March 27, 2014

Haplogroup E and Afroasiatic

This is an open access article.

European Journal of Human Genetics advance online publication 26 March 2014; doi: 10.1038/ejhg.2014.41

Y-chromosome E haplogroups: their distribution and implication to the origin of Afro-Asiatic languages and pastoralism

Eyoab I Gebremeskel and Muntaser E Ibrahim

Archeological and paleontological evidences point to East Africa as the likely area of early evolution of modern humans. Genetic studies also indicate that populations from the region often contain, but not exclusively, representatives of the more basal clades of mitochondrial and Y-chromosome phylogenies. Most Y-chromosome haplogroup diversity in Africa, however, is present within macrohaplogroup E that seem to have appeared 21 000–32 000 YBP somewhere between the Red Sea and Lake Chad. The combined analysis of 17 bi-allelic markers in 1214 Y chromosomes together with cultural background of 49 populations displayed in various metrics: network, multidimensional scaling, principal component analysis and neighbor-joining plots, indicate a major contribution of East African populations to the foundation of the macrohaplogroup, suggesting a diversification that predates the appearance of some cultural traits and the subsequent expansion that is more associated with the cultural and linguistic diversity witnessed today. The proto-Afro-Asiatic group carrying the E-P2 mutation may have appeared at this point in time and subsequently gave rise to the different major population groups including current speakers of the Afro-Asiatic languages and pastoralist populations.



AdygheChabadi said...

One thing...I think there is a typo in the paper.

They classified Oromo as Omotic. Oromo is not Omotic. Oromo has been long established as an East Cushitic language. Oromo is related to the East Cushitic "Konsoid" languages in a sub-grouping called "Oromoid" by many authors.

Other than that, great info to work with. Kudos to the authors.

astenb said...

@ Ponto - Why dont you accept Cushitic, Chadic and Omotic belonged to the Afro Asiatic family? Linguistic grounds?

What alternative family would they belong to?

Arte Watch said...
This comment has been removed by the author.
AdygheChabadi said...

There is some debate as to how many of the languages are related within the construct of Afroasiatic, but that the family is real and has a solid foundation is beyond any legitimate doubt.

Semitic is African in origin, it originated in what is known as Northern/ Lower Egypt. Upper/ Southern Egypt is where Egyptian has it's origin. There were to very distinct cultures that occupied those regions in Egypt. Also consider the Semitic substratal influence that dates back to a very early stage in the Egyptian language and that Egyptian and Semitic are rather close, but with some great fundamental differences also. Semitic seems to have spread into the Levant given a substrate that indicates that people live in the Levant before Semitic arrived. For example, the toponym, Damascus, is suspected as being pre-Semitic and is not IE either given that the site is older than IE itself (according to the dates associated with IE).

Berber is definitely African in origin.

Cushitic is is very securely Afroasiatic. Just as much as Semitic and Egyptian. Many reconstructions of Afroasiatic proto-forms would not be possible without Cushitic. Cushitic is a major and rather important primary branch of Afroasiatic.

Chadic is very securely Afroasiatic, no serious linguist would ever deny that. There is influence from other language families including other Afroasiatic languages.

As for where Afroasiatic originated. I have said before that there is strong evidence for Afroasiatic being from outside Africa as well as from Africa. Which seems rather weird I know, but consider the many languages that had contact with proto-Afroasiatic including Dravidian, Sumerian, Elamite, and Proto-Indo-European. All of those languages have lexical items that ARE NOT Semitic, but ARE Afroasiatic. There is also a rather detectable Cushitic substrate in the Arabian peninsula. There is also a Semitic substrate in Sumerian that is not Akkadian/ East Semitic nor can be connected to any attested Semitic language. Consider that Afroasiatic and Elamite may actually be related and not just related, but Elamite may have been an early branch of Afroasiatic, at least, according to V. Blazek (who is widely respected in this department). R. Blench also finds the connections highly interesting and greatly worthy of further investigation. G. Starostin considers that Afroasiatic, Dravidian, and Elamite as closer to each other than to other language families with Elamite being closer to Afroasiatic than to Dravidian.

Also consider that many of the oldest toponyms and hydronyms in the Northern half of Eastern Africa are Afroasiatic.

In my opinion, it is highly likely that Meroitic, the language of Kush, is Afroasiatic. There are fundamental aspects (those that are understood) of the language that most closely match Afroasiatic and are not at all what one would expect for a Northern Eastern Sudanic (Nilo-Saharan) language. Even what can be known of the lexicon (especially those stable words that are part of the understood lexicon) is rather closer to Afroasiatic forms than to any Nilo-Saharan forms, but my research is still we will see if my preliminary assessment holds up. So far, Kirsty Rowan seems to be right about the Afroasiatic-ness of Meroitic. Mr. Rilly's work on the Meroitic language has some downright massive methodological defects to say the least about it.

AdygheChabadi said...

As for Christopher Ehret, he has been rather vindicated in his assessment of Nilo-Saharan and African history. So many have criticized his work only for the critics to be left wrong after genetic evidence backed up what Ehret has said all along based on linguistic considerations. I must say, that in Ehret's Nilo-Saharan tome, he does not separate out obvious Afroasiatic adoptions into Nilo-Saharan. There are some obviously Arabic adoptions reconstructed as Proto-Nilo-Saharan forms. Maybe he should have had it reviewed by other authors so he could have had these errors pointed out and corrected, but he chose to not do that. The criticism of Ehret having a wide semantic net is evident in some forms, but this same criticism can be leveled at Bender's Nilo-Saharan treatise. Bender's treatise does separate out shared Afroasiatic lexemes found in Nilo-Saharan. Ehret's work is fairly excellent overall and very much of what Ehret reconstructed matches reconstructions in Bender's treatise, although, Ehret's work is wider and deeper in scope.

Arte Watch said...
This comment has been removed by the author.
andrew said...

"Semitic is African in origin, it originated in what is known as Northern/ Lower Egypt. Upper/ Southern Egypt is where Egyptian has it's origin."

The evidence on that point is not clear enough at all in my mind to say with that certainty, although this is a plausible hypothesis, and maybe even the most likely of several plausible hypotheses.

For example, an alternative narrative that is not ruled out by the evidence would be that proto-Semitic may have been spoken, e.g., in Jericho, that it spread to Africa via Egypt as part of the first wave of expansion of the Fertile Crescent Neolithic. That language differentiation heavily influenced by substrate influences give rise to Coptic from proto-Semitic, and that the other Afro-Asiatic languages could in turn be derivative of Coptic.

Also, of course, it is pretty much definitively established that the Ethio-Semitic family of languages is derived from a single South Semitic proto-Ethio-Semitic language that arrived from Arabia near the Gate of Tears, probably in the Bronze Age.

"Berber is definitely African in origin."

Agreed. But, population genetics suggest that it is also quite likely one of the last Afro-Asiatic languages to come into being and probably arose via language shift from one or more non-Afro-Asiatic languages spoken by the genetic ancestors of the Berbers, rather than demic expansion and differentiation from another Afro-Asiatic population.

"Cushitic is is very securely Afroasiatic."


"Chadic is very securely Afroasiatic."

Agreed. Genetic evidence as well as linguistic evidence also points to a closer relationship of Chadic to Cushitic than to other Afro-Asiatic languages. And, archaeological, genetic and linguistic evidence all point to a quite specific point in time for Chadic ethnogensis ca. 7000 kya in the vicinity of Lake Chad from a population including a substantial proportion of men who were recent West Eurasian migrants with Y-DNA R1b-V88, many of whom may have married Cushitic women.

"consider the many languages that had contact with proto-Afroasiatic including Dravidian, Sumerian, Elamite, and Proto-Indo-European. All of those languages have lexical items that ARE NOT Semitic, but ARE Afroasiatic. There is also a rather detectable Cushitic substrate in the Arabian peninsula. There is also a Semitic substrate in Sumerian that is not Akkadian/East Semitic nor can be connected to any attested Semitic language."

Interesting assertions. Sources?

"Elamite may have been an early branch of Afroasiatic, . . ."

I give Balzek the benefit of the doubt, since I don't know him, and respect Blench, but to the extent that G. Starostin (who appears to have done some legitimate and solid work in the trenches of particular languages) is relying on the pretty much crackpot quality theories of his late father Sergei Anatolyevich Starostin (who died in 2005), color me skeptical. I have never found the Elamite-Dravidian link convincing.

I can see three SW Asian Afro-Asiatic scenarios - (1) one with Semitic as the source of the others scenario, (2) another in which proto-Afro-Asiatic migrates to Africa with the Neolithic, mutates, and then Semitic is a back migration that is an offshoot of Coptic, and (3) one where proto-Afro-Asiatic migrates to Africa via the Gate of Tears or Sinai (perhaps as part of pre-Neolithic back migrations to Africa). In both (2) and (3), proto-Afro-Asiatic dies where it arises.

AdygheChabadi said...

@אחיקם גדליהו ישעיהו בן הלל

Corrections: "There were TWO very distinct cultures..."

Well, I should clarify. I believe Proto-Semitic originated in Northern Egypt and spread to the Levant where it later or further diversified. I think many Semites stayed behind and were later assimilated by the Egyptians which increased the closeness of Egyptian to Semitic by way of Egyptian absorbing/ assimilating a Semitic linguistic substrate. I think Semitic and Egyptian have been in contact by way of geographical proximity ever since their respective origins. Plus we cannot leave out the subsequent contact between the Northwest Semitic languages and Egyptian (in both directions). It would seem Semites have been living in northern Egypt since before there was an Egypt. S. O. Y. Keita mirrors your opinion about pre-Proto-Semitic. Helmut Satzinger states, “Historical Egypt is constituted of two populations: that of the Delta, and that of the Nile Valley. Most probably, these groups had different languages, and it is only one of them that is the ancestor of historical Egyptian. At present, many assume that Proto-Egyptian is the language of the Southerners (Naqâda culture; cf. Helck 1984; Helck, 1990). We know nothing at all about the other language. The Valley population is not indigenous. It has immigrated either from the south or from the south-west.

In the opinion of many authorities the other language was Proto-Semitic (whatever that entails).

I agree with you about an over-emphasis, especially, on Semitic. Many reconstruction are very Semito-centric. I think Ehret focused on Cushitic because it is one of the oldest branches of Afroasiatic and Saho-Afar (Cushitic) is one of the most conservative Afroasiatic languages…the others being Tuareg (Berber); Arabic, Ge’ez [no longer spoken], and Tigre [most conservative extant Ethio-Semitic language] (Semitic). Even since the publication of Ehret’s Afroasiatic opus, we know even more now than at that time about Afroasiatic. It has been almost 20 years since it was published. He should release a revised and updated version. He should do the same with his Nilo-Saharan opus.

Also, Ponto mentioned Ehret’s cattle hypothesis. I think Ponto, Ehret , and Dienekes would be interested in this…

Ancient African cattle first domesticated in Middle East, study reveals
Date: March 28, 2014
Source: University of Missouri-Columbia


See my comments on Semitic to Gedaliah. I will add to that, that Proto-Canaanite was spoken in Egypt possibly as early as 30th century BCE…

The alternative hypothesis you propose is not plausible. Coptic cannot be the source of other Afroasiatic languages for rather obvious reasons. Coptic is the clear descendent of Old Egyptian which is not much different from Middle Egyptian. Much of the Semitic substrate in Egyptian is of the Archaic period.

We agree about Ethio-Semitic, Berber, Chadic, and Cushitic which is good to see. I also believe that Omotic is securely Afroasiatic, despite minor contentions.

The statement about Elamite and (Proto-)Afroasiatic contacts is derived of R. Blench and Blazek. Mr. Blench mentions those connections in his book… He gives more sources than just Blazek in the references and notes.

I think Dravidian and Elamite similarities are areal and not genetic. They are the result of contact.

(Proto-)Afroasiatic is weird in that it seems to have been on both sides of the Red Sea. Perhaps it came from the Near East into Africa by both of the corridors you mentioned…who knows, hahaha. All very complicated.

Umi said...

Ponto's comment is seriously ignorant. The fact that you don't 'believe' in something does not discredit decades long work of serious linguists.

Dr. Clyde Winters said...

The Berbers are probably not native to North Africa.

The Berber languages as pointed out by numerous authors is full of vocabulary from other languages. Many Berbers may be descendants of the Vandels (Germanic) speaking people who ruled North Africa and Spain for 400 years. Commenting on this reality Diop in The African Origin of Civilization noted that: “Careful search reveals that German feminine nouns end in t and st. Should we consider that Berbers were influenced by Germans or the referse? This hypothesis could not be rejected a priori, for German tribes in the fifth century overran North Africa via Spain, and established an empire that they ruled for 400 years….Furthermore, the plural of 50 percent of Berber nouns is formed by adding en, as is the case with feminine nouns in German, while 40 percent form their plural in a, like neuter nouns in Latin.

Since we know the Vandals conquered the country from the Romans, why should we not be more inclined to seek explanations for the Berbers in the direction, both linguistically and in physical appearance: blond hair, blue eyes, etc? But no! Disregarding all these facts, historians decree that there was no Vandal influence and that it would be impossible to attribute anything in Barbary to their occupation” (p.69).

Andre Basset in La Langue Berbere, has discussed the I-E elements in the Berber languages. There is also a discussion of these elements in Schuchardt, Die romanischen Lehnworter im Berberischen (Wien,1918). Basset provides a few examples in his monograph. You can also consult Note di geografia linguistica berbera more ,by Vermondo Brugnatelli :

The influence of European languages on the Berber languages and the grammar of the Berber languages indicate that the Berbers are probably of European, especially Vandal origin.

Dr. Clyde Winters said...

@ Andrew
You said, “Genetic evidence as well as linguistic evidence also points to a closer relationship of Chadic to Cushitic than to other Afro-Asiatic languages. And, archaeological, genetic and linguistic evidence all point to a quite specific point in time for Chadic ethnogensis ca. 7000 kya in the vicinity of Lake Chad from a population including a substantial proportion of men who were recent West Eurasian migrants with Y-DNA R1b-V88, many of whom may have married Cushitic women.”
At this time the dominate culture in this part of Africa was the Ounanian culture which has no affinity to Eurasian cultures.
What is the archaeological and linguistic evidence for Eurasian migrants in Chad?
Y-DNA R1b-V88 is recognized as being of African origin, it is rare among Eurasians, how did Eurasians introduce the haplogroup to Africa?
You have considerable merit in your statement about Afro-Asiatic being contentious. As a result, they have not been able to satisfactorily reconstruct Proto-Afro-Asiatic.
In many books on Afrasian languages, the proto-terms for this language are primarially semito-centric. This feature of the Afro-Asiatic languages have encouraged people to assume that the alleged Afro-Asiatic languages originated in Arabia. This view is false, only 20 out of the 230 known semitic languages are spoken outside of East and North Africa

Both Ehret(1995) and Orel/Stolbova have reconstructed Proto-Afrsian. A comparison of the 217 linguistic sets used to demonstrate Proto-Afrasian lexica only indicate that 59 agree. Of Ehret's 1011 entries 619 are incompatible with Orel/Stolbova, while only 175 are complimentary.

Less than 6% of the cognate sets of Ehret were proposed by Orel/Stolbova and only 17% are complimentary. This illustrates the imaginary relationship that exist between the so-called Afrasian languages.

Ehret (1995) and Orel/Stolbova (1995) were attempts at comparing Proto-AfroAsiatic. The most interesting fact about these works is that they produced different results. If AfroAsiatic existed they should have arrived at similar results. The major failur of these works is that there is too much synononymy. For example, the Proto-Afro-Asiatic synonym for bird has 52 synonyms this is far too many for a single term and illustrates how the researchers just correlated a number of languages to produce a proto-form. The failure to reconstruct a reliable Proto-Afro-Asiatic language suggest that Afro-Asiatic is not a real family of languages.

Arte Watch said...
This comment has been removed by the author.
AdygheChabadi said...

@אחיקם גדליהו ישעיהו בן הלל

I agree with you about the need for more testing and there may even be some new studies forthcoming. We just have to keep our eyes open.

I am sort of doubtful of the initial correlation of YDNA J1 with Semitic. It may be associated with Semitic speaking populations in the present, but what if it was not in the beginning?

I think you hinted at that with your comment about E-P2?

The Levant is heavy in both Y-DNA J1 and J2 and the populations with the highest known concentrations are in the Caucasus, so I think there were other languages in the area before Semitic. If A. Militarev is correct then that presents problems for everything I just said.

Arte Watch said...
This comment has been removed by the author.
astenb said...

I believe these geneticists are linking the origin of Semitic or proto-Semitic to migrating Africans based on the distribution of E-M123.

One interesting thing to note is that Egypt, gets a full agricultural package from the Levant. This is arguably one of the most important Neolithic inputs into Africa during that time. Egypt though, does not receive proto-Semetic loanwords characterizing the agricultural package. I dont know if this is also the case with Sheep and goats or Cattle.

Also, Chadic speakers according to a recent study on Sahelians (Buckova,Cerny) have significant frequencies of M78, M35 and M123. The M78 that exist is also not of the v32 subclade. What this indicates is that Chadic speakers cannot always be ascribed to a V-88 founder effect, and along with other Afro-Asiatic speakers in Africa carry that important m35 lineage from the Horn/Red Sea.

AdygheChabadi said...

@אחיקם גדליהו ישעיהו בן הלל

It is all complicated. When looking at the genetic and linguistic evidence it, sometimes, SEEMS contradictory.

If Semitic is ultimately African and we know Y-DNA J1 is not African, barring that A. Militarev is not correct and C. Ehret is correct, how is Y-DNA J1 originally Semitic?

Again, linguistically, Egyptian and Semitic are quite close, but also have strong differences (which is expected), but despite the strong differences, they are still quite close (more so than each is to other Afroasiatic language families). One could posit that the closeness is due to substrate influences and areal contact between the two.

As an example, consider the situation of Old Nubian/ Nobiin Nubian and Kenuzi-Dongolawi Nubian, both are Nubian, and on the surface seem to belong to the same branch of Nubian and are rather close. Further analysis would show you that, in fact, Kenuzi-Dongolawi Nubian is not very close to Old Nubian/ Nobiin and is rather closer to the Kordofanian Nubian languages. Once one understands the histories it becomes clear that the closeness of the two Nubian languages is due to Kenuzi-Dongolawi adopting a massive portion of Nobiin phonology, morphology, and a good portion of the lexicon also. Nobiin also adopted some Kenuzi Dongolawi grammar, but vastly less than Kenuzi-Dongolawi absorbed from Nobiin. Old Nubian/ Nobiin speakers were already established and rather dominant in the Nile Valley before the Kenuzi-Dongolawi speakers showed up. Old Nubian/ Nobiin speakers have been in the Nile Valley since around the 3rd or 4th century CE. The Kenuzi-Dongolawi speakers came 1000 years later.

What if we have a similar situation with Semitic and Egyptian? If that is not the situation, then Egyptian and Semitic are genuinely close relatives, so (pre-)Proto-Semitic speakers could not have been originally J1 carriers, they would have similar genetic origins with the Egyptians. Would that not be correct?

I do understand that some J1 is pre-Islamic in Northern Africa, but how do you determine if Semitic is intrusive to Africa or native. The evidence from genes and language would seem on some levels to point to both, how do we reconcile that?

Going by genetics it seems straightforward, but introduce linguistic evidence it gets rather complicated.


You make some very valid points.

You speak of the Neolithic package Egypt received from the Near East...there are no Sumerian adoption there either, but is all for Lower Egypt not Upper Egypt. This is assuming that the same language was spoken in Lower Egypt and the Levant. From the Tasian period in Upper Egypt onward, it seems that the culture of Lower Egypt had strong influence on Upper Egypt as per Nicolas Grimal. This could be the archaic Semitic substrate that I was talking about.

That still does not determine from which direction Semitic came. The passing of such a Neolithic package would be rather quick and easy if the same Semitic language was spoken in both places (The Levant and Lower Egypt).

About Y-DNA R-V88 and E-M35, E-M78 and E-M123, Chadic evidences contact with Berber and Cushitic + Beja so it is not surprising that you would find both signatures among the speakers. R. Blench says Chadic speakers hail from the East...C. Ehret says that R-V88 is from Northern Africa. So one would have to postulate a plausible theory unifying both accepted theories.

cfpicks said...

AdygheChabadi said "If Semitic is ultimately African and we know Y-DNA J1 is not African... how is Y-DNA J1 originally Semitic?"
The same question occurred to me. What is the evidence that contradicts the explanation that J1 was intrusive to Africa, and that the similarities in Egyptian and Semetic languages is due to long close proximity? I believe that genetic evidence is accurate and growing, whereas linguistic evidence can often be noisy.

AdygheChabadi said...


Well, the thing is...language can tell us things that genes often cannot. For instance, about the 'Neolithic Package' from the Near East, genes cannot really tell us too much about that other than what genes may have been shared between the 'package' carriers and the in situ recipient population.

I think it is settled that Y-DNA J1 is not African. See here...

About the contact between Semitic and Egyptian, Well, that is a question that can only be resolved by a full investigation in to the respective languages. Are there any Egyptian adoptions in (pre-)Proto-Semitic, there must be. How much of the similarity is purely genetic inheritance from the parent language, apparently, most of it. See here...

Unknown said...

An Asian origin for Afroasiatic can be supported by archaeology even with a preagricultural background, while an African origin cannot at all. I could never understand why the Asian origin proponents make it seem like it was exclusively the result of farming expansions. 1. 15,000 ka at the Early Natufian stage a Near Eastern group migrates into Egypt, most cultures indigenous to Egypt at that time adopt this new practice without any other change in their culture. But in Nubia the Natufian merged with the Gemaian culture and became the Qadan. This seems to have resulted in assimilation only in that local area, because the similarity with the Natufian was evident, and this would have been PPrOmotic. 2. The group that stays behind becomes PrErythraic at c. 14,500 ka and still in an Early Natufian stage. c. 13,000 the Early Natufian evolves into the Late Natufian. Shortly after this the Late Natufian expands into the Negev and Sinai and becomes the Harifian. Now at 12,800 ka you have the bifurcation into North and South Afrasian with the Harifian representing the Southern branch in the steppe/desert zone and the Northern branch representing the Late Natufian that remained in the fertile region. 3. Around 10,000 ka the Harifian/Ounanian begins to migrate into Africa during the Green Sahara period where this industry become widespread in the Sahara but lasts longest in Niger. This industry is also apparent in the Western desert of Egypt from the same time. This would be the bifurcation of South Afrasian into PPrChadic and PPrCushitic. 4. Around 11,000 ka PPrOmotic people begin to migrate south from Nubia down the Nile into Ethiopia becoming the Besaka industry, with clear North African cultural affiliation such as microburin technique, arch- backed blades, La Mouillah points, and bedrock mortars- the later is found in Qadan and Natufian sites. 5. Around 8,000 ka the PPrBerber language bringing sheep and goat pastoralism which is adopted by all known lingustic phylums. Also the latter is associated with a new stone tool technology from 6,000 BC onward. 6. And finally around 7,500 ka farming is brought by the Egyptian speakers. The Haplogroup E that becomes associated with Afroasiatic is the subclade that originated in Egypt (E-M78). The subclades of this haplogroup were in the right place at the right time when pastoralism was first adopted, and then farming, nothing more than that.

Unknown said...

In the sense of the Ancient Egyptian language, there is most definitely a fusion within the language between Semitic languages and those of the original Nile.
For example the grammar of the Ancient Egyptian language (especially that of Coptic) shows affinities with Nilo-Saharan languages using a large amount of auxiliary verbs to produce tense which was absent in earlier stages of the written language.
Also Coptic shows alternation through vowels by contrasting between linking words or using them whole untouched which coincides with Nilo-Saharan vowel harmony. In my understanding, I take it that the Ancient Egyptian language was probably originally a Nilo-Saharan language which eventually diverged from the other Nilo-Saharan languages and began utilizing Afro-Asiatic nouns in lieu.
It is still quite dubious, in my opinion, if Ancient Egyptian ever did utilize vowel patterning the way it is used in Semitic languages as Coptic does not show very many verb/noun forms distinguishing differences in meaning due to the change of vowels. If this did exist in Ancient Egyptian it was either probably borrowed genetically into the language in an unstable manner or it was used and originally in pre-Egyptian history and then lost the distinction much earlier than when Coptic was spoken.

Unknown said...

I think this study provides even more evidence of what I have suspected for a while, that E-V827 originated in the Maghreb while E-V68 originated on the Egyptian Nile. Around 25ka E-M35 migrated to Egypt with some L3 and L2 lineages. Then around 22ka mtDNA Hgs U6a and M1 (possibly with Y Hg K* and P*)came from the Jordan Valley with Ouchtata bladelet industries very similar to the Masraqan industry of Ohalo II in the Near East. After a period of admixture in the Nile Delta E-M35 went to the Maghreb and begat E-V827. mtDNA Hg M1 begat M1b while here also at ~21ka and U6a begat U6a7, U6a1, and U6a6. The E-M35 that stayed behind on the Egyptian Nile became E-V68. mtDNA M1 begat M1a and U6a begat U6a2 at this time ~21ka. Because there are LGM dates for all these haplogroups the migration of E-M35 from Ethiopia cannot be the split between Pre-Proto-Omotic and the rest of Afroasiatic, i.e. Proto-Erythraic. Afroasiatic must have either originated in Egypt or the Near East, also because the language family cannot be ~22ka. It could only be between 15-12ka based on glottochronology and archaeology (grain grinding cultures appear at that time). Around 13ka people from North Africa migrate into the Near East this is associated with the practice of tooth evulsion (Maghreb) and Mechtoid traits in certain Late Natufian sites. Interestingly, Y and mtDNA characteristic of both the Maghreb and northeast Africa come to the Near East at this time. E-M123 and M1b lineages were brought with the Maghrebi cultural characteristics and E-M78 lineages along with M1a enter the Near East as well probably with the sub-Saharan physical traits that were detected at Late Natufian sites as well. The fact that M1b must have come from the Maghreb and must have come with an associated Y Hg; it only leaves E-Z830 as a candidate that came with M1b. The other candidates such as the E-M78 lineages are northeast African in origin so that rules these lineages out. Other lineages of E-V827 in the Maghreb also argue for its origin there. E-V257 has it's highest frequency in the Maghreb and E-M81 has its highest frequency there as well. When E-V827 descendant E-Z830 migrated to Egypt ~13ka it split into E-M123 and E-V1515. This is why I think the upper limit of Afroasiatic must be 13ka; the first significant appearance in the archaeological record of north African cultural traits in the Near East and this coincides with a split shortly thereafter of E-V1515 going southeast along the Red Sea coast ~12ka. This would be the migration of Pre-Proto-Omotic. The migration south of E-V1515 would correlate very well with the Besaka industry in Ethiopia ~9500 BC which also has Ouchtata characteristics and other traits that are only found north of the Second Cataract, such as microburins and La Mouillah points. It is also at the Early Holocene that the cultural trait of tooth evulsion from the Maghreb appears for the first time in Sudan and Kenya ~9ka, and seems to be associated with the movement southward of this Maghrebi Y-Hg E-V1515. The evidence shows that not long after the origin of E-M35 in Ethiopia it must have migrated north to Egypt during the LGM. Because of this it could only be the North African decendants of E-M35 that became a part of the Afroasiatic language family. Cushitic speakers of Ethiopia and Sudan actually have both the E-V1515 lineage the moved south ~12ka and other haplogroups that originated in Egypt AFTER ~12ka such as E-V32 and E-V22 sowing that Cushitic speakers moved south from Egypt after the origin of E-V32 ~8.5ka.

Knut Holt said...

To link the genetic trait to only Afrasian is surely too narrow, but if you associate it with Nostratic or Pre-nostratic, the genetics and linguistics fit much better together. Then you can postulate that a group of people with the raits came out of north-east Africa to the middle east, and from there spread both back into Africa and further northwards in Eurasia.