From proto-Joola to proto-Bak

• Main Author: Segerer, Guillaume
• Other Authors: Langage, LAngues et Cultures d'Afrique Noire (LLACAN), Institut National des Langues et Civilisations Orientales (Inalco)-Université Paris Diderot - Paris 7 (UPD7)-Centre National de la Recherche Scientifique (CNRS)
• Format: Conference Object
• Language: English
• Published: HAL CCSD 2012
• Subjects: linguistique Niger-Congo reconstruction comparaison; [SHS.LANGUE]Humanities and Social Sciences/Linguistics; Bak; North Atlantic
• Online Access: https://halshs.archives-ouvertes.fr/halshs-00757402

linguistics Niger-Congo reconstruction comparison So-called Atlantic languages consist of two main clusters, that is North and South, whose position in the Niger-Congo phylum is not clear yet. In addition, some of the languages traditionally classified as Atlantic are best considered as isolates, as long as no convincing evidence of their affiliation has been provided: these are Limba and Gola (for which we have a few data), and also probably Nalu and Sua (for which only short wordlists are available). The Southern cluster groups together the languages traditionally labelled as 'Mel' minus Gola. The North-Atlantic cluster, for which we propose to restrict the label 'Atlantic', may be in turn divided into two main branches: Northern proper (which includes Fula-Sereer, Wolof, Tenda, NyunBuy, Cangin and Jaad-Biafada) and Bak, which groups together the Joola cluster, Manjaku, Balanta and Bijogo. Each one of the aforementionned low-level labels corresponds to a non-controversial genetic unit. At a higher level, the Bak branch offers good perspectives of reconstruction using the standard comparative method. The present papers aims at presenting concrete results concerning the largest cluster within the Bak branch, namely the Joola cluster. I will show that despite continuous contact over the last centuries blurring the phylogenetic signal, solid sound correspondances allow to propose a subgrouping quite different from that obtained with mere lexicostatistics. Beyond the Joola case study, I will present some tentative but promising results that might lead to the reconstruction of Bak itself. In particular, the position of Bijogo as a marginal but legitimate Bak language will be assessed. These results are based on the analysis of nearly all possible lexical sources on the Bak languages, which total more than 45,000 lexical entries. This material is almost entirely available online thanks to the RefLex project, which also provides comparison-dedicated tools designed to help finding cognates and manage comparative ...