Out of Africa

There is a new paper out in the American Journal of Human Genetics, in which the authors try to determine the path by which modern humans left Africa – basically, through the Sinai Peninsula into the Levant,  or crossing the Bab el Mandeb strait (very narrow in glacial times) into Yemen.  Life is probably somewhat more complicated than this question suggests, since there were probably at least two expansions out of Africa ( judging from those mysterious Basal Eurasians) and maybe more, but the question itself is not an unreasonable one.

They try to see which population – Egyptians or Ethiopians – is genetically closer to  Eurasians. That sounds nuts, since both populations have had massive genetic inputs from Eurasians, but they try to identify and mask the Eurasian components of the genome  and compare the residual genomes.  Nuts is maybe too strong a word here, but that strategy (assuming that you can implement it correctly) is extremely dubious, because it implicitly assumes that, prior to relatively recent Eurasian gene flow , the locals have been sitting in the same place – without being replaced (or mostly replaced) by a different population – for  sixty or seventy thousand years.  In the regions where we have good information (parts of Eurasia, from ancient DNA) this never happened. Eurasians are like ogres – multiple layers of ancestry.

They talk about Ethiopian genetics. They say that highland Ethiopians like the Amhara are about 50% Middle Eastern, Somalis 30-40% Middle Eastern, but almost no Eurasian ancestry in the Gumuz (a Nilotic people that speak a (probably) Nilo-Saharan language). So they use Gumuz samples for their  comparison..  But since the Nilo-Saharan expansion is thought to be recent (a few thousand years), and to have originated in the eastern Sudan, the odds are strongly against the Gumuz having been sitting in place in Ethiopia  for the last 70,000 years.  I don’t think the Gumuz are  a good proxy for ancient Ethiopians – they look like recent immigrants from the Sudan.  They live pretty close to the Sudanese border now, and used to be closer. There may well be people in Ethiopia that have some local hunter-gatherer ancestry (there are indications of this) but the Gumuz are probably not among them. And, even then, it’s unlikely that those hunter gatherers  are a good proxy for Horn-of-Africa types from 70,000 years ago.

But their treatment of Egypt is far worse. They find that Egyptians are about 80% Eurasian – which means Middle Eastern mostly, and sounds about right. They estimate that the time of admixture (between Middle Eastern and African ancestry) was around 750 years years ago, which also sounds about right. But from what they actually say in the paper, it sure sounds as if they think that Egypt was sub-Saharan-type African until the Arab conquest ! Someone bleached all the mummies!

In fact, Egypt was settled by Middle Eastern type farmers maybe 7000 years ago (later than the Balkans) , who brought in all the standard crops, and most of that sub-Saharan component arrived after the Arab conquest, through the slave trade.

But crazy as it sounds, they must think that the black African ancestry in Egypt is old while the Eurasian ancestry is new, else there is no point at all in what they are doing.  They have it backwards.

Now those Middle Eastern farmers settling Egypt (speaking something in the Hamito-Semitic family) must have encountered some kind of hunter-gatherers living in the Nile Valley, and there’s likely some genetic trace of that. But it’s probably not very big,  and for all I know their genetics looked Eurasian as well ( looking at the Berbers, for example).  To have any chance of figuring this out you probably need to sequence mummies, if that can be done.

The idea that Egypt of the pharaohs was genetically African – by which they mean in practice sub-Saharan Africa – is lunacy. The sub-Saharan component in the Egyptians is mostly pretty new –  from the Arab slave trade, a lot of it from Nubia. In fact, that’s what it says in one of their references on this, in the abstract even:  “a migration of individuals with Nilotic ancestry into  (my emphasis) Egypt occurred about 25 generations ago (approximately 750 ya).”

Some of that Nubian ancestry is probably a bit older –  the Pharaohs hired Nubian mercenaries back in New Kingdom days – but still, nothing like ancient.

I’ve been complaining about  researchers being too narrow.  This article is a spectacular example. But it’s not just the lead author: everyone whose name is on the paper must be assumed to be equally ignorant and incurious – as well as the reviewers, and editors, and for that matter anyone who doesn’t stay up all night laughing at them.