Friederike Lüpke gives a guest lecture within the research area Linguistic structures. All interested are welcome!
In this talk, I take you on a journey to the multilingual research area in Senegal’s southern province where I have conducted research since 2010. Based on fieldwork data, I introduce this small-scale multilingual area and briefly introduce prevalent language ideas and their repercussion on what a named language or register is. This is a necessary precursor to presenting an area of grammar offering great insights into the organisation of multilingual speech: nominal classification systems.
I will introduce the morphosyntactic and semantic properties of nominal classification systems first from the perspective of named languages inherent in grammatical descriptions, focusing on Baïnounk (Nyun) languages and contrasting them with those of closely related Buy and more remotely related Joola languages. Many inhabitants of Casamance are multilingual in languages of all three families (and additional ones) and in Wolof, another distant Atlantic relative. All these languages have systems of nominal classification of varying complexity. Three languages widely attested in linguistic repertoires do not exhibit noun classes. These are Kriolu, a Portuguese-based creole, the Mande language Mandinka and the Romance language French.
Noun class systems in the area exhibit many semantic and morphosyntactic similarities, yet exhibit great variation within and across speakers. Noun class paradigms often show family resemblances yet are rarely directly equivalent in terms matter or patterns involved. I argue that we can makes sense of these convergences by adopting a perspective concentrating on linguistic features and how they are combined in multilingual speech. Multilingual speakers of weakly reified and unregulated languages build noun phrases and agreement patterns using bootstrapping strategies based on their entire multilingual knowledge and are prone to priming effects resulting from the shape of their repertoire, the discourse context and preceding speech forms. This means that rather than describing similarities in static terms, as language contact phenomena, we can capture many of them are more aptly in terms of dynamic and variable creative use of linguistic features that may be strongly of weakly conventionalised because of their overall frequency.