The TIGR corpus of spoken Italian was gathered in the Swiss cantons Ticino and Grisons in 2021-2022 to study the expression of the category 'information source' in spoken Italian. It documents face-to-face interactions in various kinds of non-experimental settings: dinner and lunch table conversations, cooking, lessons and tutoring encounters, interviews, 23.5 hours in total. The interactions were recorded with two camcorders and pocket audio recorders equipped with clip-on microphones and transcribed using the ELAN editor. The informants filled in a short sociolinguistic questionnaire. This work was completed within the realm of the 'InfinIta' project (SNSF grant no. 192771). A next phase of data processing started in 2023, aiming at depositing the TIGR corpus on the LaRS @ SWISSUbase repository to make it available to the scientific community for academic research purposes ('InfinIta' and USI ORD grant 'ShareTIGR'). In that phase, all sociolinguistic data and metadata were compiled and inserted into a set of tables; the basic transcripts generated in ELAN were processed to yield playscript style TXT-transcripts; the basic transcripts were tokenized and converted to yield XML-transcripts according to the TEI/ISO 24624:2016 standard for the transcription of spoken language; the multimedia files were de-identified using standard procedures and according to the preferences expressed by the study participants in their declarations of consent.
TIGR on SWISSUbase includes a dataset 'General documentation', a dataset 'TIGR transcripts', and separate datasets for each recorded event, which contain multimedia files and transcripts. The event datasets are available in two versions in most cases, which differ as to their multimedia files: the 'full' version contains the single audio files and mute video files recorded by the various devices alongside with a 'combined' split-screen video with mixed audio, whereas the 'light' version contains only the 'combined' video file.
