A K = 4 craniometric admixture plot is at the bottom of this post.
One of the four components (purple) is an archaic component, and it continues to be modal on average for Australians, Melanesians, Tasmanians, Teita, and Zulu.
A second component (red) is a Mongoloid component.
A third component (blue) corresponds to the racial type that I have heretofore always referred to as Ainoid. This component is indeed modal on average for the Ainu, but this analysis makes it clear that Polynesians are much purer representatives of the racial type, and I will therefore refer to it henceforth as Polynesoid.
The fourth component (green) is a non-Mongoloid non-Polynesoid modern component, or, for concision, an unresolved modern component.
It’s clear in light of the K = 4 analysis that the component I called Mongoloid in the K = 3 analysis was in reality a Mongoloid + Polynesoid component. Many of the individuals that showed up as almost entirely green in the K = 3 analysis now appear as either almost entirely blue or part blue and part red in the K = 4 analysis.
Some of what showed up as archaic in the K = 3 analysis is now part of the Polynesoid component. All of the populations for which the Polynesoid component is modal have a lot less of the archaic component in the K = 4 analysis than they did in the K = 3 analysis.
In a comment on my End to simplicity post I said that the fact that the f3 statistics indicate that Ainoids (Polynesoids) and Mongoloids interbred with different races of archaics meant that Mongoloids didn’t evolve from Ainoids (Polynesoids), as I had initially believed. But since making that comment I have returned to my initial belief. The conclusive piece of evidence that Mongoloids did evolve from Polynesoids is dental morphology. Sinodonty and Sundadonty distinguish Mongoloids and Polynesoids, respectively, from Negroids and Caucasoids. But the crucial observation is that Sinodont teeth have all of the same distinctive traits as Sundadont teeth, plus additional ones. Both Sundadonts and Sinodonts have winged upper first incisors, but Sinodont upper first incisors are also shoveled, and Sinodonts have one-rooted upper first premolars and three-rooted lower first molars, instead of the two roots for both pairs of teeth in non-Sinodonts.
So if Mongoloids and Polynesoids did interbreed with different races of archaics, the interbreeding in both cases must have taken place after Polynesoids evolved from Veddoids and the Polynesoids that would become Mongoloids split off from the rest of the Polynesoids.
In the past I have always believed that since Peking Man had shoveled incisors, Mongoloids must have introgressed this trait from East or North Asian archaics. But after reading this page, I now view that belief as stemming from yet another of Weidenreich’s blunders, which I had failed to critically examine. I now believe that the single nucleotide mutation in the EDAR gene that is responsible for Sinodonty, stiff hair, small breasts, and higher sweat gland density arose in incipiently Mongoloid modern humans in North Asia 40–25 ka, and was then driven to high frequency by sexual selection.
In this comment on one of Dienekes’ posts, I said that his f3 statistics showing that Japanese skulls were mixtures of Eskimo skulls and Peruvian skulls were the result of Eskimo skulls being like the skulls of the Mongoloid Yayoi, and the Peruvian skulls being like the skulls of the Ainoid (Polynesoid) Jomon. But this analysis shows that I got it backwards: it is the Eskimos whose modal component is Polynesoid, and the Peruvians whose modal component is Mongoloid.
That erroneous statement was one of my reasons for thinking that mt Hg B was brought to the Americas by Polynesoids and not Mongoloids. In his comments for that post, Gregory76 argued strongly that B was associated with Mongoloids in the Americas. I can now see that he was absolutely right.
Eskimos appear to be representatives of a more primitive stage of Mongoloid evolution, in which the EDAR mutation had occurred and the epicanthic fold had evolved, but Mongoloid cranial morphology, including brachycephaly, had not yet evolved.
The Buriats have a large amount of the Polynesoid component, but it’s not clear if this represents admixture, or vestigial Polynesoid features.
The Chinese specimens from South China showed little of the Mongoloid + Polynesoid component in the K = 3 analysis, but the Mongoloid component is on average modal for them in the K = 4 analysis. This shows that relationships between components at different K levels may not always be straightforward.
The error of misidentifying brachycephalic Caucasoids as Mongoloid is glaring in this analysis. The Berg specimens show large amounts of the Mongoloid component, for the Lapps the Mongoloid component is modal, and Afalou-bou-Rhummel 9 shows up as being almost entirely Mongoloid, which is of course absurd.
The Polynesoid component is found at high levels in the Norse specimens, at moderate levels in the Zalavar and Egypt specimens, and at lower levels in the Berg specimens. Also, for these four Caucasoid populations, the archaic component and the Polynesoid component are negatively correlated. I believe that the explanation for these observations is that Polynesoid cranial features are in part actually Neanderthal cranial features. Neanderthal features in the Caucasoid populations will therefore show up either as the archaic component or the Polynesoid component, but the more those features get assigned to one of the two components in a particular individual, the less they get assigned to the other of the two components in that individual.
After the archaic component, the Polynesoid component is the next most frequent for the Australians and Melanesians (Tolai). For the Australians, this is entirely due to the genetic contribution of the Murrayians of Birdsell’s trihybrid theory, who were Polynesoids who expanded into Sahul 40–35 ka. Howells’ Australian specimens came from Lake Alexandrina, at the mouth of the Murray River, where the Polynesoid Murrayian element, and not the Veddoid Carpentarian element, would be expected to predominate. For the Tolai (who speak an Austronesian language), most of the Polynesoid component is probably also due to the Murrayians, but the Tolai have an additional contribution of the Polynesoid component from the Austronesian expansion through Melanesia 4–3 ka. The Australians and Melanesians also have some of the unresolved modern component. In the case of the Melanesians, this is entirely due to the Negritoid Barrineans, who were the first humans to come to Sahul, 50–48 ka. In the case of the Australians, some of it is Negritoid, and some of it may be from the Veddoid Carpentarians.
The Tasmanians have the same components as the Australians and Melanesians, but in different proportions. They have much less of the Polynesoid component and much more of the Negritoid component, which is perfectly consistent with their overall phenotype, and also with Birdsell’s account of the peopling of Sahul, in which, according to one retelling of it, the Murrayians “drove the Negritos before them until the latter retreated to the highlands of New Guinea, the rainforests of North Queensland and to then ice-capped Tasmania”.
Upper Cave 101 is about half archaic and half Polynesoid. It has none of the Mongoloid component. This is perfectly consistent with Birdsell’s classification of Upper Cave 101 as purely Ainoid.
The Arikara have a lot of the Polynesoid component. Some of this may be due to Neanderthal admixture that didn’t end up being identified as archaic, but I think that a large majority of it is from Polynesoids that made up part of the founding population of at least some parts of the Americas.
The Zulu show a significant amount of the Polynesoid component, and the Teita show an even larger amount of it. The KwaZulu-Natal Province in which the Zulu are primarily found lies on the northeast coast of South Africa. The Teita are today found in the Taita-Taveta County of Kenya’s Coast Province. They moved there from further south in Tanzania between 1000 and 1300 AD. Madagascar was first populated by Austronesian-speaking Polynesoids from southeast Borneo between 200 BC and 500 AD, and Bantu-speaking Negroids later arrived and formed hybrids with them. But this analysis shows that the Africa-Madagascar gene flow didn’t just go one way. It shows that after crossing the entire Indian Ocean to get to Madagascar, the Polynesoids didn’t just stop there; they then made the relatively short hop to the southeast coast of Africa, and interbred with the Bantu-speaking Negroids there. The Polynesoids must have also had Mongoloid admixture, because the epicanthic fold is common among the Malagasy. Features typical of Polynesoids and Mongoloids, including the epicanthic fold, have been observed in populations from South Africa to Kenya. In the past these features have always been attributed to Capoid admixture, but this analysis shows that they must also be due in part to actual Polynesoid and Mongoloid admixture.
As an example, take a look at the photograph below of Uhuru Kenyatta, who will become President of Kenya on Tuesday. The Kenyatta family owns 74,000 acres in Taita-Taveta County. He looks like a Negroid-Polynesoid-Mongoloid hybrid, almost like Tiger Woods.
The Peruvians are strikingly different from the Arikara. Whereas the non-Mongoloid admixture in the Arikara is almost entirely Polynesoid or archaic, the Peruvians have very little of either the Polynesoid component or the archaic component. Almost all of their non-Mongoloid admixture is the unresolved modern component.
So what is the source of their unresolved modern component?
I think the answer to that question is that Thor Heyerdahl was right, and that the Peruvians’ unresolved modern component is from Caucasoids.
Evidence from many different sources now supports that conclusion.
In every photograph I have ever seen of Peruvian natives, they look like Mongoloid-Caucasoid hybrids. Mestizos, in other words. But the Peruvians were mestizos before Pizarro showed up in 1526.
The Paracas mummies are Caucasoid. Their hair is not stiff, black Mongoloid hair. It’s curly, red, chestnut, and blond Caucasoid hair.
This is genetic evidence for Heyerdahl’s hypotheses. The abstract leaves their data open to interpretation, but in the actual poster they presented at ASHG 2012, they were sure to furnish a nice, politically correct explanation, and to deny the possibility of any other. In my next post I will show why their explanation is bunk.