Ilya Yabonovich and other linguists, in examining the differences between the various members of the Afro-Asiatic family have realised that all of the old etymologies for this group were inherently semitocentric. The differences between Chadic, Omotic, Cushitic and Semitic, were wider than those seen between any members of the Indo-European family and as wide as some of the differences seen within and between separate language families, for example, Indo-European and Altaic. Certainly the exclusion of Afro-Asiatic from the controversial Nostratic family has simplified matters of phonemics, not having to include the complex patterns seen in Afroasiatic languages. It is today commonly accepted that proto-Afro-Asiatic was spoken in Africa (perhaps the Northeastern Sahara, e.g.), but in the past a number of theories were expounded.
Allan Bomhard (1996) retains Afro-Asiatic within Nostratic, despite his admission that Proto-Afro-Asiatic is very different from the other members of the proposed linguistic Nostratic superfamily. As a result he suggests it was probably the first language to have split from the Nostratic linguistic superfamily. Whatever was the case, by 8,000 BCE the Natufian culture itself had begun to disperse. In Palestine, Natufian developed into the Pre-Pottery Neolithic A (PPNA) culture, first identified by Kathleen Kenyon (1906-1978) in her 1950's excavations at Jericho. Kenyon also remarked at the hiatus and seeming abandonment of PPNA sites, and was followed by a limited extent of the PPNB culture that was very different. Rectilinear dwellings in the PPNB, from 7,000 BCE, replaced the round beehive dwellings seen since Natufian times, in the PPNA period. Since the 1960s, however, it has been shown that PPNB developed in an unbroken sequence from the Natufian cultures north of Damascus, forging a link between Palestine, Mesopotamia and the Anatolian cultures of Catal Huyuk, and Halicar, with which it shares some similarity. It has been suggested that this northern part of the range was developing as Proto-Semitic. Certainly, there is evidence that the PPNB culture, spread southwards to sites in Israel, lasting until 6,000 BCE ending with the brief spread of a more arid climate through the region. Christopher Edens (2001) has reported a bladelet tradition in South Western Saudi Arabia, possibly synchronous with the Epi-Paleolithic spread of tools of the Arabian bi-facial tradition which lasted from 5,000 - 3,000 BCE characterised by beautifully pressure-flaked arrowheads and knives. They also used scrapers and awls or drills, probably for working leather and making beads. The Bifacial tradition, seems to have been the period during which a hunting and gathering way of life was progressively replaced by a lifestyle of nomadic pastoralism from the north, which has since then characterised the Peninsula. This, it has been suggested, saw the first spread of the Semitic languages throughout Arabia.
Bomhard, following John Kern’s suggestion, proposed that further spread took the Afro-Asiatic languages across the Bab al Mandeb in Yemen into Ethiopia and thence into the Horn of Africa and further south. To the north, Afro-Asiatic languages are presumed to have crossed with the Neolithic revolution into Egypt, spreading from there into North Africa, and the Sudan, and thence across the Sahara to the area of Lake Chad.
There is thus some evidence in support of this thesis. No evidence has ever been found of a pre-Semitic substratum in Palestine, indicating a long development there. However, more recent work in African archaeology has pointed to weaknesses in the Nostraticists’ argument. Firstly it appears that the Neolithic in Africa did not develop as a result of immigrants from the Middle East speaking a new Afro-Asiatic language. Rather it developed out of a deep tradition of Egyptian Epi-Paleolithic cultures undergoing a long-process of Neolithicisation, with a full Neolithic tradition emerging with the Badarian (and possibly Tasian), about 5,000 - 4,500 BCE. It is only with the Naqada II and III periods that any evidence of incursions of people from Southwest Asia can be distinguished. By then, agricultural Egyptian Neolithic cultures had a long tradition of their own. Although earlier links can be shown to have existed between Badarian and the Western Desert and even with Merimde and the Fayyum, there are no clear early links back into Palestine or Syria.
Equally, in the Horn of Africa, although Arabian influence has now been extended before the Axumite civilisation, most of the early Epi-paleolithic links seem to have happened in the other direction, from Africa into Arabia, and it is difficult tracing a cultural trajectory sufficiently early enough to have carried the Omotic and Cushitic languages into Africa.
This archaeology seems to pose problems to a theory of a Nostratic-linked Proto-Afroasiatic language in the Middle East. There is also significant linguistic evidence that suggests that this was not the area in which Proto-Afro-Asiatic languages first evolved. Afro-Asiatic linguistic diversity is far greater in Africa than it is in the Middle East. All six of the Afro-Asiatic families are found in the African continent, only one is found in the Middle East. Even in the case of the Middle Eastern Semitic language, the diversity of Semitic languages in Ethiopia, for instance, is greater than that in Arabia, Mesopotamia or the Levant. The suggestion of a Middle Eastern origin of Proto-Afro-Asiatic of Kerns and Bomhard just represents a later continuation of the dominance in Afro-Asiatic studies by the Semiticists, and the relative depth of the understanding of the archaeology of this region, by comparison to the much less well understood archaeology of Africa and the Sahara.
The spread of Afro-Asiatic languages has recently been linked to the evolution of the Y chromosomal E1b1b Haplogroup. About 26,000 years ago, the E1b1b (formerly known as E3b) genetic haplogroup arose in East Africa (or the Middle East) and spread northward into North Africa and West Asia, splitting further into another three haplogroups: haplogroup E1b1b1c (E-M123) spent the last ice age in the Near East (where Cruciani et al. (2004) propose it arose) and the Horn of Africa; E1b1b1b (E-M81) originated in the Maghreb about 5,600 years ago, and today is the most important haplogroup of the Berbers; E1b1b1a (E-M78) originated in North Africa, around Egypt and Libya, and after the end of the ice age, it expanded into the Horn of Africa, Western Asia and Europe. The dispersal of E1b1b1a is probably connected with the spread of the Afro-Asiatic languages.
African homelands of PAA languages have been suggested. Igor Diakonoff (1988) suggested that the Urheimat of Afro-Asiatic was in the Southeastern Sahara, between Tibesti and Darfur. Martin Bernal (1980) also suggests an African origin. Quoted by Bomhard, he states that:
Archaeological evidence from the Magreb, the Sudan and east Africa [makes it seem] permissible to postulate at least three branches of Afroasiatic existed by the 8th millennium BCE”. Bomhard concludes, “The implications of Bernal’s views are enormous. Although his views are highly speculative, they are by no means implausible. Should they turn out to be true, it would give substantial weight to the arguments that Afroasiatic is to be viewed as a sister language to Proto-Nostratic rather than a descendent.
Despite this caveat, Bomhard upholds the Middle Eastern origin by approvingly quoting at length from Kerns:
If we assume that the speakers of pre-Indo-European remained in the vicinity of the Caucasus to a fairly late period (say 7,500 BCE), with the Afroasiatic already extending through Palestine and into Egypt and eventually the rest of North Africa, but with its Semitic branch still in Northern Mesopotamia, high on the upper slopes of the fertile crescent, we have an explanation for the similarity in vocabulary. That this similarity existed to a late period is suggested by the shared words for field, bull, cow, sheep and goat, animals that were domesticated first in the Fertile Crescent. In addition, shared words for star, and seven suggest a common veneration for that number and perhaps a shared ideology…. If true, it suggests an association that is social as well as geographical.