SOAS Research Online

Abstract

Languages are not discrete, nor do they exist as such separate entities in a multilingual’s mind – they interact (Grosjean 1989; 1994; 2010; Cook 1991; 1992; 2003; Cook et al. 2006; Kellerman & Sharwood Smith 1986; Odlin 1989; 2003). Bilingual speakers of European contexts have been shown to drop specific information when compared to their monolingual counterparts as a result of crosslinguistic influences (Berthele 2012; Alferink & Gullberg 2014; Alferink 2015). Yet most research in this domain has focused on Western speakers of standardised languages (Henrich et al. 2010), limiting what we know about the reality of multilinguals of different contexts. Speakers of the Lower Casamance in Senegal, West Africa, for example, know an average of 6 languages and may use these languages on a daily basis (Juillard 1995; Tabouret-Keller & Juillard 2000; Dreyfus & Juillard 2004). The question thus arises whether the nature of the speakers’ linguistic system is still one of generality considering their linguistic repertoires. This study explores how placement events are expressed by multilingual speakers in the Lower Casamance by investigating the semantic information displayed multimodally. 18 multilingual speakers residing in the Lower Casamance participated in this study. The study targeted Joola Kujireray and French, the former using both broad- and fine-grained information (sit, stand, lay, and put), and the latter encoding broad-grained information (put). I describe the linking of thematic roles to syntactic constituents and provide a lexical semantic description and comparison of placement and removal verbs in Kujireray and French following a Cognitive Linguistics framework. The information expressed in co-speech gestures is then described. Findings reveal speakers’ preference of displaying semantically specific information in gesture, while expressing semantically general information in speech – contrary to Western bilinguals. The thesis concludes with a discussion of adapting gesture-speech models to account for multilingualism and multimodal patterns.