Bantu history: Big advance, although with a chronological contradiction.

Settling an Old Debate In their article “Bantu expansion shows that habitat alters the route and pace of human dispersals,” Grollemund et al. (1) have accomplished more than just their stated intention: to identify the role of habitat in channeling the directions of the early Bantu farming settlement of the African equatorial rainforest. What is most important is that the authors essentially bring closure to four decades of debate, at least with respect to the geography of the early stages and routes of expansion of Bantu speakers across vast portions of the African continent. On the geography of this spread, their findings confirm the validity of a particular line of linguistic and historical argument initiated more than 40 y ago (2) and developed and elaborated upon by a succession of investigations in the intervening years (3–5). Grollemund et al. (1) rule out, as did those other studies, the idea that the ancestral speakers of the Eastern Bantu branchmight have reached eastern Africa by a circuitous route around the north side of the equatorial rainforest. Grollemund et al.’s (1) results broadly accord in their geographical respects with the findings of the extended syntheses of Bantu history by Kairn Klieman (6) and this writer (7, 8) that, beginning in the third millennium B.C.E., Bantu-speaking communities advanced deep into the rainforest belt. Their primary spread passed initially through the Sangha River regions south-southeastward out of Cameroon to the areas of the Sangha confluence with the middle Congo River, and farther south to the region of the confluence of the Kwa River with the lower Congo. From this latter region, the next major Bantu expansion proceeded eastward through the forestsavanna mosaic on the immediate southern fringes of the equatorial rainforest 1,500 km eastward to the greatWestern Rift of Africa (7). From that region the early Eastern Bantu then spread out all across eastern and southeastern Africa (8). The attention that Grollemund et al. (1) give to the role of savanna environments in facilitating the early phases of Bantu expansion is well placed. If the expanding Bantu communities especially exploited areas of savanna within the rainforest, that history accounts for what the reconstructed lexicon of the earliest Bantu agriculture requires: that the Bantu carried along with them into the rainforest two crops requiring savanna conditions, the cowpea (Vigna unguiculata) and the African groundnut (Vigna subterranean) (8). However, in interpreting habitat history, Grollemund et al. (1) do not take into account a particular environmental factor. Even in the wettest periods of the Holocene, intercalary savannas existed, especially on sandier soils,