Arguments for the Coherence of Nilo-Saharan

Abstract

The Nilo-Saharan phylum of African languages was identified by Greenberg in 1963. Although widely accepted, recent decades have seen researchers both inside and outside the field of Nilo-Saharan studies reject its coherence in whole or in part. This paper argues that it is a coherent phylum with at least twelve recognised branches and that phonological, morphological, and lexical evidence exists to support this claim. Three topics are examined in detail: three-term number marking, ATR vowel harmony, and the use of t- and k- affixing. The paper argues that these cannot be explained away as typological features, since they incorporate distinctive phonology and segmental morphology. In addition, the evidence from two personal pronouns is tabulated, each of which appears to have two reconstructible forms. The paper concludes with a new model of the internal configuration of Nilo-Saharan.