In the Pre-Columbian settling of the Americas, there were early layers and late layers. Let’s talk first about the early layers. Some of this is from archaeology, some from genetics (including ancient genetics), a little from linguistics. The Reich lab has played a major role in the genetic work.
The population that accounts for the vast majority of Native American ancestry, which we will call Amerinds, came into existence somewhere in northern Asia. It was formed from a mix of Ancient North Eurasians and a population related to the Han Chinese – about 40% ANE and 60% proto-Chinese. Is looks as if most of the paternal ancestry was from the ANE, while almost all of the maternal ancestry was from the proto-Han. [Aryan-Transpacific ?!?] This formation story – ANE boys, East-end girls – is similar to the formation story for the Indo-Europeans.
This new population ended up in Beringia, a land that is today the shallow Bering Sea between Alaska and Siberia, then exposed because of the low sea levels of the Ice Age. It wasn’t easy to proceed further into North America, since the path was blocked by ice sheets, and apparently the Beringians hung out there for thousands of years.
But then the ice began to retreat. The Beringians managed to get past the glaciers – probably by following a path from one unglaciated patch to the next along the Canadian west coast. The idea that they waited until an inland ice-free corridor opened up seems to be mistaken: we keep finding remains that are too old to fit that scenario.
Those early Amerindians split into two close but distinguishable populations: one that occupied parts of eastern North America (Algonquians), the other now occupying central and South America. Those Amerinds spread rapidly through the two continents, and, for the most part, they’re still where they landed. There were some regional expansions, but nothing like the vast population/language expansions in the Old World – nothing like the Bantu or Indo-European or Hamitic-Semitic expansions. In many area, the Amerindian populations today are largely descended from the people that lived there 8,000 or more years ago.
This resulted in a pattern of many deeply diverged languages, rather than a few very wide-spread language families, as seen in the Old World. Reich thinks this vindicates Joseph Greenberg’s work on classifying the native languages of the Americas. Greenberg thought the many languages in the Americas split into only three families: Eskimo-Aleut, Na-Dene (spoken along the northern Pacific coast of the US, interior northern Canada, and the American Southwest ( Navajo) – a later migration – and Amerind (everyone else). This scheme fits what we find in the genes: just three migrations and three language families. Greenberg’s method, “mass comparison” works. Another point: Greenberg thought that there was a very distant relationship between Amerind and Indo-European. If correct, that suggests that Indo-European primarily originated from a language of the eastern mostly-ANE hunters, not from a population in the Caucasus or or northern Iran. Which also accords with work suggesting Kartvelian influences on PIE, rather than origins – also with PIE having a sister relationship with Uralic. Just saying.
Back to the new world. This picture was nice and simple, but there was a fly in the ointment. Isn’t there always? A Brazilian anthropologist, Walter Neves, had studied a number of old skeletons in Brazil that looked different. The most famous of these was Luzia Woman, about 11.500 years old. Neves and others thought that she ( and other similar skeletons) looked more like Australo-Melanesians than Amerindians. Reich is dismissive of Neves’ scientific credentials – ” If I don’ know it, it’s not knowledge” – but Neves was on to something important. One of Reich’s students, Pontus Skoglund, looked more closely at native American genetic data to look for traces of a different ancestral group. He found them. Parenthetically, I’ve heard that other people had seen something weird in those Amazonian genetic samples even earlier, but seem to have thought it was too weird to publish
Some populations of Brazilian Indians were genetically closer to Australasians than to other world populations – the general group that Neves and other anthropologists had said the old Brazilian skeletons resembled. The population with the greatest affinity were the Andaman islanders, short dark people that live on islands between India and Burma.
Several of the Amazonian tribes they looked at had this admixture, at a few-percent level: the Surui, Karitiana, and Xavante. It has since been found in some other groups in or near the Amazonian basin.
Some obvious attempts at an explanation don’t work. That genetic trace isn’t from Polynesians – not a good genetic match, and the admixture is old, while the Polynesian expansion into the Pacific is recent.
The pattern of the populations that don’t have this pseudo-Andamanese admixture is illuminating. You don’t see it in the eastern branch of Amerindians, You don’t see it most of the current southern branch ( i.e. central America and South America west of the Andes). You don’t see it in ancient members of the southern branch (The Clovis-complex Anzick-1 skeleton from Montana, about 12.6k years old). You don’t see it in a Beringian that was left behind in Alaska (about 11k years old).
How can you see it in Brazil if it wasn’t already there in Beringia? Or in the early expansion out of Beringia? Or in Central America?
Because these pseudo-Andamanese were there first, before the Amerindians ever got south of the glaciers. And were then seriously stomped by Amerindians, as has happened so often in prehistory.
If you go back 18,000 years or earlier, South America was a much more appetizing target for settlement than North America, which was was glaciated and cold. Taiga all the way down to the Gulf of Mexico. While the Amazon basin was reasonably warm and far less of a Green Hell than it is today – lots of savanna.
Although likely, that’s a bad explanation, because some people won’t like it. In addition, it’s too simple. People in the soft sciences routinely use Occam’s butterknife – entities should be multiplied out the wazoo.
I’ve seen an alternative scenario : as the southern branch of Amerindians was moving south, perhaps along the west coast, a single canoe or raft full of pseudo-Andamanese landed and merged peacefully. This has to happen at exactly the right time and place – at the wavefront of people moving south. Peaceful merger wasn’t any too likely either.
On the other hand, if you assume that the pseudo-Andamanese simply arrived before the Amerindians, hat could have occurred at any time over thousands of years, and instead of having to land near the front of that wave, they merely have to hit the west coast of South America – a big target. I figure that the early-arrival scenario is thousands of times more likely than the complex-population-structure model.
There were later, smaller migrations. First, perhaps 6000 years ago, another from east Siberia that led to the Paleo-Eskimos and the Na-Dene. We’ve actually traced a clear language connection between the Na-Dene languages and a Siberian tribe, the Ket. The Neo-Eskimos arrived yet more recently and rapidly replaced the Paleo-Eskimos. Both the Na-Dene and Neo-Eskimos picked up a lot of ancestry from Amerindians.
We found that Siberian and East Asian populations shared 38% of their ancestry
with a 45,000-yr-old Ust’-Ishim individual who was previously believed to have no modern-day descendants. Western
Siberians trace 57% of their ancestry to ancient North Eurasians, represented by the 24,000-yr-old Siberian Mal’t