Second, we present the results of a new phylogenetic study including all B50-80 varieties in our sample to be later compared to an internal classification based on shared phonological innovations. We dissipate confusion surrounding some glossonyms and point out misinterpretations that led to conflation of different varieties and misnumbering in non-genetic, referential classifications (Guthrie 1971 Maho 2009 Hammarström 2019). First, we present an updated list of nearly 100 language varieties falling into Guthrie’s B50-80 groups along with corresponding updated geographical coordinate values.
According to lexicostatistical and lexicon-based phylogenetic studies, these languages belong to the West-Coastal branch of the Bantu family along with the Kikongo Language Cluster. In this paper we deal with the identification, geography and internal phylogeny of Bantu B50-80 languages. In any event, building on Möhlig (1981), we speculate that the recurrent devoicing of *g in several ancestral Bantu languages of the rainforest could be diagnostic of a pre‑Bantu hunter‑gatherer substrate. However, at this stage, there is not enough empirical evidence and especially not sufficient systematic historical‑comparative linguistic research to further test this hypothesis.
It probably also separates WCB from North‑Western and Central‑Western Bantu, where *g did devoice, but not always before *k became zero. We show that the velar merger is a unique phonological innovation distinguishing WCB from other Bantu phylogenetic groups such as South‑Western and Eastern. Based on a comparative dataset of 66 different cognate series, we demonstrate that languages previously classified as WCB according to lexicon‑based quantitative methods share at least one common phonological innovation: the phonemic merger of the Proto‑Bantu velar stops *g and *k due to the devoicing of *g when not preceded by a nasal. In this article, we assess the genealogical validity of West‑Coastal Bantu (WCB) as a major subclade of the Bantu family by means of the Comparative Method.
0 kommentar(er)
