Creating a corpus of bilingual language data to conduct linguistic research
What is it about?
A common way to learn more about people's use and knowledge of language is via a corpus, which is a large repository of language data. This book describes in detail the creation of such a corpus of speech from people who are bilingual in Welsh and English, namely the Siarad corpus which was collated by researchers at Bangor University, Wales (UK). It also details various linguistic analyses of data from the corpus, including code-switching (the use of multiple languages in the same conversation) and grammatical change.
Why is it important?
The book serves as a 'how to' guide for other researchers to emulate when building a linguistic corpus, including advice on finding and recording participants, on coding the data and on analysing the data. It contains a detailed description of our innovative auto-glossing system, which reduces the need for laborious manual grammatical glossing of language data. The book also describes findings on research in Welsh linguistics and bilingualism, including on variation in code-switching in the speech of Welsh-English bilinguals and on what the corpus can tell us about language change in Welsh. As a minority language that has endured despite long-standing pressure from the English language, research on Welsh is important and valuable as a way to understand what happens (grammatically and socially) to languages which are in intense contact with a majority language. Those who are interested in bilingualism beyond Wales will be able to extrapolate the findings to other similar linguistic situations, particularly those of minority languages.