Abstract
Written traces of the languages spoken in the Sahara date back less than three millennia, and for most of this period are limited to its Berber-speaking and Arabic-speaking northern bangs, with traces of Mande and Songhay as well as linguistic traces in the Sahara only beginning much later. Comparative historical linguistics, however, makes it possible to reconstruct little-attested or unattested languages from their better-documented descendants, using evidence of distribution and regular sound correspondences to help establish relative time depths. Reconstruction of cultural vocabulary and identification of loanword strata enable results to be compared with archaeological and written data. In the Sahara, these methods reveal a linguistic history characterized by repeated expansions that level out earlier diversity, particularly in the north, while uncovering surprising connections between distant areas, for example, between Tabelbala and Azelik, or Gao and Kanem.
