The First Ethiopians. Malvern van Wyk Smith

      Another way of approaching Egypt’s ancient Africanness is to explore the links between major demographic movements and the linguistic foundations of prehistoric north Africa. As we have seen, genetic and skeletal investigations indicate fairly clearly who the ancient Egyptians were not, and these findings are borne out by demographic and linguistic evidence. The Late Pleistocene desertification of the Sahara between about 70 000 and 14 000 years ago, and the consequent withdrawal of its populations to the east, north and south, where they continued to develop broadly in mutual isolation had the further consequence of determining north Africa’s three major language groups (Shaw et al., 1993, 18). J.H. Greenberg’s classification of these language groups into Afro-Asiatic, Nilo-Saharan and Niger-Congo (1966, 1981), of which Nilo-Saharan is the most diverse and probably the oldest (Sutton, 1981), is now widely accepted (Trigger, 1978a; Ehret and Posnansky, 1982; Diakonoff, 1989; Oliver, 1991; Iliffe, 1995; Picton, 1995) even while still being refined (Blench, 1993; Ehret, 1993). The fourth major and oldest African language phylum is a loose group of hunter-gatherer click languages, of which various Khoisan exemplars are the most familiar; we shall come to these in due course.

      Greenberg’s classification has been substantially challenged only by some Africanists (Diagne, 1981, for example), but it is not difficult to see that their unease is inspired largely by two serious quandaries for Africanist hyperdiffusionism arising from Greenberg’s work. Firstly, his Afro-Asiatic may not be of sufficiently or necessarily autochthonous African origin to satisfy Afrocentrists. Secondly, his Nilo-Saharan, which eventually extended in what Dalby calls a ‘fragmentation belt … extending across Africa from the coast of Senegal in the west to the Ethiopian and East-African highlands in the east’ (Dalby, 1981, 311), must have been in reality as much as it is now in linguistic implication a powerful obstacle to any simplistic diffusion of a pan-African ‘Negroid’ language and culture that embraced both Egypt and west Africa. Diagne’s preference for Olderogge’s ‘great Zindj or Congo-Sahara family’ of languages is inspired by the laudable persuasion that ‘ethnocentricity has gone a long way to distort analysis of the material’, and that ‘Afro-Asiatic’ is merely the old ‘Hamito-Semitic’ made more respectable (Diagne, 1981, 246–247), but it simply turns African languages into a meaningless melée.

      To return to Greenberg’s classification as elaborated by Dalby, Sutton, Blench, Ehret and others, the collective wisdom now holds that the oldest north African language, Nilo-Saharan, originated in the south-central Sahara during the Late-Pleistocene period of extreme aridity, somewhere between 18 000 and 13 000 years ago, from whence, and noticeably from the Early Holocene wet phase of about 8000–4000 BCE onwards, its derivatives would spread across Africa, from about the Niger bend to the Nile Basin ‘in a broad belt lying between a much shrunken Sahara and an enlarged equatorial forest’ (Sutton, 1981, 481). It was associated with what Sutton calls an ‘aquatic tradition’ based on the lakes and rivers of the Holocene, but in the northern and eastern parts also associated from an early stage with the cattle pastoralism developing in the eastern Sahara among speakers of Afro-Asiatic languages (Oliver 1991, 41–2). The earliest speakers of Nilo-Saharan in the Middle Nile Valley seem to have been of a Nubian-Nilotic stock – Hiernaux and Keita’s ‘Elongated African’ type.

      Despite whatever associations they may have established with the Afro-Asiatic speakers of the northern Nile area, these Saharan pastoralists would retain their quite different Nilo-Saharan languages – ‘for most of its known history, the Nubian stretch of the Nile has been occupied by Nilo-Saharan-rather than Afroasiatic-speaking people’ (Oliver, 1999, 41). In other words, despite the ancient proximities and assumed interrelatedness of the populations of the ancient Nile Valley already noted, two quite different language families became domiciled along the river’s banks.

      It was in the more northerly Nile Valley that the second major northern African language family – Afro-Asiatic – emerged. It may in fact have originated further south, in the Sudan (Blench, 1993, figure 7.2), although much controversy still surrounds the point of origin of Afro-Asiatic. Blench argues that proto-Afro-Asiatic – whose successors include ancient Egyptian, Semitic, Cushitic, Chadic and Berber – emerged in the central Sudan around 12500–10000 BP, then spread northwards down the Nile and further eastwards to become the dominant language family of north-east Africa. Ehret, however, has placed the origins of the other major North African language phylum, Nilo-Saharan, in virtually the same area, ‘in the general region of the Blue Nile between the Ethiopian highlands and the White Nile’ (1993, 108). This proposal not only explains the Nilo-Saharan character of Nubian-Nilotic languages as opposed to the Afro-Asiatic identity of those just further north; it underscores once again the intimate linguistic and cultural complexities of the Middle Nile Valley or the ‘Nilotic continuum’ mentioned earlier.

      Controversial and even contradictory as some of these proposals may still be, what matters for my purposes is that both ur-language groups seem to have emerged in relatively close proximity in the Nile domains. Elsewhere Ehret and Posnansky argue that ‘the Afroasiatic homeland was somewhere along the Red-Sea hinterland of Africa’ and speculate on the parallel emergence here of two major north African populations that in turn generated the two major language phyla: ‘two widespread groupings of related stone-tool industries, one [speakers of Afro-Asiatic] about 18 000 years ago which eventually extended from the Horn across Africa, and the other [speakers of Nilo-Saharan] before 10 000 years ago which extended across the Sudan and the southern Sahel’ (1982, 7).

      Remarkably, these two major language families remained quite different, one difference being that Afro-Asiatic would appear to have had strong southwest Asian elements in it from a very early stage, or even – pace the spectre of reviving a ‘Hamitic’ hypothesis – a partly Levantine developmental history. According to Roland Oliver, ‘It would be a fair hypothesis that [Afro-Asiatic] emerged from a series of dialects spoken in the valleys of the Jordan and the Nile during the enforced concentration of populations beside the rivers during the long dry periods between about 18000 and 8000 BC’ (1991, 40). From these nodes, the derivatives spread into the adjacent areas of the Levant, Arabia, north Africa and the eastern Sahara during the subsequent Holocene wet phases.

      It has long been held that among the users of proto-Afro-Asiatic were the ancestors of the Berbers and the Tuaregs, who moved from the Levant into North Africa and mixed with Nilo-Saharan and Nile Valley peoples (Iliffe, 1995, 30–31; Kennedy, 1997,