This project will make significant contributions to the digital infrastructure that is needed for Social Sciences & Humanities researchers to investigate the sign language of the Netherlands (Nederlandse Gebarentaal, NGT), which is lawfully recognised as of 1 July 2021 as one of the country’s official languages but not yet covered by infrastructure projects like CLARIAH.
One of the key problems in studying sign languages is that they lack a writing system. Primary data are in video format and currently need to be transcribed and annotated manually. Creating consistent annotations across datasets requires connecting annotation software (ELAN) to an NGT lexicon that does justice to the morphological, semantic, and phonetic characteristics of the language. Words in NGT and Dutch do not match in structure or meaning, making it difficult to build on lexical resources for Dutch.
The project, therefore, focuses on (i) a digital NGT lexicon with semantic and phonetic information, (ii) NGT corpora of adult language use as well as interactions between adults and children, and (iii) software for machine-supported annotation, using novel AI technologies. It connects these sign language resources to existing infrastructures within CLARIAH, enriching the CLARIN-K Centre for Atypical Communication Expertise (ACE) to maintain and bundle tools and resources for NGT.