Homo arabica

Here is a study finetuning Arabia. Per Saraceni: The study suggests that ancestors of Arabian Peninsula groups split from early Africans some 90,000 years ago, from ancestral Europeans about 42,000 years ago, and from ancestral South Asian populations some 32,000 years ago. She points us to Charles Q. Choi.

This was all Pleistocene so I had assumed that Arabia was dry but subject to monsoons. Given the monsoons I expect Oman as the Pleistocene staging point.

Also: "ancestral Europeans" is what it says it is. That 40kBC date doesn't surprise me, because the Neanders were still king from Anatolia northwest.

That South Asian 30kBC by contrast is simplistic to the point of being preposterous. The study must be ignoring the relict populations still lurking around various jungles and islands including, er, the whole of New Guinea and Australia, and maybe even Hy-Brasil. I'll buy 30kBC for the old Dravidians and Elamites but not for the Andamans. I take it that southeast Asian F holds true as the first split with GHIJK being everyone else and the true topic of this new study.

Anyway all of this just shows GHIJK phylogeny, not its locations at the time. The Great Divide's theory of Sundaland Noah is still possible, for 50kBC.

I am also unable to fit D in here, such old Siberians as now found in Japan. I do not think they ever approached southeast Asia from 50kBC to the Viet era. If nothing else the northeasterners own true Denisovan traces where the southeast has a still-undetermined (and much heavier) paraDenisovan lineage.

Trying to fit all that together, the I and K lineages map to ancestral Europe and to ancestral North Eurasia respectively; Aryan R1a, Celtic R1b, and American Q among K's sons. Non-R1a India trends high in H although Wiki tells me J can rise to about a third in Pakistan. I had thought that IJK went together but, apparently, that should be amended to IK; instead, HJ go together, splitting 30kBC.

As for Arabia back home, at the tail of the Glacial Maximum 18-10 kBC; this peninsula was moister. Or, if you like - Oman was larger then. Still too late for those earlier population emissions, but it did contribute to an expansion within Greater Oman herself. This population following the Holocene drought broke apart into tribes including, I dare say, the ancestor to our Semites. And maybe to our Cushites as well.

As for full-on Arabian Semitica, the paper implicates the male line "J1a2b", with 29 subclusters. Quraysh is J1...c3d and J1-a2b is linked to the Yemen so, -a2b will be Qahtan. That J1 / J2 split was before 18 kBC I think.