Thursday, March 14, 2019

Ancient DNA From Iberia

We assembled genome-wide data from 271 ancient Iberians, of whom 176 are from the largely unsampled period after 2000 BCE, thereby providing a high-resolution time transect of the Iberian Peninsula. We document high genetic substructure between northwestern and southeastern hunter-gatherers before the spread of farming. We reveal sporadic contacts between Iberia and North Africa by ~2500 BCE and, by ~2000 BCE, the replacement of 40% of Iberia’s ancestry and nearly 100% of its Y-chromosomes by people with Steppe ancestry. We show that, in the Iron Age, Steppe ancestry had spread not only into Indo-European–speaking regions but also into non-Indo-European–speaking ones, and we reveal that present-day Basques are best described as a typical Iron Age population without the admixture events that later affected the rest of Iberia. Additionally, we document how, beginning at least in the Roman period, the ancestry of the peninsula was transformed by gene flow from North Africa and the eastern Mediterranean.
IƱigo Olalde, et al.,"The genomic history of the Iberian Peninsula over the past 8000 years" 363 (6423) Science 1230-1234 (March 15, 2019). DOI: 10.1126/science.aav4040

New York Times summary here. One of many notable points it makes: "Before the Roman era, the Basque had DNA that was indistinguishable from that of other Iron Age Iberians. But Roman genes did not flow into Basque Country."

The Bell Beaker people are really the only plausible source of Steppe ancestry in the pre-2000 BE period in Iberia. The Bell Beaker people may have also facilitated some of the early contacts between Iberia and North Africa as there are artifacts indicating a Bell Beaker presence in coastal Northwest Africa.

How the Basque people ended up with the language they have with the genetic makeup that they have, and the nature of the migration from the Steppe that gave rise to the Iberian Bell Beaker people (and Iberia is the first place that the Bell Beaker culture is attested), is still a mystery, as is the evidence from ancient DNA (perhaps tweaked by the data in this paper that is not highlighted) that Iberian Bell Beaker DNA was fairly distinct from non-Iberian Bell Beaker DNA (with the former more similar to Neolithic Iberian DNA and the latter more similar to Corded Ware culture DNA).

Another enduring question is whether some or all of the Bell Beaker people spoke Celtic languages, or some pre-proto-Celtic language, or some extinct branch of the Indo-European language family, or some non-Indo-European language.

The immense demographic upheaval in the British Isles and Western Europe associated with the Bell Beaker people strongly suggests that there was a language shift at the time of their arrival from what was presumably a language derived, at least remotely, from the language of the Neolithic Anatolians that was presumably spoken by the First Farmers of Europe.

But, I am among a minority of observers who are not convinced that the people of the Corded Ware civilization (who almost certainly spoke an Indo-European language or group of closely related Indo-European languages) and the Bell Beaker people may not have spoken the same language, and that the Bell Beaker people may not even have spoken an Indo-European language, despite the fact that the two populations have very similar Steppe autosomal genetic makeup.

Given the significant Steppe Y-DNA proportions in Basque men, I seek basically two plausible scenarios. 

One is that Basque men descend from an early wave of Steppe men in Iberia who integrated into their local wives' society and adopted their wives' language. In this case, Basque is probably the closest remaining language to the language of the first farmers of Europe and in particular the Cardial Pottery culture first farmers of Europe (both the LBK first farmers and the CP first farmers are basically derived from Neolithic Anatolians and probably both spoke languages related to the language of that source population).

The other is that some steppe people spoke Indo-European languages, while the proto-Bell Beaker people spoke another non-Indo-European language related to Basque. I see the fairly clean (albeit imperfect) sorting between Y-DNA R1a (and more specifically the sub-haplogroup associated with late Neolithic, early Bronze Age Indo-Europeans in Northern Europe and South Asia), and Y-DNA R1b (and more specifically the sub-haplogroup associated with the Bell Beaker people), as possible evidence of a linguistic divide between the two populations. Indeed, I would be almost certain that the Bell Beaker people spoke a different language than other Indo-Europeans, and the only question in my mind is whether the Bell Beaker people spoke a lost sister branch language of Indo-European, or a proto-Celtic language, or a non-Indo-European language.

The problem with a late Neolithic/early Bronze Age source for the Celtic languages (when there was a likely language shift driven by Bell Beaker people whose geographic range is a decent fit to the historical geographic range of the Celtic language speaking peoples) is that the Celtic language family seems too have member languages too similar too each other to have that kind of time depth, and the strong association of culturally distinctive Celtic material culture with the very late Bronze Age/early Iron Age Urnfield and La-Tene cultures that didn't obviously give rise to a major population genetic upheaval. The later date suggested by the archaeology also seems like a decent fit to the time depth of the Celtic language family (with Urnfield possibly being associated with an earlier Italo-Celtic language branch).

One possibility is that the Celtic languages are the product of a late Bronze Age/early Iron Age superstrate language that was influenced in parallel by one group of related substrate languages in Celtic areas and a different set of substrate language influences in Italic language areas. 

I haven't yet read the body text or supplemental materials and may update this post when (and if) I do.


Ryan said...

If Basque is a Cardial Farmer language, then that's a pretty impressive feat of cultural assimilation. Basque people are what, 2% E1b? Possible though - it's not like England has a huge amount of Germanic R1a and I1 after all.

If Basque is a farmer language I think it's more plausible that it's from the Danubian Neolithic as Maju has suggested for a while, and that it spread into Iberia with the Bell Beakers. That would mean R1b-M269 wasn't really a steppe marker originally though.

andrew said...

"R1b-M269 wasn't really a steppe marker originally though."

Almost certainly not true.

"If Basque is a farmer language I think it's more plausible that it's from the Danubian Neolithic as Maju has suggested for a while, and that it spread into Iberia with the Bell Beakers. "

IMHO, a very unlikely scenario.

I am agnostic over the correct narrative, but I think that if Basque is EEF the narrative runs roughly like this:

1. Steppe pre-BB have relatively advanced metallurgy and learn of stunningly great tin supplies in Tartessia via Mediterranean trade routes. (FWIW, I think Tartessia is the best candidate for Atlantis too.)
2. There is a small, first wave of Steppe pre-BB men who migrate to Tartessia (which is the archaeolgical epicenter of BB culture. They integrate into a quite advanced local culture, marry local women and rise to the top based upon their skills and culture, adopting their wives' EEF languages in a 95% non-BB society. Their skills and cultural influence transforms and revitalizes local culture. (Compare the Eastern Roman Empire where the Roman invaders began to speak Greek.)
3. Chain migration from the steppe follows and the new BB synthesis spread culturally to both new migrants who are integrated into pre-Basque culture (a hybrid of EEF and steppe pre-BB), and those who are not, outside Iberia.
4. The ongoing steppe migration leads to a war-like world with lots of fortresses arising in Portugal and the less advanced non-BB influenced locals being wiped out first, while the BB lineages dominate in the hybrid culture.
5. BB derived cultures go from Portugal to France and then eventually circle back to modern Basque country.
6. At the time of Bronze Age collapse, the BB derived cultures collapse in the face of Urnfield and later IE migration, but this is hard to see genetically because the conquerers and the conquered are genetically similar and the change is mostly due to elite dominance and not mass migration, wiping out Vasconic areas linguistically in many places.

OTOH, I think that a multi-lingual steppe with a basically Northern IE and a basically Southern pre-BB is plausible since the R1a v. R1b sorting is strong enough geographically to suggest an invisible language divide keeping them from seeping into each other more with the pre-BB, non-IE folk eventually collapsing in the steppe in the face of a CWC derived invasion that makes R1b disappear from the region. If so, Basque is a distant relative of some Caucasian language.

Ryan said...

I definitely think a multilingual steppe is the most plausible (look at Mesopotamia/the Levant with the expansion of agriculture).

If you think M269 must be non-steppe though, do you consider the area around Carpathian foothills steppe or non-steppe?

That's my bet for a ground zero for M269.

andrew said...

"If you think M269 must be non-steppe though, do you consider the area around Carpathian foothills steppe or non-steppe?"

I must not have been clear. I am pretty much 100% certain that M269 has steppe origins.

Ryan said...

"I must not have been clear. I am pretty much 100% certain that M269 has steppe origins."

And where are the boundaries of the steppe in your view? Some consider Pannonia part of the Eurasian steppe.