More on gene flow

That guy on the internet was dead wrong in thinking that there was enough gene flow in human prehistory to materially interfere with local adaptation. For a polygenic trait, that would take something like a couple of percent a generation.  For an adaptive Mendelian trait,  you’d need more influx ( as a fraction) than the selective edge of the causal mutation.

But there’s another interesting question, which he would presumably also be wrong about:  how often does even a single copy of an adaptive mutation manage to travel far, and how long does it take? I talked about this here, some time ago.  But examples help.

HbS, the sickle-cell mutation, happened once about 7,000 years ago, and has spread to regions with a lot of malaria in Africa, parts of southern Europe, the Middle East, and India.

But it never got to southeast Asia or New Guinea ( until 2002) , even though there’s plenty of malaria there, and a number of local genetic defenses against malaria. Parenthetically, although HbS is found today on different haplotypes, and many people suspected that it had multiple origins, complete absence in Southeast Asia/Melanesia was always a strong hint that it had a single origin.

None of those defenses in PNG or Southeast Asia ever seem to have made it to Africa.  And during the last 7,000 years, there were an increasing number of long-distance dispersal mechanisms that didn’t exist before the Holocene:  lost-distance trade, ships, the slave trade, etc.


This entry was posted in Uncategorized. Bookmark the permalink.

41 Responses to More on gene flow

  1. RCB says:

    Almost everyone wants the answer to this question to be that there are no important genetic differences between groups. As a consequence, I think, people will readily accept whatever argument the establish geneticists will provide to support that belief. Even BS theoretical arguments. And folks who know better will mostly look the other way – happily tolerate stupid answers, if only to avoid the subject. So only a few years ago grad students were still arguing to me that there couldn’t possibly be genetic diffs in IQ among groups because “only 15% of variance is between groups”. They had heard someone like Lewontin tell them that this was a compelling argument, which of course it isn’t. (I haven’t heard anyone make that argument in a while – maybe it’s been iterated through enough. Replaced by others.)

    Also, you gotta be careful with guys on the internet. I was debating this stuff with someone who looked like they might be a reasonably smart grad student. After he showed incredible unreasonableness, I quipped something to the effect of “you’re so dumb you probably think communism deserves another shot”. To which he said “absolutely” and that “communism is a good political and economic system”! By all signs he was being sincere.

    • Smithie says:

      I think it shows something about the heritability of politics that so many want to have another go at communism.

    • Jacob says:

      “Almost everyone wants the answer to this question to be that there are no important genetic differences between groups.”

      The arrogance to place demands on God or the Universe like this is disgusting to me. Excuse me, but how fucking dare you? You are an organism, constituted of materials from this Earth, subject to the limitations imposed by physics and evolution. You don’t get to tell your own natural history what it was supposed to turn you into. You just have to suck it up. Move on. Grow the fuck up. Embrace the fact that the world isn’t exactly like you would want it to be.

      But I don’t think these people have any real investment in whether race differences are real; I think they’re invested in whether they themselves believe it’s real. They don’t care how the world actually works; they just don’t want to be thought criminals. I think they just lack the courage to have even the slightest interest in giving a fair shake to heretical ideas.

  2. Joe says:

    This is an interesting argument because the original line of defense raised by Gould, that 50,000 years wasn’t long enough to change cognitive ability through evolution, was wrong, and so now the argument goes that there wasn’t actually separation anyway. You have to admire the gumption of these folks – at least they have hypotheses that can be tested. Although the guy you got into on twitter with didn’t seem to be swayed by your argument, and chided you for your ashkenazi research.

    • gcochran9 says:

      Since it is easy to show that you need a lot of gene flow (on the order of 1%) to suppress local adaptation, the kind that is impossible without advanced transportation – easy mathematically, and since we see such adaptation all over the place – he has to be stupid or lying. Why do I have to admire that?

      I see stupid comments from biologists all the time. Lies also.

      • Joe says:

        I suppose you don’t have to admire it, but at least to a layperson, it is a step up from non testable hypotheses! Anyway thanks for addressing these. I’ve been having Reddit debates and the lines of argument now seem to be “maternal effects” (whatever that is), and gene flow. Will it never end?

        • gcochran9 says:

          Where are the comments about gene flow? Link?

          • Joe says:

            Long story but I commented on this Atlantic article ( by suggesting that it reinforced the idea of races. That got me a lot of hate responses but relevant to this was the following from a PhD student in genetics:

            “That’s not really what the atlantic article is about though, and ancestry and races don’t quite function like that either. I’m not really sure what you’re trying to defend or claim. Humans vary, but the variation in humans does not reflect a racial taxonomy nor can that variation itself really be called race.”

            Then I said:

            “Yes I suppose you’re right about the article. But I think the variation in humans can reflect a taxonomy that roughly corresponds to “racial” categories- those separated by geological barriers for thousands of generations.”

            Which got the response:

            “The variation potentially could reflect a racial taxonomy, but it does not. The barriers have not restricted gene flow enough or for long enough.”

            Then I said:

            “I’ll have to read your other response more thoroughly but I don’t see how that’s possible, the length of time seems to keep being extended by archeological findings.”

            And the response:

            “Archaeological findings can be spotty due to conditions in certain areas, and may not even represent populations that contributed to the current human lineage. Genetic patterns get around this though so looking at DNA and the way it differs around the globe then comparing that to simulated data or what patterns would look like given isolation/splitting of lineages is a good way to formally test the question of whether human races are independent evolutionary lineages and if populations have been fully isolated.

            Gene flow varies between population pairs and humans aren’t completely panmictic, but that’s a long shot from having distinctly divided populations due to isolation.”

            I’d never heard the word “panmitic” before but I’m pretty sure it doesn’t prove we are all the same.

            • gcochran9 says:

              All nonsense. The question is whether various human populations are genetically different from each other, different in ways that people give a shit about. And they are.

              German Shepherds are not a “lineage race”, but they exist.

              As for a “single lineage”, that’s meaningless.

              It wasn’t Kevin Bird, was it?

              • Joe says:

                It may have been kevin bird, because their twitter moniker contains the name “Kevin.”

                What do you think about this argument from Alan Templeton, “proving” that populations haven’t been isolated enough to make “races?” ( My sense is that he is defining races in a specific way (almost like species), although Kevin rebuffed that suggestion.

              • Joe says:

                Oh I see – Templetons gambit is the meaningless “single lineage” idea!

              • Joe says:

                I mean reddit moniker: stairway-to-Kevin. He polices the “genetics” subreddit for crimethink.

              • gcochran9 says:

                He didn’t impress the guys giving him his quals, concerning population genetics. So I hear.

              • kot says:

                “He didn’t impress the guys giving him his quals”

                What did you hear, exactly?

                He did mention at some point being rejected from every grad school he applied to, his first time around.

              • gcochran9 says:

                I heard that someone on his quals thought he didn’t know jack about population genetics, which is obviously the case.

                About grad schools rejecting him, wouldn’t know.

            • RCB says:

              IMO getting into a debate about the concept and validity and existence of races is a big waste of time. It usually invites all sorts of abstract arguments. Often the person who says the most complicated thing wins, to those reading.

              Stay on a concrete topic. The IQ differences between US whites and blacks is a conrete thing that actually matters in real life. You don’t even have to believe that blacks constitute a “real” race, or whatever, to talk about this.

          • Joe says:

            Sorry, this is a much simpler exchange on the same topic with links to research papers:


            Yes, gene flow has been allowed between virtually all populations, but it hasn’t occurred over geological barriers over long periods of time.

            You seem to be saying that there are differences among all groups, but the fact is that populations separated for long periods of time are going to be different in some ways that are distinguishable from the differences within those groups.

            And the response:

            “It has though! A few papers set toward analyzing the question of race in human populations provide some clear data.

            Alan Templeton used multi-locus nested clade phylogeographic analysis and concluded that there hasn’t been a point in which human races have been separate evolutionary lineages (basically gene flow has been sufficient to maintain a single lineage)

            from his analysis in Templeton, 2013

            Following the expansion with admixture of modern humans from Africa, there have been additional expansions, mostly into areas not formerly occupied by humans (Figure 3). Wherever humans lived, gene flow was soon established, mostly limited by isolation-by-distance but with some long-distance dispersal in more recent times. These inferences of long distance range expansions followed by gene flow mostly constrained by isolation by distance have been subsequently supported by extensive computer simulations (Hunley, et al., 2009). On a time scale of tens of thousands of years (the temporal resolution of the ML-NCPA studies), there is not one statistically significant inference of splitting during the last 1.9 million years. Hence, the null hypothesis of a single human lineage is not rejected, so there is no evidence for lineage races in humans. Furthermore, ML-NCPA strongly rejects the null hypotheses of no gene flow and no admixture under the null hypothesis that isolated lineages did exist, so there is strong evidence against lineage races in humans. Hence, there are no races in humans under the lineage definition.

            Likewise, Hunley et al. 2009 simulated data to identify patterns matching human genetic diversity and concluded

            The gene identities for the CEPH microsatellites are shown in Figure 2E. The most salient feature of the left panel is the distinctive strata of gene identity at the within- (pink triangles) and between-region (blue circles)levels. The distinctive strata are not consistent with the independent regions model, which predicts a single between-region stratum


            The Sub-Saharan African results have important implications for race. For Sub-Saharan Africans to belong to a single race, all Sub-Saharan African populations would have to cluster together on a single branch of a larger species tree, which means that each population would have about the same level of gene identity with all non-Sub-Saharan African populations, i.e., the individual population strata in Figure 3D wouldn’t exist.These separate population strata indicate that the Sub-Saharan African populations do not cluster together,and, therefore, that there is no Sub-Saharan African race in any taxonomic sense.

            For race this means

            If the independent regions model was correct, the inindividuals in the same geographic region would on average be more closely related to each other genetically than would be individuals in different geographic regions. Even in this case, the problem of finding a threshold level of gene identity for declaring taxonomic significance would remain unsolved.In reality, the between-population gene identity pat-tern is nested. Because the between-population pattern is nested in Sub-Saharan Africa, and because Sub-Saharan African populations straddle the root of the species-wide population tree (e.g., Li et al., 2008), there can be no Sub-Saharan African race under the shared genetic relationship criterion. The first division in the population hierarchy that coincides with continental locations separates non-African populations from African populations. This division is consistent with the existence of a non-African race, but because of the root, the Africans would still not constitute a race. Another major division along continental lines separates East Asian and Native American populations from all others. How-ever, classifiers would need to put East Asians and Native Americans into a sub-race, because they would already be members of the non-African race. Thus, we see that nested pattern of variation would require that the geographic groups that anthropologists traditionally considered races could not be assigned to the same level of hierarchical classification.

            What this means is these populations haven’t really been “separated” and there are few if any fixed differences between them that make them distinguishable or meaningful groups. The “bailey” you identify is not disproven in the article, as there is nothing about the patterns observed that is fixed in African populations that distinguishes them from other populations, nor are the differences somehow more significant or meaningful than any other variation observed (e.g. between African groups, within African groups).”

            • RCB says:

              Again I think this is all a big distraction.

              Suppose this claim is true: ” there is no Sub-Saharan African race in any taxonomic sense”

              Does this imply that African Americans (or “the people who identify as such” if you really want to belabor the point) have no biological differences with white Americans? Of course not. Does it mean that blacks don’t actually have lower IQ? Does it mean that genes could not possibly contribute to that difference? Of course not. We know this because they do vary in other phenotypes for known genetic reasons.

              People will pontificate endlessly about the race concept, but it doesn’t really matter. Concrete questions about real human populations matter.

            • savantissimo says:

              Seems to be arguing that there is no one, single African race, not mentioning that there are at least three, so arguing that there are zero African Races. All the nonsense about nested hierarchies seems to be a restating of the well-known family tree model in bafflegab, then concluding that there are no branches because you’ve decided to call them “nested hierarchies”.

              There is also a bizarre implication that because the root of the tree is in Africa that descendant lineages somehow can’t also be in Africa. Perhaps he thinks that the root population still exists, unevolved, in Africa. That’s testable, though I’m sure he’d be shocked at the very mention of the idea, having thought he kept that implication to himself.

    • Jim says:

      Well in the case of Australian Aborigines they have been there for something like 50,000 years with very little evidence of gene flow from outside until very recently.

      The Khoisan are probably another example of a very high degree of genetic isolation from the rest of humanity for a very long time.

      • gcochran9 says:

        They’ve picked up Bantu ancestry over the past couple of thousand years: from 10% on up. But before that, they look to have been almost entirely separated from other populations for more than a quarter of a million years.

        • Jim says:

          In the late 18th century the Australian Aborigine population was estimated to have been 500,000. So a rather small population close to totally isolated from the rest of humanity for some 50,000 years or so and with a very primitive level of technological development. But magically they are behaviorally exactly like a typical resident of Tokyo.

        • Capra Internetensis says:

          Also some other recent East African ancestry, more evenly spread than Bantu.

          Apparently there was an ASHG presentation last fall arguing for Khoisan-Pygmy-Hadza admixture dating to 30 or 40 000 years ago. I believe it. Unless current estimates are wildly wrong, the Y chromosomal divergence between old school East/Central Africa and South African groups is more like Europe vs China than a quarter million years.

      • adreadline says:

        Cochran said a possible scenario for the recent (~4000 years ago) gene flow into Australia could be that a single vessel from the Indian subcontinent and its all-male crew ended up there. Those castaways would be fecund enough to leave a big mark on present-day northern Aboriginals: just over 10% of their ancestry can be traced to India. That study was from over half a decade ago, though. HbS doesn’t seem to have reached Australia, so possibly none of those unlucky Indians carried it, or it hadn’t been introduced to India yet, or it was selected against in Australia back then.

        Also, something else: it seems HbS exists in Madagascar, and since the natives there descend, at the earliest, from people who arrived from (what’s now) Indonesia less than 2000 years ago, would it follow that HbS was introduced there even more recently from East Africa? And if so, does that also imply those recent East African arrivals overwhelmed the original, Southeast Asia-descended Malagasy population? Or not at all?

  3. Lior says:

    TBX15/WARS2 in almost all Inuits but at lesser frequencies outside of greenland.

  4. magusjanus says:

    Problem is Greg is doing it backwards. He’s looking at evidence and then coming to a conclusion.

    That’s not how ‘real science’ is done.

    Instead you already know what the conclusion is, and then you try to find any evidence to support it while arguing away any evidence that doesn’t. If one argument gets taken down, replace it with another. And another. And another. What you don’t do is change your conclusion.

    So: conclusion is racial equality no matter what, period. Which means 50k years isn’t long enough for adaptation. Oh wait it is? well that’s okay cuz too much gene flow inhibits differentiation. Oh but there is some obvious differentiation? well that’s ok cuz more variety within races than between. Also something something epigenetics do you even keep up with the literature bro?

    • Michel Rouzic says:

      It’s always like this, before they start investigating they’ve already concluded that one option cannot possibly true because they find it distasteful. They do scientific research the way a lawyer would, and never pause to think about how their approach is so flawed that they could only possibly be right by accident (that would involve their preconceptions coinciding with reality, for once).

      I’d love for them to answer the question “Imagine a parallel world in which black people are genuinely generally less intelligent, what would the facts have to be for you to admit that it is the case?”. It’s a question of falsifiability, and we know what the answer is, no matter what the facts could possibly be the distasteful conclusions can never be reached.

      • gcochran9 says:

        How would you tell the difference between that world and ours?

        • Michel Rouzic says:

          I didn’t say there would be any difference, just that it would be a parallel world ;). But I’d verify how different it is by comparing test scores, academic achievements and real world achievements, after all that parallel world could well have smarter black people than ours.

          Btw I just read Taleb insist that medical doctors have a high IQ because of “circularity”, as in you need a high SAT/IQ score to get into medical school so that’s the only reason why MDs have high IQs. We should encourage him to put skin in the game and start a medical school for lower IQ applicants, if he’s right he will find a vast untapped resource of perfectly good medical talent, if he’s wrong we’ll have fun reading news articles about low IQ medical students using bleach as a decongestant. Of course all this could perhaps be avoided by making him actually talk to people with below average IQs instead of his usual high IQ Ivy League entourage.

          • RCB says:

            “Btw I just read Taleb insist that medical doctors have a high IQ because of “circularity”, as in you need a high SAT/IQ score to get into medical school so that’s the only reason why MDs have high IQs.”

            This is such a dumb argument that I had to actually look for myself that he said it. He did.

            I mean, to be clear, it’s true: only people with high test scores become doctors, and only people who have high IQ have high test scores. The idiocy lies in his belief that this somehow invalidates IQ…

    • gcochran9 says:

      ” you already know what the conclusion is, and then you try to find any evidence to support it while arguing away any evidence that doesn’t. If one argument gets taken down, replace it with another. And another. And another. What you don’t do is change your conclusion.”

      Ron Unz is a shining example.

  5. Joe says:

    Here’s the link you asked for, 7,000 words later:

  6. Unladen Swallow says:

    The guy you are talking about, is it Graur or Taleb?

  7. Citizen AllenM says:

    LoL- now get this people-

    The real question is if this is a good foundation for the “Genetic Super Soldier” fast growth so we can get to the clone warz, lulz.

    Seriously, what is the value of fast maturity and a predisposition to violence…..

    Now who was saying that genes don’t confer advantages and disadvantages?

    Right, people who are idiots.

  8. ThoseSicklesThough says:

    My brother and father have the sickle cell trait. My father is from Dresden. That part of the family doesn’t look South European at all, they’re grey/blue eyed, brownhaired East Germans.

    Is the single origin theory proven already? How many generations would an ancestor of mine have to have it and still roughly be in Eastern Germany to be evidence for disparate origins? How could I figure out, if any of my ancestors did have it or not (other than identifying one node should about halve the search space)? It’s not maladaptive or at least not very. They probably shouldn’t do ultra endurance events from what I understand. And who should?

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s