BONA SAWA

By Christopher Ehret

Linguistic evidence is the most powerful indicator of the centrality of the eastern side of Africa in the early dispersals of Homo sapiens. Recent investigations suggest that four late Middle and Upper Pleistocene episodes of radical population relocation are mirrored in the histories and relationships of the African language families. We consider these proposed episodes in reversechronological order.

Episode 4: The most recent of these episodes dates to around 20,000 years ago, at the time of maximum dry climate in Africa. In this period, it can be argued, northeastern Africa became a refugium into which populations from distant parts of the continent retreated.
According to the most recent findings, each of the four established families—Nilo-Saharan, Afroasiatic, Niger-Kordofanian, and Khoisan—divides at the deepest level into two primary branches. In each case one primary branch is spread widely across Africa, while the other primary branch is restricted to one adjacent set of regions in northeastern Africa. The Omotic primary branch of Afroasiatic is limited to southwestern Ethiopia (Bender 1974, Fleming 1974, Ehret 1995b). One Nilo-Saharan primary branch, Koman, is spoken at the edge of the Ethiopian
highlands, immediately adjacent to the Omotic lands (Ehret 2001). The Kordofanian primary branch of Niger-Kordofanian is spoken in the Nuba Mountains, only 200-300 kilometers from Koman and Omotic languages (Williamson and Blench 2000). In Khoisan, the restricted primary branch, Hadza, is found somewhat farther off, but still nearby, in East Africa (Ehret
forthcoming).

The inescapable import of these findings is that most probable origin place of each of the African language families lay in one composite African region, comprising the southern Middle Nile Basin, the adjacent western and southern parts of the Ethiopian highlands, and certain nearby areas of East Africa (Ehret 1984; Blench 1993).

Making this case still more compelling, we now have strong evidence for postulating a fifth distinct African language family, consisting today of a single remaining language, Shabo. This language is spoken by a small community of hunter-gatherers located in far southwestern Ethiopia, right in the middle of the origin areas we must postulate for the other four families (Ehret 1995a).

Those scholars who have studied the issue of time depth in the four established families agree that very long chronologies must be postulated (Fleming 1977; Ehret 2000b, 2003). Ehret has argued from proposed archaeological correlations that the minimum time depth of the Khoisan and Afroasiatic language families is 20,000 years, while the Niger-Kordofanian and Nilo-Saharan time depths may possibly be somewhat shorter, at 15,000 to 20,000 years (Ehret 2000a).
From these adjacent regions, the various families of Africa then spread out to repopulate the continent during the periods after 20,000 years ago.

Episodes 2 and 3. Two previous episodes of population relocation, much earlier in time, are implied in recent conjectures about the relationships among the African language families and between them and families outside Africa (Ruhlen 2003). According to this proposal, the Niger-Congo and Nilo-Saharan families belong to a common Afro-Pacific macro-phylum that includes several southern Asian families and the Indo-Pacific family. Afroasiatic belongs to a second macro-phylum along with Indo-European, the various northern and central Asian language families, and the Amerindian languages. This hypothesis requires two initial streams of human movement out of Africa, one following the southern margins of Asia, eventually as far as New Guinea and nearby island chains. The second stream would have passed into western Asia and from there into Europe and across central and northern Asia and eventually into the Americas.

Episode 1. The argument has recently been made (but not yet published) that the Khoisan family may form one primary division of the existing human languages, and that all other language the families belong to a second primary division. The argument relates to the question, why do click consonants exist uniquely in the Khoisan family, even though it is known that children of all genetic backgrounds, raised in Khoisan societies, easily learn and use clicks? The most economical explanation is that the original proto-Human language from which all existing languages derive divided, at the first stage of human dispersal, into two daughter languages. One of these was distantly ancestral to the Khoisan languages. This language would have maintained the original click consonants. The other was the ancestor of all of the rest of our human languages. It would have dropped the click manner of articulation, accounting for why its descendants all lack clicks today.

The genetic implications of this explanation are that human beings initially diversified into two groupings of communities within Africa. One spoke a set of languages found across southern and eastern Africa, of which the Khoisan languages are the last remaining offshoot. the other primary group would also have been purely African in origin, but it would have accounted for the settlement of the rest of Africa and, subsequently at stages 2 and 3, for the
spread of human beings into the rest of the world. (Because no original "Pygmy" languages survived, the evidence is lacking for locating them in this picture. Their no longer extant languages conceivably might have formed a third primary branch of proto-Human.)

Implications for Human Population History

Episodes 1-3 implied by the linguistic hypotheses have partial correlates in genetic studies already published (e.g., Tischkoff and Williams 2002, Underhill et al. 2001). Episode 4 has particular implications for northeastern Africa population genetics: it implies that modern Northeastern African populations should derive from as many as four or five distinct very early human lineages. In addition, however, they should evince significant cross-population gene flow dating to c. 20,000 BP (e.g., Underhill et al. 2001), a consequence of the retreat at that period of these different populations into adjacent areas in and around the southern Ethiopian highlands.