The expansion of Bantu-speaking peoples over large parts of sub-Saharan Africa is still a matter of debate—not only with respect to the propelling force behind it and the route(s) taken, but, also, in terms of the question whether there actually was a demographic expansion of peoples, rather than just a cultural expansion involving the spread of languages and technologies. In this paper, we provide a critical review of the extant linguistic and molecular anthropological data on Africa and discuss the insights they provide concerning the expansion itself as well as the demographic processes involved in it. Contrary to some assumptions by historians and cultural anthropologists, the genetic data speak in favor of an actual movement of peoples during the expansion of the Bantu languages over Africa, rather than a spread through language and culture shift. Furthermore, the molecular data indicate that sociocultural practices such as patrilocality and possibly even polygyny played a role in shaping the genetic diversity of Bantu-speaking peoples. These sociocultural practices might explain why, in Africa, there is a correlation between Y-chromosomal (i.e., paternal) lineages and linguistic affiliation, but not between mtDNA (maternal) lineages and language.
The Bantu expansion, a major topic in African archaeology and history, is widely assumed to correlate with the spread of farming, but archaeological data on the subsistence of these putative early Bantu speakers are very sparse. However, finds of domesticated pearl millet (Pennisetum glaucum) in southern Cameroonian archaeological sites, dated between 400 and 200 BC, open new perspectives on the history of agriculture in the Central African rain forest.
Linguistic evidence suggests that pearl millet was part of early agricultural traditions of Bantu speakers, and has to a great extent been distributed during the course of their expansion over large parts of western Bantu-speaking Africa, possibly even originally from their homeland in the Nigerian-Cameroonian borderland.
In combining archaeobotanical, palaeoenvironmental and linguistic data, we put forward the hypothesis that an agricultural system with pearl millet was brought into the rain forest during the first millennium BC, and that its spread across Central Africa coincided with the dispersal of certain Bantu language subgroups.
Click consonants are one of the hallmarks of “Khoisan” languages of southern Africa. They are also found in some Bantu languages, where they are usually assumed to have been copied from Khoisan languages. We review the southern African Bantu languages with clicks and discuss in what way they may have obtained these unusual consonants. We draw on both linguistic data and genetic results to gain insights into the sociocultural processes that may have played a role in the prehistoric contact. Our results show that the copying of clicks accompanied large-scale inmarriage of Khoisan women into Bantu-speaking communities and took place in situations where the Khoisan communities may have had relatively high prestige. In the Kavango-Zambezi transfrontier region, these events must have occurred at an early stage of the Bantu immigration, possibly because small groups of food producers entering a new territory were dependent on the autochthonous communities for local knowledge.