COMMENTARY Bantu history: Big advance, although with a chronological contradiction Christopher Ehret1 of the Great Lakes region of East by Department of History, University of California at Los Angeles, CA 93012 2500 B.P. (11), and south and southeastward as far as southern before 2300 BP (12). The prior expansion of Bantu speakers Settling an Old Debate facilitating the early phases of Bantu expansion eastward from the Congo–Kwa confluence In their article “Bantu expansion shows that is well placed. If the expanding Bantu com- to Lake Tanganyika, 1,500+ km away, had habitat alters the route and pace of human munities especially exploited areas of savanna thus to have begun not just before 2500 B.P., dispersals,” Grollemund et al. (1) have ac- within the rainforest, that history accounts but probably well before 3000 B.P. The even complished more than just their stated in- for what the reconstructed lexicon of the earlier, initial period of Bantu expansion tention: to identify the role of habitat in earliest Bantu agriculture requires: that the through the Sangha River corridor and into Bantu carried along with them into the rain- channeling the directions of the early Bantu the regions along the lower could forest two crops requiring savanna conditions, farming settlement of the African equa- only have taken place still centuries before the cowpea (Vigna unguiculata) and the Afri- torial rainforest. What is most important is that. Klieman’s linguistic dating of this ini- can groundnut (Vigna subterranean)(8). tial expansion to the third millennium B.C.E. that the authors essentially bring closure to However, in interpreting habitat history, four decades of debate, at least with respect comports well with these chronological re- Grollemund et al. (1) do not take into account quirements (6). to the geography of the early stages and routes a particular environmental factor. Even in the of expansion of Bantu speakers across vast wettest periods of the Holocene, intercalary Summing Up portions of the African continent. On the ge- savannas existed, especially on sandier soils, To sum up, Grollemund et al. (1) perform ography of this spread, their findings confirm the crucial service of resolving once and for the validity of a particular line of linguistic and Grollemund et al. per- all the geographic history of Bantu dispersal, historical argument initiated more than 40 y form the crucial service and their proposal that savanna environ- ago (2) and developed and elaborated upon by ments helped channel early Bantu expansion a succession of investigations in the inter- of resolving once and for through the Sangha and lower Congo River vening years (3–5). Grollemund et al. (1) rule all the geographic history regions seems very well taken. The “Sangha out, as did those other studies, the idea that of Bantu dispersal. Interval,” however, came into being only after the ancestral speakers of the Eastern Bantu Bantu peoples already had passed through branch might have reached eastern Africa by a in parts of the Sangha corridor and in the those regions all of the way to the edges of circuitous route around the north side of the areas around the stretches of the Congo be- eastern Africa. The existing intercalary sa- equatorial rainforest. tween the Sangha and Kwa confluences. So, vannas of the third and second millennia, Grollemund et al.’s (1) results broadly ac- even without climate change, Bantu settle- seem, therefore, the most probable enablers cord in their geographical respects with the ments could have hopscotched ahead from of the initial Bantu expansions through the findings of the extended syntheses of Bantu one patch of savanna to another. western equatorial rainforest. history by Kairn Klieman (6) and this writer A Contradicted Chronology (7, 8) that, beginning in the third millennium Not accounting for this factor leads Grollemund B.C.E., Bantu-speaking communities advanced “ 1 Grollemund R, et al. (2015) Bantu expansion shows that habitat et al. (1) to propose ascribing the initial north- alters the route and pace of human dispersals. Proc Natl Acad Sci deep into the rainforest belt. Their primary south migration of Bantu speech communities USA 112(43):13296–13301. spread passed initially through the Sangha ” “ 2 Ehret C (1972) Bantu history and origins: Critique and across the Equator to the period of the Sangha interpretation. Transafr J Hist 2(1):1–9. River regions south-southeastward out of River Interval” of around 2500 B.P. (1). Un- 3 Heine B (1973) Zur genetischen Gliederung der Bantu-Sprachen. to the areas of the Sangha conflu- fortunately, that idea is not chronologically Afr Uebersee 56(3):164–185. ence with the middle Congo River, and farther 4 Heine B, Hoff H, Vossen R (1977) Neuere Ergebnisse zur supported in the archaeology. Territorialgeschichte der Bantu. Zur Sprachgeschichte und south to the region of the confluence of the Here’s why. Yes, very little archaeology as Ethnohistorie in Afrika, eds Möhlig WJG, Rottland F, Heine B (Dietrich Kwa River with the lower Congo. From this yet exists for the Sangha River region itself (1). Reiner, Berlin). 5 Ehret C (2001) Bantu expansions: Re-envisioning a latter region, the next major Bantu expan- However, large bodies of evidence do exist central problem of early African history. Int J Afr Hist Stud sion proceeded eastward through the forest- for other regions of early Bantu settlement. 34(1):5–41. savanna mosaic on the immediate southern 6 Klieman K (2003) The Pygmies Were Our Compass: Bantu and Archaeologists have known for three decades Batwa in the History of West , Early Times to c. 1900 fringes of the equatorial rainforest 1,500 km thatalreadybythe10thcenturyB.C.E.—500 y C.E (Heinemann, Portsmouth, NH). eastward to the great Western Rift of Africa (7). before 2500 B.P.—the forefront of Bantu ex- From that region the early Eastern Bantu then pansion had reached the eastern side of Lake Author contributions: C.E. wrote the paper. spread out all across eastern and southeastern Tanganyika (9–11), more than 1,500 km The author declares no conflict of interest. Africa (8). beyond the Sangha and Kwa confluences See companion article on page 13296 in issue 43 of The attention that Grollemund et al. (1) with the Congo River. Offshoots of this volume 112. give to the role of savanna environments in settlement then spread farther east, into more 1Email: [email protected].

13428–13429 | PNAS | November 3, 2015 | vol. 112 | no. 44 www.pnas.org/cgi/doi/10.1073/pnas.1517381112 Downloaded by guest on September 29, 2021 7 Ehret C (1999) Subclassifying Bantu: The evidence of stem 9 Van Grunderbeek MC (1983) Le Premier Age du fer au et 11 Schoenbrun DL (1998) A Green Place, A Good Place: morpheme innovation. Bantu Historical Linguistics: Theoretical au (Institut National de Recherche Scienetifique, Brussels and Agrarian Change, Gender, and Social Identity in the Great COMMENTARY and Empirical Perspectives, eds Hombert JM, Hyman LM (CSLI Butare, Publication No. 23). Lakes Region to the 15th Century (Heinemann, Publications, Stanford, CA), pp 43–147. 10 Van Grunderbeek MC (1992) Essai de délimitation chronologique Portsmouth, NH). 8 Ehret C (1998) An African Classical Age: Eastern and in dl’Age du Fer Ancien au Burundi, au Rwanda et dans la région des 12 Katanekwa NM, et al. (1981) Radio carbon dates for Zambia. World History, 1000 B.C. to A.D. 400 (Univ Press of Virginia, Charlottesville, VA). Grands Lacs. 28:53–80. Archaeologia Zambiana 20:23–25.

Ehret PNAS | November 3, 2015 | vol. 112 | no. 44 | 13429 Downloaded by guest on September 29, 2021