Rapid Change in Polygenic Traits

A few people have occasionally claimed that natural selection takes a long time to change any trait that might bother them – sometimes, they say this of polygenic traits, ones influenced by many genes, like height and intelligence. Kevin Mitchell has said this (“at least 100 millennia to evolve appreciably”), so has Brad Delong (an economist at Berkeley). David Reich mentions Joseph Graves, in a talk at Harvard, saying something similar in a lecture in 2016. They’re all 100% wrong: it is easy to select on polygenic quantitative traits, and we do it every day. Dogs vary a lot in height and most of the big differences are the result of fairly recent selection: nobody had Chihuahuas or Great Danes back in the Ice Age.

Human populations vary in height – not just from differences in nutrition. We know [work from the Reich lab] that northern Europeans are around a standard deviation taller than southern Europeans ( given adequate food), and we know some of the alleles behind that. Andaman islanders are four standard deviations shorter than the Dutch , Efe Pygmies almost five std shorter.
Probably the total range in humans, from the tallest (Dinaric Slavs, 6 1, to Efe Pgmies, 4 8) is six standard deviations. I’m sure that David Reich can explain to me how this is a small, even insignificant difference.

You may ask yourself why fairly prominent people manage to say instantly falsifiable things about human genetics. Good question. I might be partially protected from this error by preadaptation, having spent years and years walking past the Morrow Plots on my way to the main undergraduate library at the U of I. Delong has the excuse of not being any kind of biologist, but then again he ought to know that. A couple of us were trying to educate him on Twitter: he started accusing the mathematician involved of innumeracy, so the mathenaut called him “just another regression monkey”. Good times. As for Kevin Mitchell and Joseph Graves – well a lot of people in biology and medicine just don’t know much about selection. And of course in this context they don’t want to know.

Quantitative selection is a lot easier than people think. If I kidnapped a year’s worth of National Merit Scholars and dropped them on a deserted but fertile island, a new race with an average IQ around 130 would develop ( unless those little brainiacs escaped. You have to watch them all the time). If I dropped a lot of NBA and WNBA players, you’d see the tallest race, if we could just get them to reproduce.

But… there are some subtle points here. Great Danes exist and persist, but they have a bundle of health problems, and they don’t live too long (8-10 years). Wolves last around 15-16 years in captivity, with a record of 20. If you wanted to create a new race with an average adult height of 7 feet, I’m sure you could, but I’d bet money they’d have bad knees.

On the other hand, if they stayed 7 feet tall for a couple of million years, they would not be particularly prone to bad knees. There would be gradual selection for tougher knees: changes in development, changes in bones and tendons and cartilage, eventually perhaps fundamental changes in the architecture of the knee. There would be lots of little changes that made development among those giants more robust, changes that reduced the incidence of many problems that centers fall heir too.

Brain size in ancient and archaic humans was plenty big, but we don’t really see signs of rapid innovation, art, and decent fast food until fairly recently, 50,000 years or so. Some anthropologists says that humans were just as smart 100 or 150k years as they are today ( which may be true for some people) – but the notion that people managed to get smarter without showing any practical sign of it for 50 or 100k years is, in my opinion, just plain stupid.

So I think Kevin Mitchell ( not the other two) has a point. It’s possible, even likely, that the populations that have relatively high IQs today haven’t had them for very long, and that they’re not terrible well adapted to their new mental horsepower. Susceptible to various mental problems and illusions that would probably be a lot rarer if natural selection had had time to iron out the bugs.

This entry was posted in Uncategorized. Bookmark the permalink.

125 Responses to Rapid Change in Polygenic Traits

  1. I imagine the NBA players would persuade the WNBA ladies to reproduce, given a free hand and a finite island.

  2. Ursiform says:

    For example, giraffes didn’t just evolve long necks, they evolved a circulatory system that lets them get blood to their brains. Evolution favored ancestors who could not only reach higher in trees to eat leaves but could also get enough blood to their brains to not pass out and who didn’t die of heart attacks due to high blood pressure. Pretty crafty, that evolution …

    • gcochran9 says:

      Just being able to drop their head down to a water source is quite a trick.

      • dux.ie says:

        The mechanism of giraffe drinking water is facinating. They do not ‘suck’ the water but by pumping the water using the tongue as plunger and oesophagus squeezing the water up against the gravity.


        Yesterday I also happened to watch a TV program on how Heston designed food for astronauts. He wanted to experience how astronauts eating curry by lying down on a platform rotated 45 degree with his head down. In upside down position Heston could not swallow the curry fast enough that it burnt his mouth and the drink of water nearly drowned him. He had to be put upright immediately. The plunger effect for human is too weak to be deployed in a hurry.

    • crew says:

      Let’s say we drop a bunch of giraffes (say, 1,000) on an island that is otherwise uninhabited by browsing species but has plenty of trees of medium height.

      How long before we start seeing giraffes with shorter necks?

      • gcochran9 says:

        We see massive shrinkage of elephants and mammoths isolated on islands in just a few thousand years. Also hippos. Key deer. Possums evolved longer life span ( 3 years instead of 2) within the last ten thousand years on an island.

  3. Frog says:

    In that twitter exchange with DeLong I think you could have actually helped him understand if you had given a numerical example with thousands of alleles of small effect, where having an extra few percent would correspond to a noticeable phenotypic effect.

    Yes, he’s being close-minded and wrong, but I think he wants a refutation of his own wrong argument on its own terms more than the voluminous examples of the processes his argument denies.

    Why not give such a numerical example and tweet him?

    • gcochran9 says:

      Feel free to do that yourself. I’ve run into him before: he’s useless. I mean, he couldn’t even bother to read the wiki.

      • Frau Katze says:

        He blocked you? That shows how close-minded he must be. That’s dreadful. Maybe he thinks anyone who read this site must be an idiot and can be safely ignored.

      • crew says:


        blockquote>Recent positive selection on newly arising alleles produces a strong genetic signature: a long haplotype of unexpectedly high frequency



        Did they mean lone?

    • gcochran9 says:

      One example of a result that he considers impossible ought to make him think.

    • magusjanus says:

      He’s (brad) not an honest debater. He’s not interested in conversation as a pursuit of ‘truth’. To him it’s basically a form over verbal combat between the side of good and righteousness (progs) and those against them (anyone else). So it’s a bit of a waste of time IF the goal is to convert him.

      BUT if the goal is to show him for the fool he is so thousands of casual readers see and thus score a small win for the real good guys, then by all means have at it. But go in with a clear mind as to what it is you’re doing: you’re not convincing Brad. Cuz that’s not happening…. until of course social forces switch to such an extent that we’ve always been at war with East Asia, whence he’ll turn on a dime I imagine.

    • peterschaeffer says:

      I have some knowledge of De Long. Very smart guy. Complete and utter a######. The bottom line is that De Long is an identity politics extremist and a raving neoliberal (he was actually one of the original authors of NAFTA). Science is trashing identity politics. Reality is trashing neoliberalism. Brad is upset. Go figure.

      Check out the comments over at ‘Rate My Professors’ for Brad. They are hilarious.

      “This guy is a legend at Cal…because everyone knows he’s the worst prof on campus. He doesn’t care about anyone, doesn’t even know his GSI’s names. Truly arrogant and doesn’t really cover anything at all in lectures. He spent more time telling us how much money he makes than talking about class material. Don’t take his class even as a last resort.”

      “DeLong’s lectures are little more than incoherent ramblings about his various accolades and pedigreed family. Pretty poor professor, but in fairness a very smart man. Boring classes with horribly derailed lectures. Easy tests though, wants you to succeed. Count on your GSIs. ”

      “Professor DeLong is a brilliant man; however, his lectures are nonlinear: goes off into narcissistic tangents regarding his decades of accomplishments, his many renowned acquaintances, his blogs, and members of his family. But if you can somehow manage to pay attention to his lectures, you should fare well on the exams. Get help from your GSI.”

      “DeLong was a horrible teacher – lectures were incoherent and problem sets were littered with mistakes. Luckily, he was an easy grader and the curve saved everyone. I barely touched the textbook because exams were open-book/open-note and ended up getting a B+ ”

      Ream them all. Who says students don’t have a sense of humor?

      • gcochran9 says:

        A priori, brilliance isn’t all that common. Has Delong done something that smacks of brilliance?

        • peterschaeffer says:

          gcochran9, Let me try to answer your question substantively. First, read the student comments over at Rate My Professors. The same themes come up time and time again. “Brad is brilliant”, “Brad is utterly pretensions and arrogant”, and “Brad is a horrible professor”. If just one comment hit on these themes, that would be potentially be an anomaly. However, the student consensus is rather strong (and hilarious).

          Second, I don’t think most brilliant people do anything notable in their lives. Some do of course. However, I think they are the exceptions. Of course, depending on how you define ‘brilliant’ and ‘notable’ this answer might change. However, there are plenty of Math PhDs with IQs over 160 who won’t change the world (and Physics PhDs). That doesn’t mean they are not brilliant.

          Third, Brad has done a few things. He was a Deputy Assistant Secretary for Economic Policy, United States Department of the Treasury from 1993 to 1995. In that period, he was one of the principal architects of NAFTA. He still defends NAFTA by the way. He doesn’t appear to be familiar with the PISA data for Mexico. Of course, he may know the PISA data and may just be ignoring it.

        • peterschaeffer says:

          Just a few more notes. For fun, type “j bradford delete” into Google (with the double quotes). He (Brad) has quite a reputation for deleting anything that makes him look bad. Over at his blog, he claims that general intelligence could not yield even a difference of 0.1% in survival rates. He actually states this as a fact. I have no idea what the correct percentage might be… However, it might be a lot more than 0.1%.

  4. Shropshire says:

    The idea that high IQ people are more susceptible, on average, to nutty ideas and/or eccentric behavior was Charles Murray’s intuition. He writes somewhere that Herrnstein disabused him of this notion: He was simply more apt to notice the smart crazy people. If he could actually look at the admissions folders of Harvard undergrads, he’d see there wasn’t much of a relationship between greater intelligence and heightened eccentricity. It’s probably true that Ivy students profess to believe certain false ideas more than State U students, but they probably also have a better grasp of the social consequences of not professing those ideas. It’s hard to disentangle nuttiness from cynical preservation of power.

    Or maybe not. Lately, on Twitter, Murray has tweeted some stuff that indicates he’s back to his original intuition and agrees with Greg.

    • Dave Pinsen says:

      It’s a really interesting theory, but my anecdotal experience makes me wonder if mental illness is more common among those with less mental horsepower.

      I worked in a group home for retarded adults once, where one of my jobs was to give them their pills everyday. Maybe a third were on psychiatric drugs in addition to the usual stuff people their ages would be on.

      Similarly, the last time I was in my local emergency room, the black guy yelling about how he made more money than anyone in the hospital before begging for cab fare from the nurse’s station didn’t seem like he had a lot of mental horsepower to spare.

      The more parsimonious explanation for Reich’s fudging seems to be Sailer’s: that he wants to keep his funding.

      • gcochran9 says:

        Schiz is definitely more common at lower IQ, but bipolar seems ( I think) to be more common at high IQ. But I’m more talking about becoming an adherent of a crazy and destructive ism rather than being classically, individually psychotic. Listening to Stalin, or the New York Times, rather than the voices in your head.

        • Halvorson says:

          If you ask the geneticists, I think it’s only autism risk genes that also increase IQ:

          “Autism had positive genetic correlations with the same two Biobank variables: 0.34 for educational attainment and 0.19 for verbal-numerical reasoning. Bipolar disorder and schizophrenia showed a pattern of having a positive genetic correlation with educational attainment (0.22 and 0.13, respectively) and a negative genetic correlation with verbal-numerical reasoning (−0.11 and −0.30, respectively). Schizophrenia was genetically associated with slower reaction time (−0.24) and poorer memory (−0.34). Bipolar disorder was also genetically associated with poorer memory (−0.24). Major depressive disorder was genetically associated with slower reaction time (−0.25).”

          • mtkennedy21 says:

            Was the schizophrenia association measured before psychosis ? I doubt it. It could just be part of the syndrome.

            • Halvorson says:

              This is measuring the effect of schizophrenia/autism genes in unaffected people. Genuinely autistic people with epilepsy and mental retardation do not have higher IQs or educational attainment than the general population.

          • Interesting. Do you have a cite for that?

            Given that autism has a strong negative impact on reproductive fitness, I suspect something akin to heterozygote advantage is going on – having some of the ‘autism alleles’ increases fitness (perhaps by boosting IQ, but also by boosting learning ability, since autistic brains form connections more easily), but having most of them sends you over the autistic fitness cliff. The end result is that the alleles remain in the gene pool, but at a low level.

          • Both schizophrenia and bipolar disorder have typical onsets in late adolescence, when people are finished or nearly finished with their education.

        • Patrick L. Boyle says:

          As you no doubt know, African Americans have a schizophrenia incidence much higher than that of whites. Blacks, of course, also have IQs about one standard deviation lower. So when you see the relationship between schizophrenia and IQ, which way do you think the causality is running?

          Is the observed correlation because the population of dummies happens to contain a great many blacks who bring with them a lot of schizophrenia – or are the low IQs of blacks a result of whatever makes them have so much schizophrenia?

    • Linda Seebach says:

      Michael Shermer has said something similar, observing that Mensa audiences seem more likely to be gullible. He suggested that it’s because smart people are really good at thinking up plausible arguments for things they already believe.

  5. Eric Ruttencutter says:

    Did you live at Florida Avenue Residence Hall?

    Sent from my iPhone


  6. poster says:

    I wonder what the average height of the Dinka would be if they lived in a richer area.

  7. Regarding the comments about great Danes in the article. Here’s a thought experiment: does anyone know what might happen if you inseminated a chihuahua, or any other small breed, with semen from a great Dane? If the mother dog has no way to slow the growth of the embryos the puppies would be larger than she. I would bet that the mother has some fail-safe method to prevent too much vascular tissue being absorbed by the embryo and they would die of intra-uterine hypoxy?
    Or maybe the puppies would be stunted by some mechanism the mother has to prevent the embryo from sucking too much energy and nutrients form her. Similar to the way there’s a runt in a litter. But if the mother can in fact stunt the growth of her puppies it’s probably not adapted to such extreme biological differences in size as between a great Dane and Chihuahua.
    I’m surprised nobody has actually tried this. I could find no reference to any experiment like this. It’s relatively simple; And if the mother’s life is in danger from the growing embryos then you could just abort them and a dog doesn’t have to die from a medical experiment. So you probably don’t even need a license to try it?

    • gcochran9 says:

      A big sire and small mother is not rare, and it causes trouble.

      • It would be physically impossible in the case of a great Dane and small breed. I know somebody with a toy Yorkshire terrier and she was told large male dogs might kill her if she was outside and in heat.
        By a “big sire” and small mother I think that would be like a Labrador and Schnauzer or something which might not be mechanically impossible. I’m talking about a case where only artificial insemination would be possible. Would the pups be larger than she? I know that smaller breeds have smaller litters anyway. But I wonder if the pups did survive then would they grow normally once born? I guess they would be intermediate in size just as a Beagle and Labrador, but how could they possibly be born even if they are half the size of a normal great Dane puppy?
        I could find no examples the largest and smallest dog siring pups. But there’s a lot of instructions for dog breeders about how to facilitate natural mating when the female is much smaller. And artificial insemination is common in dog breeding anyway.

        • Whyvert says:

          Size is bound to cause complications. Some old farmer’s wisdom I have heard of is to keep a small bull on the farm for a heifer’s first pregnancy, in order to get a small calf and an easy first birth. Then for subsequent pregnancies buy semen from larger sires. (This may be obsolete now.)

          • But what kind of complications? I mean with artificial insemination you could theoretically have a chihuahua mother and great Dane father. With a small bull he’d be like half the size of the cow or something. With a Great Dane and Chihuaha his penis would be bigger than her whole body. And any puppies would be larger than she is after little gestation. The average size of a pups relative to the mother’s body weight has to be restricted somehow. And that’s why small dogs have smaller litter sizes even when mating with the same breed or a breeds of similar size.
            I guess the pups of a larger and smaller dog are almost always intermediate in size. The mother has to have some mechanism to avoid giving the pups more energy and nutrients than would be in her interest to give them. And she has to have some mechanism to avoid pups larger than she can physically give birth to.

            • A dog breeder told me that if you cross a male Great Dane or other large dog with a Yorkshire terrier or other small bitch, the bitch will die. A 6 foot Caucasian friend of mine married a 5 foot Japanese lady and after painful hours in labor the OB decided C-section.

              • So how does your dog breeder friend of yours know this? The Japanese lady is only 18% smaller than the man. The great Dane is 100X larger. The hobbits Homo floresiensis and the 6 foot Caucasian guy would be interesting as an experiment but she would only be half his size or so.

    • peterschaeffer says:

      What you are describing happens in humans with some frequency (more so of late). A large man (typically American/European) marries a small female (typically Asian) and she gets pregnant. Ob/Gyns know exactly what to look for and schedule a C-section to protect the mother. The term is CPD (cephalopelvic disproportion). Note that the prevalence of CPD is disputed. Not being a doctor I can’t comment on this.

  8. Lauris says:

    To be more precise, there are at least three problems related to fast single-trait selection:
    1. Physiological interactions. For example, if you are “creating” a 7 feet tall race by selecting only by height, other organs will not be optimally adapted to new body dimensions (as in example above)
    2. Up-selecting minor alleles. Even if many traits seem to have additive genetics, these are actually heterozygous variants that are beneficial. Minor allele homozygosity may be maladaptive – but because of the very low allele frequency there are almost none of these in population. As soon as you start up-selecting, there appear more and more homozygotes that may cancel out beneficial effects. The most famous example is sickle cell anemia, where heterozygotes have certain resistance to malaria, but homozygosity of minor allele is lethal.
    3. Pleiotropic genes. You cannot, in general, select for just one “pure” trait. Most genes influence many traits and thus fast-selected new breed will have many more differences from original than the original “selected” one.

    • gcochran9 says:

      For example, there has been very strong selection for milk production in dairy cows, that has resulted in a particular screwy variant becoming common. There’s a (now common ) 660kb deletion that A. significantly increases milk production and B is embryonically lethal in homozygotes. I know of a number of similar cases in stronglyselected domesticates.

      • Ursiform says:

        The health of lactating cows depends on people milking them. Turned out into the wild some changes would happen in a hurry.

    • Ursiform says:

      I think for the case Greg gave it would be more like recovering from a population bottleneck than out-competing a less selected group. But continuing outside intervention can overcome maladaptation–lots of domestic animals survive that way

  9. Christopher B says:

    Re high IQ and believing crazy stuff (not being crazy)

    I think there is pretty strong evidence that high IQ people profess adherence to ideas they don’t actually put into practice. If you take high SES as an indicator of IQ (reasonable given the link) you find those folks almost always follow the successful life script of obtaining an education, delaying parenthood until after marriage, are likely to have one parent devoted to raising the kid, and seldom divorce. No UMC family is proud of raising dropouts, unwed mothers, or baby daddies while they give lip service to the idea that these are acceptable life style choices. As one rather famous UMC father said, he didn’t want his daughter punished by being forced to be a single mother.

    • Roland H says:

      Never attribute to craziness that which is adequately explained by self-serving hypocrisy.

    • SMack says:

      Ashkenazi make an interesting example of high IQ bundled with eccentricity.

      If we define mental illness as “your mind becoming the enemy of your life”, there is a certain scorpion-and-the-frog pattern that’s hard to miss.

      Example 1: the only known way to make Ashkenazi poor is to let Ashkenazi design the economic system under which they live. Their own invention of communism managed to do something that eluded generations of jealous anti-semites.

      Example 2: What’s the definition of chutzpah? Killing your parents and demanding mercy on the grounds of being an orphan? Flooding Europe with Muslims and protesting a rise in anti-semitism? Promoting multiculturalism and lamenting the end of your materilneal group identity because all your sons are falling for asian girls.

      One of the smartest people I know is a South African of Lithuanian background, who fled his home in the early 2000s because the place was fast becoming no country for smart people. With a straight face he tells me “it’s really ironic, because my parents were in the forefront of the struggle against apartheid.”

      I said “yeah man, that’s some crazy inexplicable twist of fate there, one of those things nobody could see coming.”

      • biz says:

        “Flooding Europe with Muslims”

        Can you justify that charge? With some general analysis please, not one or two possibly fake quotes from one or two leftist Jews.

        • Frau Katze says:

          I would say that gentiles alone are quite sufficient for insane immigration policies. The “progressive” Christians are similar to Reform Jews. There’s few Jews where I live (west coast of Canada). Even if you allow for more of them in the East, they still don’t have the influence of American Jews. There aren’t many left in Europe (most of the ones in France came from Algeria after WW II).

          I present you with Justin Trudeau as proof of insane gentiles.

          • magusjanus says:

            maybe, but it’s certainly been the case that you know who have served as a catalyst at the very least.

            • Ivan says:

              Who was a catalyst in China’s embrace of communism ? Mongolians ?

              • magusjanus says:

                Pretty sure Soviet tanks and US foreign policy idiocy (courtesy of spies like Currie and naive progs like Marshall with that “save Mao” truce) fit the bill.

                That’s proximate cause. Ultimate cause you’re back to square one of course, with obvious high Jewish role in revolutionary movement in USSR (Solzhenitsyn’s 200 years) and yes, US.

                You don’t have to be a full blown alt-right but to just see that.

              • gcochran9 says:

                “naive progs like Marshall” You’re funny.

  10. JayMan says:

    I’m 6’4″ and have bad knees.

    • R. says:

      I’m same height but my knees are fine. So far. Though my patellas are hypermobile.

      No trouble with walking or running, but then, I’m stilly fairly youn and fairly light (about the same as the average US female) for my height.

    • Dave Pinsen says:

      Try the Starting Strength linear progression. Done correctly (with your hips rather than your knees taking most of the stress), squats will likely make your knees feel better.

    • Patrick L. Boyle says:

      I’m 6’4″ and have bad feet. Basketball instead of football. Or it could be that I wear a size 16 shoe.

  11. Bukephalos says:

    Ah the National Merits left on an island idea, again. Reminds me of Lee Kuan Yew’s Social Development Unit and how it failed to spur assortative mating between graduates and raise educated women’s fertility. It was later scrapped. Perhaps it could work with much stronger incentives. I can see it in a third world country where these youth are typically rare anyway, so it’s affordable: select and identify them rigorously, distribute generous welfare proportionate to the numbers of kids raised. This helps stave off brain drain, the likely prospect for them anyway. At the same time promote classic family planning for the less endowed masses, and a few generations down the line you improve the general cognitive profile, in essence doing the opposite of what the welfare state does in the West…

  12. Cpluskx says:

    If there was a 7 feet tall human not as a result of selection but hybrid vigor, would he still have bad knees?

    • MawBTS says:

      Yes. Height mostly causes problems because of non-genetic reasons (the square-cube law: an object’s load-bearing area increases as a square, while its mass increases as a cube.)

      You’d need special adaptations to offset this, and all hybrid vigor does is reduce the number of deleterious recessive alleles expressed. It doesn’t add any new genetic information.

    • RCB says:

      Is there hybrid vigor for height in humans?

  13. Citizen A says:

    BdL- no sense of ever being wrong, for it has fallen on him from on high- I was banned over a decade ago over pointing out the failure of rationality in economics. He was one of the ones tarring and feathering Larry Summers, so to expect any rational analysis is risible.

    And he continues to be a grand large example of genius thinking that it has answers to everything, even while failing at what he is supposed to be an expert.

    A smart fool is still a fool, full of folly and senseless noise.

    If you fool yourself first, then you can advocate your position unto the end of the world, because you can’t admit you fooled yourself.

    If the facts change, the analysis should follow the new facts, not the old consensus.

  14. Unladen Swallow says:

    Graves isn’t a prof at Harvard, he must have just been invited to lecture there, NC A&T and UNC-Greensboro. His “critique” of the Bell Curve is the last thing mentioned in Wikipedia entry on the book, he accuses the authors of “scientific racism”.

  15. Rich Rostrom says:

    ISTM that the rapidity of evolution is dependent on the intensity of selection pressure. When reproduction is consciously controlled by breeders, that’s about as intense as it can get, and there is experimental evidence of rapid evolution in response – Belyaev’s domestication of red foxes.

    Of course no degree of selection pressure can speed up the development of wholly new traits, as that is dependent on mutation.

  16. I didn’t know that my experience with DeLong was common. On my blog, I criticised a post of his ten years ago, linking to it. He came over and responded rather rudely and dismissively. One of his points wasn’t bad, but most of them were answerable. He never came back for that.

    As for rapid selection of intelligence, the standard explanation for Ashkenazi Jews is that their restriction to trade, moneylending, banking, and the like selected for being able to do math and complicated planning in one’s head. As local trade, and even ocean-going world trade increased gradually in Europe, a similar though less intense pressure was likely exercised around who could keep such matters in their heads – or do the arithmetic at all. Those would support more descendants. Intelligence is likely selected for in all societies, but when there are fearsome diseases, and not much surplus to trade with anyone, other traits would be selected for more strongly.

    That much trade is not many centuries old, and seems to have elevated NW Europe a few IQ points in a short time. NE Asian IQ, may be similarly derived – I’m just guessing here. A thousand years is plenty for 3, 5, maybe even 10 points.

  17. blueiris says:

    This article is a reasonable interpretation of the data it presents. But, as the saying goes, there are three types of lies: lies, damn lies, and statistics. Here, we see straight-forward biology being used with too little context, resulting in flawed conclusions.

    The author correctly points out that there is regional genetic variation. They fail to mention the often unusual and inconsistent nature of that variation. For example, the shared regional traits between Europe, India, and Latin America in accordance with the path of empire. Depending on when a mutation occurred, it may or may not have spread between multiple populations. But, the core point is true: there are some mutations that have only spread in certain regions. This is as far as you can take this argument while remaining true to the science. Here’s brief explanation of why that is true:

    1) You should always be suspicious of people who attempt compare physical attributes with mental ones. We have identified 16 genes that control eye color, and perhaps as many as 700 may be responsible for height. There are over 6,500 genes that are primarily active in the brain. It’s nearly 1/3 of our functional genes! Because the brain is so complex, any given mutation is far less likely to have a noticeable, much less significant, effect. The brain simply doesn’t evolve that fast. Evolutionary biologists are often looking at changes across much larger timeframes, typically 500,000-1M years. Over the last 10,000 years, incidentally, human brains have shrunk– a downside to developing agriculture.

    Humans, at the point of advanced tool usage, began skirting natural selection in a variety of ways. People often wrongly believe that natural selection would have simply began selecting for intelligence, but the aforementioned tiny changes would not have been enough to proliferate in a population, nor would it have outweighed the benefits of things like civilization in protecting those who would have otherwise died before reproducing.

    Civilization, essentially, made natural selection a less powerful force.

    2) Mutations in the brain are more likely to be fatal. A gene in the brain that does something wrong is far less likely to be passed down. Even today, this is starkly true among people with triploidy or trisomy. Negative mutations of the body, such as a differently shaped hand or blindness are less likely to impact someone’s survival and reproductive possibilities. Again, this is particularly true in settled societies. Most mutations are not good– they either have no effect or do something negative. It’s hard to improve on a species that is working in the moment! Even if brains were not much more complicated (more genes), brains will still evolve slower because of the high risk mutations pose.

    TL;DR– Yes, there is geographic genetic variance. But they are far more cosmetic than we believe (and they surely do not meet the biological definition for races).

    Finally, you shouldn’t trust anyone who tells you that genes and IQ are reliably linked. Just stop and ask yourself for a moment, “What does it mean to be intelligence?” and then “Has intelligence been used to select mating prospects in our society?” or “Is our species selectively only allowing the most intelligent to breed?” We haven’t. And frankly, why should we? Intelligence, as I’m sure you know, is no promise of happiness.

    • gcochran9 says:
      1. You’re wrong. It’s perfectly easy to select rapidly on a trait influenced by many genes. Check out the breeder’s equation.

      2. Civilization radically changes the experienced environment. A big change in environment results in strong selective pressures – the population isn’t yet very fitted to that enw environment. Strong selective pressures speed up evolutionary change.

      3. Intelligence is highly heritable, almost as much as height.

      4. Intelligence is not a sure route to happiness – but it sure does help you win arguments.

      • blueiris says:

        The breeder’s equation? Seriously? You need to stick to anthropology and stop trying to be a geneticist. Please tell me what you’d plug into the breeder’s equation to reach that conclusion for any of the areas you’ve spoken about.
        That is nonsense– your making the assumption that fitness is the same standard inside and outside of civilization. Civilization changed the environment, yes, but mostly in ways that protected and insulated humans from our ecologies. The standard pressures were smaller, not larger, which resulted in the last 10,000 years of brain shrinkage. Civilization does not apply a selective pressure for intelligence. There’s zero data to back up the idea it does, and a wealth that suggests the opposite.
        Intelligence– what do you even mean by that? IQ? Musical talent? Parenting ability? Creativity? Intelligence is a variable concept with no clear way to measure it. IQ, as you must know, is not reliable and it’s not accurate to place genes in any sort of causal relationship with IQ.
        Nice dig.

        Now, there are several arguments of mine you left unanswered:

        Mutations in the brain are more likely to be fatal and/or still caught up in selective pressures, meaning they are less likely to spread through a population.
        The number of genes involved in intelligence is higher than most physical attributes, making even strong selective pressure take such a long time.

        I’ll give you a challenge: tell me what regionally variable gene links to a given definition of intelligence, and at least you’d have a leg to stand on.

        • Pincher Martin says:

          Intelligence is a variable concept with no clear way to measure it. IQ, as you must know, is not reliable and it’s not accurate to place genes in any sort of causal relationship with IQ.

          You’re an idiot. IQ is highly “reliable” – which can be defined as IQ tests having stable and replicable results linked to important independent variables everyone associates with being smart.

          And it’s completely uncontroversial to anyone who knows anything about IQ that different population groups (i.e., races) test at different levels. The only controversy is over why. And the only reason that is controversial is because a large segment of the intelligent populations averts their eyes from hard truths. (I do not include you in that group. Sincere idiots are in their own class.)

          If you don’t understand this topic, don’t post about it.

        • gcochran9 says:

          None of what you’re saying is correct.

          You say that highly polygenic traits respond more slowly to selection – but that’s not true. Go read up on polygenic selection.

          Civilization put people in new conditions that they weren’t closely adapted to, so it created new and stronger selective pressures. You can’t see that – but I can. For example, originally people weren’t able to fully use the extra food value in milk, because they couldn’t digest lactose as adults. New situation – genetic changes let people take advantage of it. Booze was available to farmers – genetic changes let people drink without as much risk of alcoholism. The fraction of carbohydrates in the diet tripled: genetic changes eventually reduced some of the unpleasant effects of that change. Wheat was deficient in ergothioniene: a more-active ergothioniene transporter became common in some farming populations. There was more hierarchy than there had been among hunter-gatherers – I guarantee that was a genetic response. Infectious disease became a greater threat – people changed to become more resistant. While isolated populations did not.

          I know medical genetics fairly well, at the level of Vogel and Motulsky plus a few hundred newer papers on more recently discovered Mendelian things in ASHG. As far as classical genetics & pop genetics, at the level of Crow and Kimura, Fisher, Charlesworth and Charlesworth. Evolutionary genetics at the level of Maynard Smith and Hamilton, with again a few hundred newer papers and books. I looked at the big Neanderthal paper (Green 2010, ~ 150 pages into the online supplement) and noticed that the San-Papuan genetic distance was anomalously high, suggesting that the Papuans had experienced yet another dose of archaic admixture, presumably Denisovan. It means that John Hawks and I managed to successfully predict that A. there had been Neanderthal admixture and B. some of it was adaptive and common today. I successfully predicted that the Tibetans had picked up useful archaic variants that had helped confer a superior kind of altitude adaptation.

          Quantitative genetics at the level of Falconer.

          So, where were your drops?

        • RCB says:

          “The number of genes involved in intelligence is higher than most physical attributes, making even strong selective pressure take such a long time.”
          Again, if you understood polygenic selection, you’d understand that very small changes in very many loci can equate to large changes (say, 1 SD) in phenotype. No single gene has to evolve much if you have 100s of genes contributing to evolution.

          Or as Charlesworth and Charlesworth say in their chapter on quantitative traits in the Elements of Evolutionary Genetics, “a substantial change in trait mean over the course of a selection experiment that lasts only a few generations can be caused simply by shifts in variant frequencies, not fixations.”

          Turns out that it does pretty much all boil down to the breeder’s equation.

        • You seem to think no one here has ever heard these arguments before and examined them. I think that shoe is on the other foot.

        • teageegeepea says:

          blueiris, Greg co-wrote a book titled “The 10,000 Year Explosion” on selection after agriculture. You should read it.

      • Fifty years ago (i.e., before genetic engineering) I asked a Professor of Agriculture how long it would take to recreate modern maize from its wild cousin, teosinte. He answered: about 100 years.

    • Frank Remele says:

      Steve Hsu said that the number of alleles that could potentially influence intelligence are far greater than the amount that explain the variation in IQ differences we see today. This being the case, your argument isn’t nearly as powerful. We wouldn’t necessarily need thousands.

    • RCB says:

      For traits influenced by many loci, each locus has a small effect… but then there are many more loci on which natural selection can act. The effects cancel out; breeder’s equation holds. Guess what: the mathematics of this has been understood for about 100 years. Go look up the infinitesimal model!

      “Civilization, essentially, made natural selection a less powerful force.”
      People need to stop saying this. Wherever there is heritable variation in reproductive success, there is evolution by natural selection.

    • Deckin says:

      “Just stop and ask yourself for a moment, “What does it mean to be intelligence [sic] ?”” OK, I stopped and still can’t imagine it.

    • Harold says:

      “Because the brain is so complex, any given mutation is far less likely to have a noticeable, much less significant, effect.”

      “Mutations in the brain are more likely to be fatal.”

      Seems pretty significant.

    • Yudi says:

      It occurs to me that people like blueiris (and Brad Delong, whose writings I used to enjoy) think that natural selection increasing the frequency of a single mutation is like a man laboring to pick up a very heavy rock within a short space of time: difficult, but possible.

      But for quantitative traits, now the man has to pick up 2,000 heavy rocks within the same length of time! There’s no way that’s possible! Whereas it’s really like natural selection works like a force field that can levitate most of the rocks with equal efficiency and speed.

  18. luisman says:

    Reblogged this on Nicht-Linke Blogs and commented:

    they’re not terrible well adapted to their new mental horsepower. Susceptible to various mental problems and illusions that would probably be a lot rarer if natural selection had had time to iron out the bugs.

    • mtkennedy21 says:

      There is an interesting experiment going on right now. High IQ couples from “assortive mating” are having babies and then assigning their care to low IQ immigrant nannies. We should learn whether nature or nurture is more important in intelligence as opposed to raw IQ.

      • I think a number of British writers have reflected on how much they loved their dear old nannies even though they eventually decided they weren’t very smart. And those nannies were likely closer in intelligence to master and mistress than the modern immigrant is to the modern equivalents.

  19. Thank you for this post. Lucy, my Great Dane, died of cancer perhaps four years ago.
    A question has been nagging me lately. Large mammals (elephants, whales) outlive small mammals (mice, shrews), yet when humans breed dogs for size, they get short-lived dogs. I suggest that human-directed breeding for size means, in effect, breeding for rapid growth and therefore for cancer susceptibility. Related (?): malignant neoplasms are among leading causes of death in very young children.
    My question: how does adding desired traits to the target (larger size, greater longevity, greater speed, agility, tractability, etc) affect the time (in generations) to the target?

  20. Flemur says:

    I saw “Judge Starling”/Dan Graur mentioned at Coyne’s blog.

    Reich op-ed: “The ancestors of …[almost everybody]… almost completely isolated from one another for 40,000 years or longer, which is more than sufficient time for the forces of evolution to work.”

    “What a load of nonsense! For selection to operate and counteract the effects of random genetic drift, the effective population size should be large. Unfortunately, the long-term effective population size for all the humans in the world is barely 10,000—lower than that of chimpanzee. By necessity, the effective population size of each “race” separately is much smaller. So, the chances that 74** loci will experience significant changes in allele frequencies simultaneously in each of the four populations is zero.”

    ** 74 genetic variations that are over-represented in genes known to be important in neurological development.

    Any comment? Especially re. “long-term effective population size” ?

    • RCB says:

      It is true that heritability within populations says nothing about genetic differences between them. The demonstration that IQ is highly heritable is not in itself evidence that differences between races have a genetic component. To demonstrate that, you’d want good adoption studies. This has been widely understood for a long time.

      “For selection to operate and counteract the effects of random genetic drift, the effective population size should be large.”

      It is true that larger populations are better if you want to select for something. Both because it reduces drift and (as Cochran has argued repeatedly) because it makes new mutations more likely.

      Of course, you know any argument that claims “there hasn’t been enough time to select for racial differences” must be nonsense, because there are obvious phenotypic differences between the races due to natural selection. That’s why we are able to identify them in the first place. Skin color, body size and shape being the obvious cases, but Cochran has talked about many others.

      “Unfortunately, the long-term effective population size for all the humans in the world is barely 10,000—lower than that of chimpanzee.”

      Again, this doesn’t matter because we already know the conclusion is false. But, for educational purposes: what the author doesn’t know is that there are (at least) two measures of effective population size (EPS): “Inbreeding EPS” and “Variance EPS”.

      When the author says “EPS = 10,000”, (s)he is referring to Inbreeding EPS. This is the equilibrium theoretical population size that yields the same amount of genetic variation as the actual population in question. This depends a lot on the population’s recent history. It’s small for humans because we weren’t very numerous until recently. It’s not relevant if you’re trying to understand drift vs. selection.

      Variance EPS tells you the theoretical population size that yields the same noise in allele frequency change, right now. It basically measures how little a population is susceptible to drift. It depends little (not at all?) on history. It’s the relevant measure if you want to talk about barriers to selection. Humans actually have a very large Variance EPS (large populations, relatively low reproductive skew, long lifespans), and have for quite some time. It’s been well over 10,000 on every continent for millennia, no doubt.

      “To imagine that 74 alleles change frequency in concert and that the 74 alleles have successfully battled genetic drift and recombination in merely 2,000 generations requires an extremely naïve and unsophisticated view of the evolutionary process.”

      “merely” 2000 generations? 1/2000 of a stdev change in IQ per generation would be enough to explain observed differences between groups today. Given available additive genetic variance in IQ today (and probably in the past), that’s easy.

      (Also, selection doesn’t have to “battle” recombination because recombination doesn’t affect allele frequencies. And this person claims to be some expert in pop gen…)

      • gcochran9 says:

        I’m going to steal a little from you when I post about Graur’s comment. Ok? Bye the way, you should be scared: “Graur served as associate editor of the journal Molecular Biology and Evolution from 1995 to 2011. Since 2009, he serves as associate editor of the journal Genome Biology and Evolution. From 2009 to 2011 he has held the position of Councillor for the Society for Molecular Biology and Evolution.”

        • RCB says:

          No problem. I left a similar comment on his Tumblr blog, at the source. Maybe you wait until he responds there before you post…

          More useful things mentioned:
          -There’s this impression that people get that if evolving one gene is “hard”, then evolving 100 genes in the same direction is really hard. Evidently this Graur guy believes that (“To imagine that 74 alleles change frequency in concert and that the 74 alleles have successfully battled genetic drift and recombination in merely 2,000 generations requires an extremely naïve and unsophisticated view of the evolutionary process.”) But that’s just wrong. If 1000 alleles additively contribute to a trait and that trait is under positive selection, all 1000 alleles will experience a small amount of positive selection. Drift will by chance cause many to move the opposite direction, but on the whole a large number will increase a bit. And a little bit of average movement among 1000 alleles can cause a big movement in phenotype.
          -Recombination actually helps natural selection because it allows different loci to evolve independently. Provides an escape from Muller’s Ratchet.

          As for his credentials…. when you’re wrong, you’re wrong.

      • gcochran9 says:

        A while back, one of my kids did better on a particular standardized test than any adopted black kid in the US. Guess how I know this odd fact.

        • magusjanus says:

          she aced it. and no black kid did (cuz if they had we’d have heard about it in every major media ever).

        • RCB says:

          The Minnesota trans-racial adoption seems like the best available common garden experiment:
          Take a bunch of white adoptees and raise them among white parents with an average IQ of 115. Now do the same with a bunch of black adoptees. Both adoptee groups are unrelated to the parents, and went through the adoption process at a young age (in case that environmental effect is important). What’s the difference between the mean IQ’s of these two groups at age 17? 106-89 = 17. About an SD.

          Of course, it’s possible that black adoptees consistently experience a different early-life environment than white adoptees. Or that black kids raised in white families don’t actually experience the same total environment as white kids raised in white families – and that this has a 17 point effect. Differential attrition doesn’t seem to explain it, because change in IQ of the cohorts over time didn’t differ significantly between races (they all dropped, as we expect from the finding that heritability increases with age).

          • gcochran9 says:

            I know about that study.
            Iq-like differences show up pretty early, well before school. I say IQ-like because they’re looking at vocabulary differences at age 3, not an actual IQ test. At age three, black-white vocab difference is about a standard deviation.

      • catte says:

        Variance EPS tells you the theoretical population size that yields the same noise in allele frequency change, right now. It basically measures how little a population is susceptible to drift. It depends little (not at all?) on history. It’s the relevant measure if you want to talk about barriers to selection. Humans actually have a very large Variance EPS (large populations, relatively low reproductive skew, long lifespans), and have for quite some time. It’s been well over 10,000 on every continent for millennia, no doubt.

        Variance EPS is affected by population history too — in the idealized case it’s the harmonic mean of generation sizes, which means bottlenecks are important. But yes — our long lifespans matter too, and that brings it up.

        • RCB says:

          If Variance EPS is the size of an idealized Wright-Fisher population that yields the same amount of variance in allele frequency change as the reference population across a generation right now, then how can population history matter?

          I suppose one could say that if the population consistently goes through a population cycle, or has some stationary probability distribution of population fluctuations through time. Then at this “equilibrium”, yes, you’d use the harmonic mean of the cycle/distribution to get the Variance EPS that describes the whole cycle. But in any given generation, you could also say that the Variance EPS is the population size at that moment – because that’s what determines drift at that moment.

          Take the human case. Right now, we are extremely un-drifty because of our enormous population sizes. This fact would be true even if our actual population size were 10,000 only 100 generations ago. Our Variance EPS right now is huge. It’s not anywhere near 10,000, and hasn’t been for a long time.

          • RCB says:

            To put it another way: if we’re talking about the effect of drift during the few thousand generations that human races have been separated, then we only care about human population history during that time, and not before. Any bottlenecks that occurred prior to the split would have affected (and still do affect) the Inbreeding EPS, but they would have no effect on how drifty the populations would be.

        • gcochran9 says:

          Inbreeding EPS is the harmonic mean, Variance EPS is not.

    • catte says:

      I suppose this is why it’s impossible to create different breeds of dog. The effective population size is just too small.

    • catte says:

      Looks like this Starling can’t hear the Falconer.

      • Yeah. The way extremely educated people are, er, discussing these issues makes me think of “lies, damned lies, and statistics”, but it’s also interesting.

        Apparently “Starling” is a(n) SJW, and posted some anecdotes which I don’t believe happened:
        “judgestarling. tumblr. com/ page/ 7” (remove the spaces).

  21. Flemur says:

    Ouch! I forgot what wordpress does with links – sorry!

  22. simontmn says:

    I see what you did there.

  23. Ivan says:

    As I recall from reading Falconer a long time ago, the effective population size should be about a hundred or so for selection to overcome random drift. His formula is s = 1/(4*Ne), so the tiny coefficient of selection > 0.0025 is sufficient to compensate drift.

  24. Brian says:

    “if natural selection had had time to iron out the bugs” …I wonder what such human beings would be like? What if they already are present, as out of place now as the first beings like us were circa 50k ya. Wish I could write… there’s a story somewhere in there.

  25. Philip Neal says:

    The Guardian has issued its View on recent developements.

    “big, creepy… absurd… The Bell Curve… a tall claim… problematic on many levels… sophisticated statistics… an intimidatory atmosphere… troubling… pernicious and incendiary… tendentious… it is wrong to think… power requires a sense of responsibility… we must not elevate discrimination to a science… need a more sceptical eye on the science… we all have a duty…”

    A new fallacy is needed, and soon. No doubt one is being cooked up.

  26. sprfls says:

    Ha! I used NBA centers as an example of genetic tail effects as part of a presentation at a small conference in 2015.

    According to my rough back-of-the-napkin calculations, players from the Dinaric Alps comprise ~7% of NBA 7+ footers, making them ~54x overrepresented relative to the global population (and that’s being very generous with Dinaric Alp population size estimates).

  27. sprfls says:

    If I kidnapped a year’s worth of National Merit Scholars and dropped them on a deserted but fertile island, a new race with an average IQ around 130 would develop ( unless those little brainiacs escaped. You have to watch them all the time).

    Was Huxley very wrong re: The Cyprus Experiment in Brave New World? I think yes, he was, but I can see an argument for the other side as well.

  28. Pingback: Whence Polarization? | POLITICS & PROSPERITY

  29. Pingback: The Balderdash Chronicles | POLITICS & PROSPERITY

Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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