The “Ethio-Somali” gene and Semitic

So if you’ve paid attention to my blog at all and read the most recent comment replies I’ve given, you’re aware of a specific paper that will be central to this post.

In part, what I had said was:

“Intermixing between peoples of the Horn of Africa and peoples from Arabia dates back to the late Paleolithic according to Hodgson, Mulligan, al-Meeri, & Raaum (2014) and while there is evidence of Sabaean traders and artisans moving to the Horn of Africa, it was only after the advent of Di’amat and it was never mass migration. Such correlations are mostly only predominate amongst Amharas and Tigrays like myself, such results surely wouldn’t be represented in speakers of Soddo, Sebat Bet, or Silt’e. Genetic arguments are weak regarding Semitic anyways because they don’t reflect linguistic or historical reality as much as people would like. The last link you posted is mainly misleading because Mota cave is in Southernmost Ethiopia… where the gene pool is more similar to that of South Sudan or Kenya and up until recently was outside of the realm of Northern Ethiopian influence let alone then genetic influence of even the most southernly Semitic-speakers. There was even a paper that used this nice, neat, and convenient 3,000 year figure and then attempted to claim it’s linked to the Kebra Nagast… which as I’ve said a thousand times is a fabricated narrative. One could even say the 3,000 year figure isn’t actually linked to language as much as it is to the trading populations from Yemen who became clients of the local Ethiopian Semitic-speaking elite in the 1st millennium BC and therefor did mix into the indigenous population. In fact, the paper I mentioned regarding late paleolithic gene flow into the Horn of Africa that ended up producing the “Ethio-Somali”ancestral marker also mentions that this specific ancestral marker is in fact also distributed in the Levant and is attributed by the paper to Semitic-speakers, but in an awkward fashion fitting the framework you seem to agree with. The difference is that this can also be explained by the “Out-of-Africa”theory of “Proto-Semitic” as well.” – my comment on “The ‘Sabaean Theory’ is no bastard”

This of course is a very small part of the paper and a very short explanation of the discussion part of the paper, which comes to a conclusion that not only is the greater majority of the “non-African” genetic ancestry in the Horn of Africa from the late Paleolithic, but that it also may play a part in understanding the introduction of Semitic languages into the Middle East, represented roughly by a ~7% distribution of the so-called “Ethio-Somali” autosomal marker in the Levant and Northern Arabia. As said at the end of the paper, in a very weak and somewhat confused tone, it attributed the existence of this marker in the Levant with such an even distribution to the origin of Semitic languages being in the Levant via a migration of Afroasiatic-speakers into the Levant… but this is not the only interpretation possible as I had noted, and I think this post will be a good way to argue it.

If you read through the entire paper and comprehend it in its entirety (I had a terribly hard time in doing so), it becomes very clear what the argument based on the data is: the majority of non-African genetic ancestry in the Horn of Africa, across all Afroasiatic-speaking groups, is from the Paleolithic. This being said, the paper also brings up the argument that because of the lacking of evidence regarding any major genetic admixture within recent millennia on the basis of the differentiation between the non-African genetic components of populations in the Horn of Africa and the Middle East, it is unlikely that any recent admixture can be attributed to relatively any population outside of a handful of recent cases that were small scale and seem to have contributed little if anything to the linguistic geography of the Horn of Africa outside of written inscriptions in languages such as Sabaic. What must be noted in the paper is it’s ending provisional linguistic hypothesis:

” A later migration of a subset of this population back to the Levant before 6 ka would account for a Levantine origin of the Semitic languages [18] and the relatively even distribution of around 7% Ethio-Somali ancestry in all sampled Levantine populations (Table S6). Later migration from Arabia into the HOA beginning around 3 ka would explain the origin of the Ethiosemitic languages at this time [18], the presence of greater Arabian and Eurasian ancestry in the Semitic speaking populations of the HOA (Table 2S6), and ROLLOFF/ALDER estimates of admixture in HOA populations between 1–5 ka (Table 1).”

But here is what is funny about this, being none of the researchers are “bona fide” historical linguists, they actually site a paper that I had in my opinion cast reasonable doubt upon in my last post that isn’t a republished version of an older one. They literally based their hypothesis on the link between the “Ethio-Somali” genetic marker and Semitic on a dubious paper! When I read this I admittedly laughed but again, they aren’t linguist and they will go with what they as geneticists are familiar with (i.e. mathematic models over comparative linguistics) so I can’t necessarily blame them, entirely. But this of course doesn’t resolve the question brought up earlier about the existence of “Ethio-Somali” in Arabia:

 This relationship between genetic and geographic distance between HOA and Arabian populations might support a hypothesis of long-term equilibrium gene flow among these populations in an isolation-by-distance model. However, if this hypothesis were true, we would expect the highest levels of pairwise gene identity to be between HOA and Arabian populations, but this is not the case. The highest levels of shared gene identity are between HOA populations and the Levantine Palestinian and the North African Mozabite population samples (Figure 5B). Thus, it is more likely that the genetic-geographic HOA-Arabia distance gradient reflects secondary admixture of Arabian migrants into HOA populations already carrying substantial non-African ancestry or already admixed HOA populations sending migrants into Arabian populations.”


So the Levantine question makes sense to them, but the Arabian one doesn’t? But of course, they’re working with the assumptions that:

  • Semitic came to the Horn of Africa via a back-migration, after a Levantine origin.
  • Semitic-speakers inherently have more Arabian ancestry than other populations.

But this is of course questionable because well, the Semitic-speakers represented are from two majority groups who do not represent all Ethiopian Semitic-speakers. I myself am a Tigray and a carrier of J1 and L03, that doesn’t mean a Soddo or Silt’e-speaker would reflect any amount close to my amount of non-African, Arabian-specific ancestry. My ancestral population had close ties with trading colonies in Arabia and a minority population of Arabian artisans and merchants unlike theirs.  So there’s reasonable doubt in these assumptions especially being other non-Semitic populations in the region are noted in the paper as having a similar amount of Arabian-specific ancestry as well.

But what am I saying here? Why am I writing another coffee high induced post about some methodology I question? Well, because my hypothesis can actually work with this genetic data. It explains the Arabian question, and the Levantine population in question can also be dealt with via my hypothesis. I just needed to say that because it has been on my mind all week but nonetheless I seriously didn’t see the implications of this paper even though I’ve been aware of its existence for over 2 years now.

So, that being said. This post is done.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s