https://www.nature.com/articles/s41586-021-04108-8

... So they're trying to pinpoint the spread of the (disputed) Transeurasian (Altaic) language family. This has to be much older because the extant languages are rather unrelated. At a glance, there's honestly no discernable relationship between these families, other than maybe shared contact with & borrowing from the same sort of language family at different times.

"The origin and early dispersal of speakers of....Japanese, Korean, Tungusic* Mongolic and Turkic—is among the most disputed issues of Eurasian population history" Indeed.
* Tungusic incidentally refers to a family of languages/peoples in Manchuria which are historically related to the Mongols to the North. (The Manchu now have almost been assimilated entirely with the Han Chinese.) This family clusters closely with Mongolic and Turkic.

The expansion of the Tungusic peoples into Siberia displaced the previous, much older and more unrelated-looking Paleosiberian languages. One remaining Paleosiberian language, Ket has no more than 200 speakers left. Scholars think Paleosiberian may be related to the Eskimo-Aleut languages, and some even think they see evidence for a link between these and Na-Dene languages in North America.

"Recent assessments show that even if many common properties between these languages are indeed due to borrowing, there is nonetheless a core of reliable evidence for the classification of Transeurasian as a valid genealogical group" So, the computers say there is some kind of structural connection. OK. Now what do they say about the dating of these links? That's why I read the article, the blurbs didn't give me the dates and places.

----

1) Proto-TE: ".. the common ancestry and primary dispersals of Transeurasian languages can be traced back to the first farmers moving across Northeast Asia from the Early Neolithic onwards... this shared heritage has been masked by extensive cultural interaction since the Bronze Age. " First farmers could be far back as 12,000 BC.

They actually estimate using computer analysis: " Our results indicate a time-depth of 9181 BP (5595–12793 95% highest probability density (95% HPD)) for the Proto-Transeurasian root of the family; "
...So that's 7181 BC. That's pretty close.

This paper says that agriculture actually didn't provide bulk calories for most people for 4 millennia after the first traces. https://www.jstor.org/stable/10.1086/659965?seq=1#metadata_info_tab_contents Some have suggested that millet spread as livestock feed first before becoming a staple grain.

Next is: " 6811 BP (4404–10166 95% HPD) for Proto-Altaic, the unity of Turkic, Mongolic and Tungusic languages." This is getting into the nomadic era, by the way. These languages could have been broken up by migration. (Nomadism occurs after agriculture. Counterintuitive as hunter-gatherers follow herds, too, but migrant lifestyles & animal breeding are great for trade with farmers. Think camel caravans and so forth.)

Then: "5458 BP (3335–8024 95% HPD) for Japano-Koreanic." So the split would be about as old as any major Indo-European language branch (Greek, Germanic, Italic, Slavic, Indic.), without lots of other language branches to help fill in the linguistic gaps?

Where was this purported proto-TE group? "We find support for a Transeurasian origin in the West Liao River region in the Early Neolithic." This is north of the Bohai sea (western arm of the Yellow Sea). THat would make spread and constant contact with North and NE in Siberia likely, as well as north and east into the steppes (for Turkic languages) feasible, as well as intervening groups coming from the SW (China) disrupting cultural and genetic links between Turkic and the other languages,
The earliest traces of human settlement in Shengyang area (near West Liao) date back to 8000 BP which puts it a little later than the language split. The oldest wood sculpture, a bird, was found in this area. So the Xinle culture there might be identifable with early Tungusic or Koreanic languages, but not proto-TE.

So what's the proof?

"... we examined agropastoral words that were revealed in the reconstructed vocabulary of the proto-languages.. Common ancestral languages that separated in the Neolithic... reflect a small core of inherited words that relate to cultivation (‘field’, ‘sow’, ‘plant’, ‘grow’, ‘cultivate’, ‘spade’); millets but not rice or other crops (‘millet seed’, ‘millet gruel’, ‘barnyard millet’); food production and preservation (‘ferment’, ‘grind’, ‘crush to pulp’, ‘brew’); wild foods suggestive of sedentism (‘walnut’, ‘acorn’, ‘chestnut’); textile production (‘sew’, ‘weave cloth’, ‘weave with a loom’, ‘spin’, ‘cut cloth’, ‘ramie’, ‘hemp’); and pigs and dogs as the only domesticated animals.

Some of the authors have published previous papers on this, and think that millet agriculture spread to the Russian Far East by 5000 BC (7-8000 years BP), linked to spread of proto-Tungusic languages, and to Korea by 35000 BC.

"By contrast, individual subfamilies that separated in the Bronze Age, such as Turkic, Mongolic, Tungusic, Koreanic and Japonic, inserted new subsistence terms that relate to the cultivation of rice, wheat and barley; dairying; domesticated animals such as cattle, sheep and horses; farming or kitchen tools; and textiles such as silk.... linguistic interaction between Bronze Age populations speaking various...languages."

Okay that looks good and completely logical. Hmm. Now the split in the West Liao:

" We find a cluster of Neolithic cultures in the West Liao basin, from which two branches associated with millet farming separate: a Korean Chulmun branch and a branch of Neolithic cultures covering the Amur, Primorye and Liaodong. This confirms previous findings..."

And the Amur would continue to have significant exchange and flow to Korea. They use other evidence that millet farmers are more apt to expand (slash and burn) and disperse outwards than rice farmers are, as paddies are time-intensive and labor-intensive-- more labor can yield more from paddy fields. (I wonder how climate and multiple growing seasons play into this too.)

" Neolithic population densities increased across Northeast Asia before a population crash in the Late Neolithic." Hmmm. I tried to look up the references for this and I'm not seeing the data for that. West Asia (PIE peoples) around 6000 BC, yes.

---- Why did this crash affect millet farmers nearer the coast more than denser-packed rice farmers? Famine from crop disease? Millet requires little human intervention. Crop pests? (locusts, birds?) Tsunami? Accelerated sea level rise? Hmm. Childhood illnesses requiring higher birth rates?

From another source: "Accelerated rises in sea level rise, called meltwater pulses, occurred three times.... The last one, Meltwater Pulse 1C, which peaked c. 6000 BC, produced a rise of 6.5 metres in only 140 years. It is believed that the cause was a major ice sheet collapse in Antarctica." That wouldn't have done the weather any good, but it doesn't seem to make a discernable difference in sea levels in that area. The Bering Land bridge disappeared by 8000 BC. The population back in 6000 BC is estimated to have been just 10 million worldwide, but that's up from previous estimates.

The link of Jomon to the Amur river population (North of Korean Peninsula) is interesting. They also find that Jomon ancestry was almost entirely replaced over time in Korea, with only small pockets left (not due to Japanese migration.) The research also suggests massive migration from Korea into Japan during the Bronze age.

In southern Japan, neolithic Miyako Island had strong Jomon presence (2300-1500 BC) and later migration of a non-pottery culture from Taiwan with Rykukyan languages (along with later agriculture.) around 800 BC onwards through 1100 AD.

Because this happens during agricultural expansion and climate change, being able to connect what people knew and used (re language, archaeology, and genetic flow) and what happened to them and how they adapted or not, may be helpful to us.

We have over 800 times the people on the planet now than they did back then and very different technology, of course.