About

The recordings were transcribed according to a newly developed transcription protocol – highly urgent in times of rapidly progressing dialect loss! – to then be linguistically enriched using existing tools with information on the word type of the individual words (‘pos-tags’) and with information on the syntactic functions of the word groups and their interrelationships (‘parsing’).

Compared to existing data collections on Dutch dialects, the GCND is unique in that it contains only spontaneous speech. Since the dialect recordings represent a historical stage of the language (in the case of Franco-Flemish even the last testimonies of a now almost extinct language variety) and the recordings are now efficiently searchable, the GCND makes it possible to (i) map language change processes geographically, (ii) quantitatively investigate the functionality of dialect features, and (iii) detect new, previously unnoticed and thus unrequested structures. Audio, transcriptions and annotations are freely available and searchable online. The GCND thus constitutes an unprecedented historical dialect corpus.

A follow-up project (GCND+) has started in cooperation with the Instituut voor de Nederlandse Taal (INT) and LT3 – Language and Translation Technology Team (UGent) to extend the collection further north.

Using the GCND

The corpus can be accessed via the link below: https://gcnd.ivdnt.org

It can only be accessed with a username and password. Users employed by a university, a college or a research institute can log in with their organisation’s username and password. Users not associated with an academic institution can also access the corpus, but they must first apply for an account at www.clarin.eu.

Detailed information on the Spoken Corpus of Southern Dutch Dialects (GCND) can be found here.