What “heritability” actually means

I often get brought up short when someone tells me that intelligence, for instance, is 80% determined by your genetics. I’m held up because that someone clearly doesn’t understand the magic word “heritability”, which every geneticist understands but every layman seems to consciously misunderstand, preferring to play games with folk etymology than actually understand the math, or the concept. So here’s a nice video that explains the background clearly, and as a bonus, show that Sam Harris doesn’t know diddly squat about the science behind the racism he endorses.


  1. raven says

    Sam Harris is always an idiot.

    American conservative Ron UNZ 2012
    “…. attacking my Race/IQ analysis is the overwhelming focus of these critics upon my Irish data. Although I discuss similar ethnic IQ evidence regarding the Greeks, Balkan Slavs, Southern Italians, Dutch, Germans, and various other European peoples, attacking my Race/IQ analysis is the overwhelming focus of these critics upon my Irish data.

    At various times, many European groups have been proven to be low IQ by testing. In fact, it is most of them, Greeks, Slavs, Italians, Dutch, Germans, and especially the Irish. The Irish for nearly a century in dozens of studies have scored low in IQ tests and have been considered to be sort of dumb.

    Except modern Irish and American Irish score average on IQ tests.
    What is going on here is obvious.
    There are a vast number of variables that effect average IQ.
    Socioeconomic status, childhood nutrition, rural living, education, culture, racism, and so on.

    The heritability of IQ only has meaning when all other variables have been controlled for.
    This is difficult to do and rarely has been done.

  2. Jazzlet says

    Good find PZ, from this one and the other two videos he’s done so far it looks as if this will be a useful resource.

  3. PaulBC says

    Good video. I made an exception to my normal rule of not watching video content, but I still would have rather seen it written.

    My actual feeling is that a lot of this is besides the point. I don’t think we all have it in us to be, say, Richard Feynman. I sure don’t. But within the wide latitude of normal intelligence, it is possible to stay informed and reach rational decisions. The inability (sadly common) is due to a lack of education and easily fixable through education.

    The arguments for innate intelligence are almost always used to deprived the underprivileged (whether by class or race) of adequate educational resources and keep them stupid. It is not a waste of time to educate anyone. Whatever their potential, they’ll still be better off with it than without it. And I don’t mean “elite” education either. That some tiny fraction can make it into Harvard or whatever is also missing the point. Whatever kind of “bell curve” you’re talking about, I think we should be putting a lot more focus on the middle and not the upper tail.

  4. says

    Given the huge number of environmental influences on IQ test scores, which must be controlled for, wouldn’t it be most rational to assume that any deviation from average is a result of some as-yet un-controlled-for environmental factor, rather than a genetic difference?

    Hint: one would expect education (in countries with socialized education) to be heritable, as well as the effects of wealth which is also heritable. There are all kinds of things we get from our parents that are not in our DNA.

  5. says

    As far as I can tell, scoring high on an IQ test means that you’re good at taking IQ tests. In other words the test measures your ability to take that particular test.

  6. pilgham says

    There does seem to be more correlation between how much money you’ll make and your IQ score. Can we just dispose of the whole “intelligence” thing and call it MQ? Although you’d have to claim it was logrithmic to explain current salaries.

  7. Michael says

    With regards to nature vs nurture, I once heard it described as nature sets the limits on what nurture could do. So the limits on your height are set by your genes, and while they can be affected by your diet, posture, lifestyle, etc., if you had a pair of identical twins, one of which grew up to be 6′, unless you made drastic and unethical changes to the others environment, it isn’t likely the other would grow to be less than 4′ or more than 7′. On the other hand, while eye and hair colour are heritable and variable in a population, I’m not aware of any normal changes to an environment that would affect those, ie. two identical twins reared apart will still have similar hair and eye colours.

    Considering the high proportion of Jewish Nobel Prize winners, perhaps we should focus on identifying the factors that have influenced that, and/or model the education system on whatever they are doing right.

  8. says

    I’ll probably get banned for this, but that podcast makes a very strong case that there is cognitive stratification going on. My favorite quote from Sam Harris @1:40:

    “… I think the last person I knew on Earth who smoked cigarettes, was Christopher Hitchens .”

  9. PaulBC says

    The video mentions wearing glasses. It wasn’t the point, but let’s assume good vision is highly heritable. I think it is. On the other hand, it’s basically irrelevant now, because if you are myopic, that can be fixed. How many cognitive traits are like that? I’m not sure everyone has the memory to be a bard passing down epics through oral tradition (though I’m not sure this memory feat is as difficult as it sounds). I do know that we don’t need it as much now, because most people can be taught to read and write. Why don’t we work with what most people have already instead of developing filters for cognitive traits that may be useful under some circumstances but can be replaced by simple devices and environmental changes?

  10. lucifersbike says

    Jonathan Norburg, pilgham. My family all score well above average on IQ tests (up to 158 in at least one case). Our suspicion that IQ tests just measure how good you are at doing IQ tests is partly supported by the fact that most of us are stony broke.

  11. KG says

    I’ll probably get banned for this – canman@9

    There are few things as infuriating, or as stupid, as this kind of pre-emptive whining about persecution. If ithis were my blog, you’d get banned for that alone.

  12. chrislawson says

    The original IQ test was designed by Binet to identify struggling students so that they could be given extra help in their education. That progressive, remedial view died in translation to the US where educational psychologists adapted it to the Stanford-Binet test as a tool for “curtailing the reproduction of feeble-mindedness and in the elimination of an enormous amount of crime, pauperism, and industrial inefficiency”. How poisonous of Harris, Murray and Hernstein to recycle that hateful philosophy.

  13. A. Noyd says

    PaulBC (#10)

    On the other hand, it’s basically irrelevant now, because if you are myopic, that can be fixed. How many cognitive traits are like that? I’m not sure everyone has the memory to be a bard passing down epics through oral tradition

    Kids come with all sorts of baggage that affects how they learn stuff, and teachers come with baggage that affects how they teach stuff. A lot of that baggage stems from cultural beliefs about what certain groups of people can or can’t do or what talents good learners exhibit.

    In turn, pedagogy itself has baggage. It can get stuck in ideas that particular teaching methods are best for all students. It can design tests that value arbitrary factors that favor one type of learner over another. It can make invalid associations between individual characteristics and individual potential.

    So you can end up with kids who are perfectly capable of learning things but who never try because they believe they’re incapable. Or kids who are written off as unteachable because they’re never taught in a way that makes sense to them.

    How do you tell apart the person who is truly cognitively incapable of memorizing oral epics from the person who is cognitively capable but who has trouble with the single method by which they’re taught? How we learn and how we teach are extremely sensitive to outside influence and thus also to correction.

  14. chrislawson says

    The better an education system gets, the more heritable a score like IQ becomes. In a perfect education system, each student will receive the maximal environmental assistance which means the only variables left in play will be genetic, so heritability will be 100%. Which of course says nothing about the underlying genetics of intelligence other than that it has some non-zero influence.

    As for regression to the mean: if IQ really was genetic, we would expect to see a high-IQ person’s children move back towards the mean. Conversely we would expect to see low-IQ people have children “smarter” than them on average. Because of the nature of regression to the mean, this applies for environmental as well as genetic factors. Which means that any multi-generation persistence of a trait like IQ is overwhelming evidence not of environmental or genetic influences but a severe lack of sorting. That is, it’s all about class rigidity, not innate intelligence.

  15. unclefrogy says

    very nice he reminds me a little of “three arrows” and that is good no such thing as too many clear rational YouTube Channels
    @17 what does that mean in this context?
    uncle frogy

  16. Hj Hornbeck says

    Work can sod off for a bit. I hope.

    Michael @8:

    Considering the high proportion of Jewish Nobel Prize winners, perhaps we should focus on identifying the factors that have influenced that, and/or model the education system on whatever they are doing right.

    Eh, I don’t think there’s much mystery there. This theory sounds pretty solid:

    The real explanation of Jewish success in science lies elsewhere. The 20th century began with massive migrations of Jews, to the United States, to the cities of Russia (and then the Soviet Union), and to Palestine. In each of these new lands, Jews turned to science in great numbers because it promised a way to transcend the old world orders that had for so long excluded most Jews from power, wealth and society. Science, based as it is on values of universality, impartiality and meritocracy, appealed powerfully for Jews seeking to succeed in their new homes. It is not so much what Jews were (smart, bookish) that explains their success in science, as what we wanted to be (equal, accepted, esteemed), and in what sorts of places we wanted to live (liberal and meritocratic societies).

    There’s likely selection effects at play, too. Any Jewish person that can survive in an environment of constant discrimination has to be a sharp cookie, whereas gentiles can float along on their privilege. The net effect is that the average Jewish academic can be smarter than the average gentile, even if there is no biological difference in intelligence.

  17. Hj Hornbeck says

    Also Michael, whether you know it or not you’re playing into a racist dog-whistle. It’s a common tactic for race realists to switch from talking about why some races are inferior to why some are superior.

    Recently, however, several researchers have argued that Ashkenazi Jewish intelligence is not entirely socioculturally caused, but rather is the result of unique selection pressures faced by the ancestors of modern Ashkenazi Jews (…).

    Cochran et al. (2006), for example, contended that because Ashkenazi Jewish people between 800 and 1700 were largely endogamous and were forced into a few, cognitively challenging occupations (e.g., accounting, money lending, and management), there were strong selective pressures on Ashkenazi Jewish intelligence. That is, because Ashkenazi Jews who were highly intelligent flourished in their limited occupational niche, they reproduced more prolifically than less intelligent Ashkenazi Jews, which ultimately led to a population level increase in intelligence. Lynn (2011) and Glad (2011) have also forwarded evolutionary accounts of Ashkenazi Jewish intelligence. […]

    This is a dog-whistle on two fronts. Intelligence is a complex emergent phenomenon dependent on five hundred genes at last count. What group is more likely to stumble on a set of mutations that lead to a significant improvement in average intelligence: a small isolated group, or a large one with plenty of interbreeding? Race realists are essentially arguing that your best chance to win the lottery is buy as few tickets as possible. So either these scientists don’t know the first thing about the science they claim to have a keen interest in, or they’re invoking the language of science in order to disguise their racism.

    Secondly, “Ashkenaz” is Herbrew for Germany. In other words, “Ashkenazi Jews” referrs to a group that is predominantly white and European. Jews with Middle-East or Asian ancestry apparently do not display the same level of intelligence, despite sharing religious practices that are almost identical. Talking about the intelligence of Ashkenazi Jews is thus a coded way to argue for white superiority, one that flies under most people’s radar.

  18. mvdwege says

    Also disregarded is an enormous cultural factor in the migration of Jews to jobs with high intelligence requirements: a culture that encourages young boys to study religious law in an ancient tongue and expects them to at least somewhat competent in understanding it at 13 is a culture that highly values reading complex texts.

    Not surprising that all those Jewish Nobel winners are men, then. With the growing amount of girls doing bat mitzvah in the same way as the boys do bar mitzvah, one would expect this difference to slowly equalise away.

  19. yaque says

    Hj Hornbeck:
    “Race “realists” are essentially arguing that your best chance to win the lottery is buy as few tickets as possible”
    Good point.
    Generally speaking, if your nation wants Olympic/Nobel prize/etc. level people in some pursuit, you must encourage it on a wide scale. Sorting from a population as large as possible. And lots of practice. Russians have no particular “genetic” talent at chess, the Soviet government, for reasons of prestige and culture, just widely established chess clubs at the elementary school level and up. It’s a pyramid. You want a higher peak, widen the base.
    Let everybody participate. Encourage those who do well, whatever their origin.

    Whatever you do, don’t limit it to the few with money or leisure! Rich people don’t have any special, inherent “talent” at anything. Today’s “meritocracy” is bunk. Legacy admissions to university? Please.
    It works for most things, tennis, basketball, swimming, scolastics, athletics, ballet. To win the lottery, buy lots of tickets!
    Ashkenazi culture encouraged book study for most everybody as much as possible for generations, so no wonder. No need for “genetics”, just practice and sorting.

    The idea that Ashenazi Jews are somehow “genetically” White Europeans is a rather weird and recent antisemetic trope. The “White” “race” is, of course a completely bogus category, and changed over generations according to the needs of racists. In the US, Irish and Italians once weren’t White. Now they are.
    Latinos? Somebody explain to me when they’re White and when they’re not. I’ve given up.

    Of course Jews were always regarded until the last few decades as always non-White. Today they are a kind of Schrodinger’s White people, depending on the needs of the antisemite, In Europe or the US, not White!, foreigners! invaders! in the Middle East, White! invaders! foreigners! Khazars!. All bogus.

    Genetically, Ashkenazis are more related to other Jews than to anybody else, and Jews, including Ashkenazis, are generally related to Levant populations:

  20. chrislawson says


    Adding to your argument: charging interest on loans was considered a mortal sin in mediaeval Europe (and other places, but let’s look there as the Ashkenazi have been invoked). If it’s a crime to bill for interest, then nobody will want to lend money. Thus non-Christians became necessary to a functioning financial system. And Jews were a very convenient non-Christian population. Nothing to do with genetics. It was 100% social norms.

    (And let’s not forget, most Jews were not bankers. Many were cobblers, tailors, importers, etc. And at the same time Jews were giving loans to Christians, they were often persecuted and expelled or murdered. It seems odd to claim that there was evolutionary pressure on Jews to be smart bankers when clearly the most intelligent decision from a purely reproductive success view would be to convert to Christianity…which would then make it impossible to continue banking.)

  21. DanDare says

    I’m reminded of the golden arch concept for selective education. Imagine the school is a gold arch and there is a queue of potential students. Only the best are selected to pass through the arch. Then the students comming out of the arch are tested and found to be all best students. The claim is then made that the arch produces these students, in some way making them so, rather than the test for entry exactly matching the test for best in the output.

  22. Matthew Herron says

    The arguments against Harris’ naive interpretation of heritability are on the money. One minor quibble, though: in the island example (starting around 5:05), heritability would not be 1.0. That is itself a misunderstanding of how heritability works. In an obligately sexual species (such as humans), the heritability we are concerned with is narrow-sense heritability, h^2, which is additive genetic variation divided by total phenotypic variation, Var(A)/Var(P). Since Var(P) includes things like non-additive genetic variation, maternal effects, and stochastic variation, Var(P) is always greater than Var(A), even when environmental variation is zero.
    Side note: who’s the nitwit who thinks evolution denial is primarily on the political left?

  23. mountainbob says

    I remember a point made by a prof. at the UofM way back when. She said it was likely, in a state of nature, that humans who live to adulthood were above average in intelligence, despite the fact they wouldn’t know how to deal with a SAT. Those who weren’t so swift would likely succumb. Doesn’t say anything about modern life, so don’t jump on me, it’s just a speculation about a set of conditions that are very rare in the world these days… and which would become extinct soon after contact with “advanced” society.

  24. John Morales says

    mountainbob, you’re doing that thing where you focus on a single factor

    (e.g. no less true that the humans with a better immune system would live to adulthood, or, more abstractly, those who were more liked)