Substack cometh, and lo it is good. (Pricing)

Open Thread – 10/01/2021 – Gene Expression

Over at my Substack I’ve posted my interview with Steven Pinker on Rationality. You can read my review at UnHerd.

I would appreciate it if readers of this weblog leave positive reviews on my podcast for Apple or Stitcher.

Also, I posted a long piece on the arrival of humans to the New World, and the finds that suggest humans were here before the LGM.

48 thoughts on “Open Thread – 10/01/2021 – Gene Expression

  1. “上等人,本事大,脾气小。中等人,本事大,脾气大。下等人,本事小,脾气大。” – 杜月笙

    “The upper-class people, the ability is big, the temper is small. The middle-class people, the ability is big, the temper is big. The inferior people, the ability is small, the temper is big.”-Du Yuesheng

    Have been living in ghetto, middle class, wealthy neighborhood, my observation is quite consistent with Mr. Du’s. Little bit different though, middle class people actually have bit better temperament than low class already.

    Mr.Du was famous shanghai mafia boss in 30s.

    “穷山恶水出刁民“ 乾隆帝。

    This is another one from Chinese Emperor Qianlong, which is talking about the same observation. He said nasty people were produced in poverty.

    No wonder Western Ancient wisers are unemotional bunch with stoicism (apathy). Indeed, emotion is the enemy of rationality.

  2. @Razib

    Have you seen this LA Times article by seasoned journalist Michael Hiltzik?

    He criticizes the press and Sanjay Gupta for giving credence to the Wuhan Inst. of Virology lab leak theory. He also criticizes Alina Chan , stating she is not a virologist and has a book coming out which CNN failed to disclose.

    “On Sept. 19, CNN aired an hourlong documentary entitled “The Origins of COVID-19: Searching for the Source.” Hosted by the channel’s star science anchor, Sanjay Gupta, the program carries the veneer of an evenhanded approach.

    Proponents of the zoonotic origin theory are given airtime, including Kristian Andersen of the Scripps Research Institute in La Jolla and Peter Daszak, a prominent grant maker in the virology field.

    But so are proponents of the lab-leak theory. They include Alina Chan, a researcher at the Broad Institute, a biomedical research center, and Josh Rogin, a Washington Post columnist. Neither has any experience in virology. Chan is co-writing a book about COVID’s origins that is expected to feature the lab-leak theory prominently, a fact not mentioned by CNN.”

    https://finance.yahoo.com/news/column-evidence-emerges-against-covid-130058105.html

    FWIW : Since Feb 2020 I considered the lab leak theory as a valid hypothesis, though I am not a virologist. I have disdain for the press smearing those suggesting said hypothesis as pushing Trump’s agenda or encouraging anti Chinese racism. The infamous Lancet letter from Daszak was the cherry on the top.

  3. @Razib: Re:DanEngber

    Evidence: https://www.americanbar.org/groups/public_education/resources/law_related_education_network/how_courts_work/evidence/

    The heart of the case is the presentation of evidence. There are two types of evidence — direct and circumstantial.

    Direct evidence usually is that which speaks for itself: eyewitness accounts, a confession, or a weapon.

    Circumstantial evidence usually is that which suggests a fact by implication or inference: the appearance of the scene of a crime, testimony that suggests a connection or link with a crime, physical evidence that suggests criminal activity.

    Both kinds of evidence are a part of most trials, with circumstantial evidence probably being used more often than direct. Either kind of evidence can be offered in oral testimony of witnesses or physical exhibits, including fingerprints, test results, and documents. Neither kind of evidence is more valuable than the other.

  4. You can be convicted of a felony and sent to jail for a very long time based solely on circumstantial evidence.

    There is no direct evidence available to those of us outside of China as to the origin of SARS-CoV-2. The lab notes are under lock and key if they haven’t been burned. The Bat Lady and the other higher ups probably operate out of complete regime loyalty, if not, they have been suitably dealt with. The underlings who did not die from the initial wave of infection are spending their time in Xinjiang province making sneakers for Nike and studying Mao Tsetung thought.

    Of course the Chinese regime is in a tough spot. If the lab leak theory is replaced by the zoonotic theory, China looks like a country of barbarians that cannot run a sanitary food supply.

    Therefore the Chinese will continue to do what they have done to date. Foreclose objective inquiry and emit obfuscating dezinformatsiya.

    The real resistance to the lab leak theory is coming from people in Washington who fear that if Americans accept the lab leak theory they will turn not just on China but on the Federal Bureaucracy — our “Deep State” — and that they will make Saint Anthony Fauci, M.D. the poster boy for out of control Federal Bureaucrats who spend their days on a frolic of their own. The fear must also be shared by the entire community of people who feed off NIH.

  5. Interesting post on the New World migrations, although I can only imagine the furor there will be if they find a mass grave of people genetically identified to the earlier population with injuries obviously from Clovis people weapons.

  6. @DaThang, OK, I’ve graphed the data from the Marciniak 2021 paper (https://www.biorxiv.org/content/10.1101/2021.03.31.437881v1.supplementary-material) but just to be very clear this is not data I’m generating myself.

    Here’s the figures for those sets: https://imgur.com/a/iGeit8P

    Decline of height by the early Upper Paleolithic.

    The only step I’ve added to treat data is normalized the heights to account for male female difference in mean and Standard Deviation so they can all easily be on one plot, and changed some group labels.

    (Pastebin for data: https://pastebin.com/M5yAW3k7)

  7. Another presentation by David Reich, from July – https://www.youtube.com/watch?v=rXsNKNZtdM0

    He doesn’t really give much new (from the previous one I mentioned in comments before here in Feb – https://www.gnxp.com/WordPress/2021/02/28/open-thread-02-28-2021-gene-expression/) but he does preview at around 16:30 why their lab thinks there’s a sudden pulse of population change around 1000 BCE. (Specific slides: https://imgur.com/a/dTTxeQR). I think it’s a tricky argument though as does rely on *very* close curve fitting with their sample set… (Another talk Reich gave in June only nailed it down to a rise between “3400 and 3000” BP, a much wider interval).

    Some small changes in proportions – I still do wonder if more steppe-rich people than they think entered slightly earlier and Neolithic ancestry did survive a little better.

  8. @Matt
    So there isn’t a strong link between the polygenic score and measured height? Could low coverage be a cause or are the predictions not that good yet?

  9. Been waiting for the open thread post…

    So folks, who has been keeping up on news of the supply chain crunch? From talking to a few people I know who work as truckers and longshoremen, I’ve begun stockpiling certain essentials.

    Honestly, on the subject, I’m starting to feel like it’s January-February 2020 again, there’s this ominous feeling of something looming that not enough people are paying attention to.

  10. @DaThang, yeah, I’m not sure how much that’s an issue with coverage vs the low prediction of current polygenic scores.

  11. @Matt. The only other possibility is that WHGs were legitimately stunted but I have not read about unhealthy bones among them so this much can be discarded I think. Have you read anything to the contrary?

  12. Chinese scientists successfully produced starch in the lab, which was just published in Science.

    https://www.science.org/doi/10.1126/science.abh4049

    Compared to corn fields, this lab starch production is 5 times faster in speed for the same space unit, 4 times more efficient in solar energy utilization, and 8 times faster in carbon fixation. 

    This provided the potential for food production in factories the same way as cement production. Traditional farming might be replaced,  as horses are replaced by automobiles.

  13. @IC: No thanks. I prefer mine in the form of potatoes, baked, with butter and sour cream, and a juicy rib eye.

  14. @DaThang, I’ve never heard that they were stunted from malnutrition more than EEF or anything. Possibly their high homozygosity might reduce their height?

    Although since its mostly short runs, I don’t think it would be vast – https://www.research.ed.ac.uk/en/publications/directional-dominance-on-stature-and-cognition-in-diverse-human-p -“In each case, increased homozygosity was associated with decreased trait value, equivalent to the offspring of first cousins being 1.2 cm shorter and having 10 months’ less education.” If first cousin offspring around 1.2cm depressed then I don’t think WHG had as much homozygosity as that?

  15. “So there isn’t a strong link between the polygenic score and measured height? Could low coverage be a cause or are the predictions not that good yet?”

    Diet is huge. Consider Korea and Japan which have had almost no significant change in population genetic mix in the last century, but in which you can very accurately assign someone to the correct last three or four generations based upon their adult height, just by looking at them, because it has increased so steadily in that time frame due to better public health and better diet.

    Also, the time frame in which diet and health are important is pretty short. For better or worse, you are looking at a 15-20 year interval after which environmental effects don’t really matter (shorter for women, longer for men). My inlaws were food insecure refugees during the Korean War and are very short as a result. Their children are also short (suggesting an epigenetic factor), but their grandchildren and grandnieces and grandnephews are tall.

  16. @Matt
    So its most likely something like low coverage.

    @ohwilleke
    Korean diet vs whg diet is a big difference. by stunting, I meant some kind of signs in whg bones. Korea’s height is probably still increasing a bit while Japan has levelled off. East Asian height peaks in Northeast China I think.

  17. Is Beijing on Borrowed Time? Is the U.S.?
    By Richard Fernandez | Oct 04, 2021 | https://pjmedia.com/richardfernandez/2021/10/04/will-china-invade-taiwan-n1521434

    Will China Invade Taiwan? Despite the recent saber-rattling, probably not any time soon. It lacks the amphibious capacity to land the 30-plus brigades it will need to overcome the island’s defense. “China does not appear to be currently investing in the equipment likely required for a direct assault on Taiwan, such as large amphibious assault ships and medium landing craft necessary for a large beach assault.”

    Beijing’s current naval build implies that the prerequisite for taking Taipei is neutralizing the U.S. navy. “China’s recent spate of military exercises and the PLA Navy’s focus on building large aircraft carriers, escort cruisers and amphibious transport dock (LPD) ships suggest the military, for now, is geared toward blue water naval operations and smaller expeditionary missions.”
    In other words, the CCP thinks Washington must “fall” before Taipei can be stormed. That strategy could change, but not overnight. Since an invasion of Taiwan would have to be on the scale of Operation Overlord or Iceberg (the invasion of Okinawa) to overcome 22 defending brigades, the buildup could not easily be concealed, and there should be no surprises….

    Looming large over the scene is the perception that China is becoming more dangerous because it is a declining power. As the American Enterprise Institute put it, based on a recent piece in Foreign Policy, “The United States needs to prepare for a major war, not because its rival is rising but because of the opposite.” …

    If China is actually declining, not only economically but demographically, then it doesn’t have the time to strangle Taiwan before itself collapsing. Taiwan becomes a problem of Now or Never for Beijing, with all the disastrous risks desperation elicits. But it’s not game over if Joe Biden can match Chinese decline with a collapse of his own. Alex Lo of the South China Morning Post poses the question: “Suppose both China and the US are in decline …?”

    ‘So, I ask our wise political scientists: if both superpowers are in decline or have peaked, are they more or less likely to go at each other’s throats? Well, judging by the aforementioned discussions, if both sides are in decline, wouldn’t both be the problem? And that’s the real problem for the rest of the world.’

    … A scenario where entropy wins means that, rather than achieving an End of History similar to what followed the fall of the Soviet Union, the struggle between the U.S. and China will result in a collapse of the old Global World, succeeded by an unsettled period like that after World War 1….

  18. Alex Lo is old, semi-retired and living in Toronto, and he runs hot and cold on most topics. He plagiarises a lot of stuff. Just saying. I wouldn’t recommend him as a primary source on China. I don’t have anything against him, particularly, and usually read his column because it’s short, but he’s not brilliant. Tanner Greer is normally a lot better, if you don’t mind the long reads.

    I realise that everyone probably has virus fatigue, but I was alarmed to read this: “Johns Hopkins University, which has been tracking the Covid-19 cases and deaths since the start of the pandemic, said on Wednesday that more Americans have died of the virus in 2021 than in all of 2020 – regardless of Biden’s mask mandates and more than half of the US being fully vaccinated.” That really shocked me.

    I have to say that initially, the USA did very badly. But the speed of production of vaccines was sheer brilliance, and then when it began vaccinating, it was doing extremely well (certainly in comparison to Hong Kong – you need to vaccinate 330 million, we only need to do 7.3 million). Now it seems to be doing very badly again.

    HK’s vaccination effort has been shameful – we had the capacity to vaccinate the whole population by July with ease; we had enough vaccines, we have a super efficient network of community vaccination centres…and yet we still have fully vaccinated only 63% of the eligible population (i.e. excluding kids younger than 12 and the small % of people who can’t be vaccinated due to health conditions). The age distribution is bizarre – the least vaccinated are the 70+ and 80+ year olds. https://www.covidvaccine.gov.hk/en/dashboard

    But we have succeeded in keeping Delta out, at least so far. HK has the world’s most fierce entry/quarantine requirements. In 2021, more people have died in hiking accidents so far than from Covid-19. But of course that means that we are totally cut off from the rest of the world. And, not trivially, from the Chinese Mainland.

    But when Delta does get into HK and gets going, and it will sooner or later, it’s going to be a sh1tshow. The hospitals will be overwhelmed, and many people will become severely ill and/or die.

    And with such a low % of the whole population vaccinated, the government’s panel of ‘expert’ advisers are not even willing to consider administering boosters, even to the elderly and immunocompromised. They don’t see a point. They say they will reconsider if HK has an outbreak of Delta, failing to understand that by that stage it will be too late, because Delta is too difficult to control – if the Mainland has struggled to control Delta outbreaks with the population controls at its disposal (which it has), you can bet we would. I understand that, and I’m a civil engineer, not a virologist.

  19. @Matt

    A very hypothetical partition of BEL024 by adding up respective “northern” and “southern” sources it picks up in a model using all modern Europeans, including Jews and related populations like Cypriots and Anatolian Greeks.

    https://pastebin.com/Zb6ad7PW

    There are obvious issues, e.g. a population on a cline between the southern and northern ancestors might be the actual southerner though I’d say no European populations are exactly on a cline between the Greek islanders the southern one comes out closest to and Belarusians. Still the Balkans isn’t far off from that so there are obvious issues by picking extremes in that kind of model. Or it might be something relatively similar as that hypothetical southerner, like something Jewish-related without some excess eastern European (or generally northern European) related ancestry the moderns have that might even be getting absorbed by the putative northern source.

  20. One southern paternal great-grandparent, with the other 7/8ths being Belarusian(-like) under this very hypothetical model.

  21. @Forgetful, nice idea. And now I’ll be a bit annoying and say why I think it doesn’t *quite* work.

    So my suspicion is that the flow that makes Belarus_Medieval:BEL024 slightly different to later Belarusians is something which isn’t well represented by modern Europeans – it’s “paleo-Carpatho-Balkan”, without the “East Med” ancestry that Greek Islanders and the modern Balkan cline has.

    To sort of confirm this, I’ve run Vahaduo against a set of *all* ancient European averages (excluding outliers, low_res and contaminated samples), and I’ve take the ancient results and split them like you suggested into a Northern and Southern ancestor: https://imgur.com/a/uKVKqdR

    In this case the Northern ancestor is a composite of Copper Age to Medieval sources chiefly from Russia and the Baltic, while the Southern ancestor is a composite of Bulgaria_Beli_Breyag_EBA and early Bronze Age (pre-Mycenaean) Aegeans.

    Crucially, under this set up, although there is the option for it to take Roman_Imperial, the model doesn’t want it. This model does not select this at all as a proxy for East Med ancestry, and just sticks to the EBA Bulgaria+Aegeans.

    Now testing your two based on moderns against mine based on ancients: https://imgur.com/a/uObn7Oh

    The two “ancestors” based on ancients seem to be superior to the “ancestors” based on moderns. Not by any vast supremacy, but its detectable.

    Not a major difference, but based on this I would expect that it’ll make more sense, to me at least, for this Belarusian sample to have a bit more ancestry from a “paleo-Balkan” group, rather than the later East Med infused groups from the Balkans. There might be some East Med like pulse into it, but it looks more Paleo-Balkan.

    I would expect this sample, knowing nothing about it, to potentially be representative of a population at its time, and later Belarusians to have more ancestry from the Northeast. But we’d need more samples to know if we should abandon our “Copernican” assumption that we’re not currently standing at any special place in regards to the sample record.

    (Pastebin: https://pastebin.com/N3CSzNSY)

  22. @Forgetful, I replied to your post but think it hit the filter. The jist of it is that I was inspired by your post to try with *all ancients*, and then made a Northern and Southern “ancestors” based on that. The Southern one seemed to pick purely Aegean+Bulgarian EBA sources, despite availability of Romans. The results seemed to give better fits than the ancestors based on modern people. So I think it’s more likely that the sample represents a larger northward population movement from Balkans rather than an isolated guy who is an outlier that we’ve just picked up, with an unusual grandparent.

    (Bit of an odd conversation we’re having here between two sets of blog comments 😉 ).

  23. https://www.cell.com/ajhg/fulltext/S0002-9297(21)00342-6 / https://phys.org/news/2021-10-artificial-intelligence-narrative-africa.html – “Revisiting the out of Africa event with a deep-learning approach”.

    Think this was a preprint (think I remember something Pagani and Yelman like this on Biorxiv). Two layer model for West Africa involving migration from a proto-OoA population from NE Africa and a Southern African population which contributes less to them today (10% seems the estimate).

  24. @Matt

    Not at all annoying since I feel it’s a more reasonable a priori point than my attempt via moderns anyway, I saw you making it in your previous posts too so I definitely had it in mind when using that other kind of method to basically see what it’d throw up. The northerner I created via moderns does assume a somewhat more western position (both in west Eurasian and northern European PCA), at least, than the modern Belarusian set in G25 too so it basically agrees with what you’re saying. A priori it also makes more sense to assume it’s not an unusual sample of the area than that it is, even if the sample’s Y-DNA might give the impression of some potentially more southern connections, especially since it doesn’t have anything that’s very clearly “exotic”.

    In extremely slight defense of the model via moderns I will add that the plausibly early Slavic individuals of the same period that are more HG-rich like three of the Krakauer Berg ones or Av2 plot a bit like that northern ancestor, in which case we might assume the Belarusians have received further admixture from both directions to re-assume a somewhat early Slavic-like position (another more general qustion that might have some bearing here and has come up in other conversations is how varied we assume the expanding, early Slavic population to have been; do Av1, Sunghir6 or even RISE569 represent the existing variety or very early admixtures while expanding where Av2 and the Berg samples are somewhat closer to the actual early average? even the LVA and EST BA sets are quite varied which means we probably shouldn’t pay attention only to the more Baltic-like individuals though some more samples wouldn’t hurt to see what’s up).

    And even if we do suppose some recent flow compared to the general population, the model via moderns might point to it from very close-by regions, not something far off like that, if we assume the modern eastern European cline has been basically established by that point with potentially small differences to the existing one or it not being captured that well with the two hypothetical endpoints since the model’s distance isn’t fantastic anyway.

    Now if we had a few more samples from the same necropolis…

  25. Yeah, I did run off all pop average distances against our ancestor simulations, and the modern model NorthernAncestor is closest to Belarusians (among moderns) and DEU_Kraukauer_Berg_Middle_Ages (among ancient dna), while the ancient model NorthernAncestor is more Lithuanian/Estonian. The Southern Ancestor for modern is like Greek Islanders and Imperial Roman while teh ancient Southern Ancestor is less well proxied by any group but of course close to its primary contributor, Bulgaria_EBA. (Distances: https://imgur.com/a/9pVZ7Z3)

    Think we definitely would need more unrelated samples from the same necropolis to sort it out!

    On other things IE related:

    1) https://www.cambridge.org/core/journals/antiquity/article/return-of-the-beaker-folk-rethinking-migration-and-population-change-in-british-prehistory/ABF13307796A0476353FA8D2DA38A21A – This, which was a preprint, got published; no new adna, David Reich and Ian Armit talking about “Steppe Drift Hypothesis” vs “Beaker Colonisation hypothesis”.

    2) https://ioa.ucla.edu/content/marija-gimbutas-magnificent-vindication – some virtual event around Marija Gimbutas work on 22/10/2021; David Anthony might say something interesting about unpublished adna.

    (Massive vindication though, yeah, kind of – “One of her most original (and controversial) contributions has been validated recently by aDNA: the Kurgan Hypothesis and the arrival into Europe of the Proto-Indo-European speakers around 3500 BC. Gimbutas suggests 3 waves of “Kurgan culture” – https://en.wikipedia.org/wiki/Kurgan_hypothesis – over about 2000 years, of which only the least one seems to have actually happened as far as I can tell from adna, and which as Anthony already suggested, only the latest was actually likely from linguistic “paleontology”.
    And check out this review of Childe by David Anthony in 1986 – https://www.jstor.org/stable/2743045 – “Childe (1926) assumed that large-scale migrations would have been associated with the prehistoric diffusion of the Indo-European languages, and he therefore searched the archaeological record for a material culture horizon that was distributed widely enough to qualify as a the archaeological manifestation of that difusion. Linguistic evidence suggested that any such horizon should be located in the temperate zone, should represent a culture familiar with copper/bronze and wheels, and should predate the 2nd millennium BC. The spread of what was then conceived as the Corded Ware-Battle Axe-Tumulus Burial complex, at the beginning of the Bronze Age, provided him with a qualified candidate. He saw this complex as originating in the Ukrainian North Pontic steppes, a region favoured by some linguists … Moreover he identified the bearers of the Pit-Grave (or Yamna) culture, a then poorly understood Bronze Age culture from the Ukraine, as the probable original speakers of Proto-Indo-European languages, and as the crucible in which the Corded Ware-Battle Axe-Tumulus Burial horizon was formed.” This paper also contains a comment but Gimbutas where she mentions her likely wrong 3 wave theory, and the gender-religious stuff that seems likely to be quite wrong. But eh, probably close enough and split the difference. The full fledged developer and bearer of the theory deserves a fair bit of credit.)

  26. @Matt

    In his podcast episode with Razib, I recall Mallory mentioning that Gimbutas was a “lumper rather than a splitter” and you certainly get that impression reading her work. That probably had an advantage in this case (versus splitting up the steppe in dozens of variants like some Russian archaeologists seemed to do), even if she was too inclusive of various things that don’t seem steppe/IE-related at all. Though at least in some situations where she has been shown quite wrong that have been brought up in these discussions, she was a bit more reserved I feel. Something like her suggestion that GAC might have been “Indoeuropeanized” comes to mind where she didn’t exactly seem that forceful about it.

    Has been discussed a bit before but I think we can tentatively at least argue for one more “separate wave” with regard to some potentially early Balkan samples and that low-quality Kumtepe (Kum4) one, so plausibly the Anatolian one. I don’t recall her specific dating for the three waves, though certainly some of the supposed far-reaching migrations of these earlier waves (all the way to parts of far western Europe, reinforced by the last one) don’t seem to be there.

    I think the anthropological difference I mentioned in a previous thread that she saw happening between Dnieper-Donets and Sredni Stog (more robust but shorter -> taller but more gracile and with likely connections to the east) is in her response to that article by Anthony too. In general the impression I got from reading that article a while ago is that Anthony seemed to consider more a local Dnieper-Donets origin of PIE while Mallory and Gimbutas in their responses argued more for a Don-Volga/eastern PC steppe one, or at least in Mallory’s case a more tentative need for consideration of the eastern part of the steppe too. The Wheel, the Horse and the Language seemed to move more towards that latter kind of argument too I feel though I wouldn’t want to misrepresent Anthony’s evolution in his thoughts about more specific origins. In recent presentations and discussions, he also mention that he thinks this eastern Don-Volga “mating network” is at the root of PIE. Even harder to argue with that now!

    Childe’s The Aryans was an interesting read from a historical (and maybe not just, I recall thinking that some of the specific arguments holding up well even if the knowledge at the time was much more limited, not quite unlike the small part Anthony relates there) perspective though I also remember very little from it now. My recollection is that he didn’t seem that certain that a northern European/Scandinavian (“that the Germanists insist on” to paraphrase the kind of popular sentiment at the time he rails against in his work a bit) origin could necessarily be excluded at the time compared to a steppe one though he preferred the latter and thought that the early Indo-European area would end up being too large and differentiated for a people that seemed quite united until recent times to include both general regions at the same time.

    Shame neither of them are around to see much of what they argued for be vindicated. Even other important archaeologists like Renfrew in a 2017 presentation (“Marija Rediviva”) seemed quite happy that we’re moving towards a conclusion of some very long debates, despite their specific argument in that case being wrong. “A magnificent vindication” does seem quite appropriate, the failure of more specific arguments aside. Let’s see if we get any video of the segments.

  27. Probably entirely fair to be honest (and you have a superior knowledge of this lit)! My tendency to nitpick.

  28. Just some recollections triggered by the interesting links you mentioned from having read some on the subject over time as your average, interested amateur naturally without any expertise on it at all really.

    That kind of specific criticism seems pretty fair and Gimbutas herself was much more savage when e.g. reviewing Renfrew’s work. Certainly, most of us wouldn’t know what we’re even talking about in general terms about this issue if it weren’t for the contributions of all those people (so some more egregious comments you can sometimes come across in other places about this or the other well-known archaeologist involved with the general issue at some point do rub one the wrong way) but they have obviously been far from always being right.

  29. Re; Kum4, yeah, that’s true, although I really wouldn’t want to bet the farm on her – 11,000 SNPs plus “Very_High_Contamination” in ContamLD (Reich lab anno), plus some strange behaviours in modern linked G25 dimensions (she always seems to bounce around in odd ways relating to the Balto-Slavic dimension when I look at G25) is, er, hair-raising to me. I think the Balkan samples are there as well as you say for this sort of thing (unless they’re all becoming labelled as contaminated too – Smyadovo: I2181, I2176 and other Bulgarians Varna: ANI163, Dzhulyunitsa: I2509, I2519 all have a fair bit of uncertainty in the anno). Kind of nervous about this given the prevalence of labelled contamination and low SNPs at the moment.

    As a very banal statement, I think one thing that might seem surprising is how differentiated the groups in contact with/via the steppe were. We’ve got: 1) EEF farmers, 2) Ukraine_N: EHG-like Dneiper-Donets fishing groups, 3) possible EHG-CHG mixed group since the Mesolithic on southern PC steppe/Volga Basin (Nalchik / Khvalynsk), 4) EHG proper, 5) Darkveti-Meshoko and then Maykop cultures, 6) Steppe_Maykop like people with EHG-CHG-West Siberian. In some ways could we say this is the region with the most interacting genetic diversity at this time? (Certainly relative to its population size?) Even though 3) and to a smaller extent 1) left the longest term genetic trace. It seems like there is an unusual degree of trade between many different separate communities that didn’t really merge on the whole, to spread domesticates and ideas.

  30. @ Spike – You are right. Some people are doing more than paying attention now, they are starting to panic, and they have good reason to.

  31. @Forgetful, e.g. trying Kum4 in Vahaduo with Global25 – https://imgur.com/a/M0alHgV . Just gets this insane high distance, and combinations of populations which partly make sense (about 50% Turkey Late Copper Age), and partly don’t (50% some combination of Native American and present day Eastern European and African etc, rather than actually a sensible steppe group of the time). I think it’s still the case that we don’t really have steppe ancestry in Anatolia, confidently at anything like this time.

  32. @Matt

    To continue your theme of doubt (and a lot of this has been discussed but might as well receive some more feedback on the dating aspects etc. since you keep very good tabs on that), in their supervised ADMIXTURE analysis in Mathieson et al. those Balkan CA samples including the Dzhulyunitsa samples, and with the I2181 exception, all come out as having some local intermediate HG-related ancestry rather than necessarily the “steppe” component (represented by Yamnaya_Samara at the time) so they probably don’t have any of that either. And ANI163, which always looked oddly almost modern Balkan Slavic-like from what I recall might very well be wrongly dated as they’re claiming now so it’s probably a good idea not to consider it at all.

    We do have those more recently published late 4th millennium Trypillian samples with some apparent steppe-related ancestry from Pocrovca and Gordinesti that date roughly to our earliest Balkan BA samples from Mathieson et al. like I2165 and I2520 (this one also from Dzhulyunitsa, and with reliable steppe ancestry). I suppose it depends on whether we also consider this a separate “wave” or just part of the general movements that start affecting Europe roughly during this period. Arguably something like the unreliable Kum4 (dated ~3500-2900 BC, basically contemporaneous to those), if it actually has any steppe-related ancestry, might derive from that.

    In other words, you’re very right as far as I can tell. Still, we have some very tantalizing samples from areas that would be nice to be sampled more in the future during that period. Maybe there’s nothing to it and that early sort of “wave” just petered out without producing any later, known branches. Even Gimbutas with all her waves and far-reaching, constant impacts thought there were those sorts of migrations too that ended up being fully assimilated instead in “Old European” societies. Maybe the one Protoboleraz I2788 (early 4th millennium) can be brought up here again too, because it does seem to deviate towards more of a steppe-like population rather than the HG cline, compared to the others; another very maybe representative of those potential, early short-term impacts that were fully assimilated. It might be a complete mirage even but interesting nonetheless if we ever manage to capture that kind of thing, if it exists, more reliably.

    Rough dating-wise, if we take some of the less certain samples or uncertain admixtures at face value, _maybe_ even “three waves” could be argued? One during the early 5th, one during the middle to late 4th and another (the big one) during the late 4th to early 3rd? That later period looks hard to split up like that though, it just looks like steppe-related horizons and ancestry are expanding all over the place at roughly the same time, like you said.

    The last part of your comment is interesting. With the exception of that Don-Volga group, picking up some maybe 10-15% potential ancestry from those CT groups and maybe some extra UKR_N-related ancestry at least in its early CW variant, most of those groups seem to have much more localized impacts in the long run. A community that at the time might have seemed peripheral from at least the southern and western ones’ perspective, and surrounded by all those other groups too, ended up carrying some potential advantages (some of which, like the wheel and the wagon it might have gotten from these others in the first place) quite far. We kinda expected that in linguistic and even archaeological terms at least under a particular model, but we can see it in genetic ones now too.

  33. @Forgetful, interesting to note around I2181. I put all the Bulgarian samples that made it through to Global25 on Vahaduo PCA yesterday – https://imgur.com/a/CDOpMxR – and as ANI163 is “contaminated”, that seems to leave only I2181 as the evidence of an early sample.

    Actually even the later EBA sample I2165 with a substantial pulse at 2910 BCE is contaminated.

    The only samples who are left that are “uncontaminated” who also have steppe ancestry are two samples from approx 3100 BCE, I2520 and I2175, who have about 13% Steppe ancestry between them (roughly equivalent to one great grandparent) and have y-dna H and I2a1b1a2 (not sure if the latter is a steppe related clade).

    Those two at 3100 BCE are pretty dispositive towards the idea that it’s the late 4th millennium that mattered for the dispersal of steppe ancestry, just on the cusp of (possibly within as little as 3 generations of) steppe ancestry dispersal.

    So it does seem like I2181 is the crucial sample for evidene of early steppe ancestry (well before late 4th millennium) at the moment, and the higher coverage version of him that is unpublished would be important. Like you say, the earlier samples seem to fit with a HG population that’s on the WHG->IronGates->Ukraine cline and not “steppe”.

    The Trypillians are interesting in that the two samples that are dated early, I1926 and I2110, at 3750 BCE to 3650 BCE, basically reject steppe ancestry (they’re just Barcin+IronGates in Vahaduo), while the later samples after 3500 BCE, closer to the cusp of steppe expansion, dated at 3340 BCE to 3230 BCE, *do* have steppe ancestry, as you say, they date roughly to the earliest Balkan samples. It’s all very convenient for us thinking that its the period after the big expansions in mobility that Anthony talks about after 3500 BCE, that is actually most important. Agree, we could talk about this as waves or a continuous change that builds up with some early signs beginning in the mid 4th millennium. Good point re Protoboleraz I2788 at 3680 BCE, that is a *bit* more precocious.

  34. @Matt

    Can you share these preliminary G25 coordinates for the Gordinesti-Pocrovca Trypillian samples? Hadn’t come across them.

  35. As another comment, the two Trypillian samples I1926 and I2110, from Verteba cave who seem to me without steppe ancestry, coincide with the “Middle Trypillia phase (c. 4000 to 3500 BCE)” of large settlement building, while the folks with some steppe ancestry come from the “Late period (3500–3000 BC)” where, per wikipedia “Members of the Cucuteni–Trypillia culture who lived along the coastal regions near the Black Sea came into contact with other cultures. Animal husbandry increased in importance, as hunting diminished; horses also became more important. Outlying communities were established on the Don and Volga rivers in present-day Russia. Dwellings were constructed differently from previous periods, and a new rope-like design replaced the older spiral-patterned designs on the pottery. Different forms of ritual burial were developed where the deceased were interred in the ground with elaborate burial rituals. An increasingly larger number of Bronze Age artefacts originating from other lands were found as the end of the Cucuteni–Trypillia culture drew near.”.

    That’s interestingly concordant with an expectation (if we had it) that it’s the late phase that drew in people from the steppe/pre-Yamnaya culture. (And also since the Anatolia:HG proportion seems highly variable, with diversity within the CT horizon? The early samples are more like TRB/Funnelbeaker with maybe a touch more eastern HG ancestry?).

    Unfortunately all the late samples are female, so we can’t say much about haplogroups (I had got the idea from somewhere they were all G2 and E1, but it seems that only the males from Verteba were!).

  36. Thanks a bunch! Those 3 might have been in past versions that I didn’t catch but they don’t seem included anymore. I notice Pocrovca1 in paticular gets pretty high distances to other samples, even if the relative distances and the position on the WE PCA seem about right, so I’m guessing there are quality considerations with it at least.

  37. Razib, a while back you talked about a book which (at least in part) dealt with the economic impact of Buddhism. Do you remember the name?

Comments are closed.