistic annotation of recorded original interviews and listening to them, because it contains 'recordings, transcriptions and annotations' (Allwood, 2008: 207). The corpus is part of the international LangGener project on changes in morphology and syntax in Polish and German in the speech of bilinguals over two generations. It contains seventy-eight hours of material, equating to 679,741 words. The material was collected during fieldwork in Germany and in the former German territories incorporated into Poland in 1945. The article describes the demanding procedures and the complex work on the recorded interviews, the tools used in the project, the annotations of the phenomenon of mutual Polish–German linguistic contact and dialectal remnants. In order to relate the results to the language biographies of the interviewees, sociolinguistic annotation was made in parallel, presenting the modes of acquisition and language use in the speech of bilinguals. The corpus also enables a general typologisation of language contact phenomena for Slavic languages in contact with Germanic languages. [ABSTRACT FROM AUTHOR]