par Guitang, Guillaume
Référence Annual Conference on African Linguistics (51-52: 8-10 April 2021: Online (The University of Florida))
Publication Non publié, 2021-04-10
Communication à un colloque
Résumé : In many contemporary Chadic languages, reduplication is a common productive process for deriving new words and encoding various morphosyntactic categories (Allison, 2012; Dougophe, 2015; Frajzyngier, 2002; Newman, 2000; Schuh, 1998; Viljoen, 2013). Newman (1990) has suggested that Proto-Chadic was likely to have used prefixal *CV- reduplication as the primary means for forming pluractionals (see Schuh, 2002 for a different view). Suffixal *-VC(V) reduplication has also been considered, though unavailingly, as a means for forming noun plurals in Proto-Chadic (Newman, 1990). While our reading of the place of the process in Proto-Chadic is still fragmentary, its evolution in present-day languages appears to be puzzling. One issue, for example, is the fact that some present-day languages do not have it as an active derivational and/or inflectional process. This is the case of Masa North (Melis, 2019) languages, and especially, of Gizey, the language under study here. Gizey contains only a few frozen reduplicatives in the noun and ideophone classes. Gizey thus stands out among Chadic languages in that active reduplication is neither used to mark pluractionality in the verbal class nor any other morphosyntactic feature in other classes. After a comprehensive description of the formal and semantic properties of frozen reduplicatives, I will argue that Gizey now mostly relies on syntactic repetition to mark pluractionality, unlike many Chadic languages which are known to do so via reduplication. The data presented here mainly include a) primary data (narrative texts) collected by the author and b) secondary data consisting of lexical items deriving from previous studies of Gizey (Ajello & Melis, 2008), Masana (Melis, 1999), Musey (Roberts & Soulokadi, 2019), Marba (Melis, 2006), Zimé (Vincent, 2000) or comparative wordlists like (Ajello et al., 2001).