African variation

Generally, sub-Saharan Africans have low scores on IQ tests, educational assessment tests like PISA, etc.

Does that mean that all Africans are the same? No. Could some population in Africa do considerably better, say more like Europeans or East Asians?

Sure: there’s no law of physics that would make it impossible. If local selection pushed in that direction, it could happen.

Has it happened? Maybe on a sub-national,tribal scale. Not on a national scale: every country in sub-Saharan Africa does poorly.

Occasionally I hear people talk about Africa’s great genetic variety. It exists: the genetic difference between Bushmen and Bantu is bigger than the difference between Bantu and Finns. A couple of thousand years ago, before the Bantu had arrived in South Africa & mixed with the Bushmen, it was even bigger: looking at ancient DNA from those unmixed Bushmen, looks as if they split off from the rest of the human race at least a quarter of a million years ago. Well before anything that looks like behavioral modernity, by any definition. Half as divergent as Neanderthals.

But the most divergent populations are small. There are fewer than 100,000 Bushmen, on the order of a million Pygmies, around 1000 Hadza. Most people in Africa are Bantu or related populations: next after that are Nilotic peoples.

This entry was posted in Uncategorized. Bookmark the permalink.

113 Responses to African variation

  1. Some say that Igbo in Nigeria are above the rest

    • AppSocRes says:

      I’ve read that the Igbo have been described as the “Jews of Africa”. I’ve also read they may have average IQs eceeding that of Europeans. It’s certainly true that two children of Igbo immigrants were finalists in recent Science Talent Searches (Both finalists mentioned here have Igbo names:,, while other Negros don’t appear to show up at all

      As for the Ashkenazim, cultural selection might play a role if these observations are correct. The Igbo have long specialize in intricate iron smithing techniques and are famed for their prowess in this area. I could imagine that the skills associated with this industry, at least as practiced by the Igbo, are improved by intelligence.

    • Chanda Chisala makes the best-known argument for the idea that some African groups do better. I am only partly convinced, but she makes some good points.

      • JW Bell says:

        Chanda Chisala is a man.

      • syonredux says:

        Yeah, Chisala…… Let’s see,success at scrabble proves that Blacks are smart….except for Black Americans, who are strangely dim…..

        • Toddy Cat says:

          Well, it’s a shame that we never got to find out how well the Ibo would perform on their own, thanks to the usual suspects in Moscow, Washington, Peking, and London. Had Biafra become independent, based on the Ibo that I have known, my guess is that it would be well ahead of most of Africa, perhaps doing as well as Mexico or Brazil, which is certainly not damning with faint praise in the African context. But equaling Europe or Asia? I doubt it. If I had to guess, I’d put the Ibo/Biafran IQ at around 90 or so, about the same as Eskimos or Maoris, which means, that, yeah, there are plenty of good Scrabble players in that lot…

        • mtkennedy21 says:

          Did the dumb ones get more easily caught and sold by the leaders ? I have wondered that.

      • pyrrhus says:

        No, he doesn’t. JayMan and many other commentators have dissected Chanda’s nonsense in detail….

        • szopeno says:

          No, they didn’t. I was actually very dissapointed by Jayman’s and dr James Thompson replies. They attacked the weakman instead of strongman, i.e. they ignored the strongest examples while dissecting the easy to dissect examples. If you were following the discussion, you would notice that I have said the same – Chanda Chisala surely have shown that in at least some countries the IQ is impossible to be in the ranges claimed by Lynn.

          • szopeno says:

            In short, there are seven top players from Gabon. Gabon have small and very young population. Also, every mental activity is correlated with g, even though the correlation may be small. Gabon has 1.8 million people; taking into account only those male and over 20y/o, i guess we are talking about population of 450 thousand or so.

            I find it extremely unlikely that all intelligent people, or even very large proportion of them, woudl decide to play scrabble instead of choosing other careers.

            After taking that into account, I have calculated the following number of the possible players above some IQ threshold (those are all rough estimation, just to give you general idea of order of magnitude)

            MEAN IQ/SD : above 115IQ Above 130IQ Above 145 IQ
            110 IQ 12 SD: – – 0.17/722
            105 IQ 12 SD: – – 0.042/178
            100 IQ 12 SD: 10.5/44625 0.62/2635 0.00884/37
            100 IQ 15 SD: 15/63000 2.27/9647 0.135/573
            85 IQ 12 SD: 0.62/2635 0.0084/35 0.0000287/1
            85 IQ 15 SD: 2.27/9647 0.135/573 0.0032/13
            80 IQ 12 SD: /722 /6 /0

            If you carry this calculation, you would find that at IQ70 and SD 12 (as is often postulated for black SD), there would be only something like 2820 people with IQ over mere 100 in whole of Gabon (182 over 110, 40 over 115…). I bet there are more Scrabble players in France. Therefore, the argument “scrabble is more popular in Gabon than in France” does not explain the results if it postulates the Gabon’s IQ at 70 and SD at 12. No way.

            I can think of several explanations, but other commenters went instead for easy pray and saying that in huge Nigeria there surely will be enough intelligent Nigerians, while ignoring the example of Gabon.

          • Criticizing Lynn without supplying better analysis for MOST country serves no use
            Note that Lynn’s IQs are phenotypic, not genotypic.

      • magusjanus says:

        Chandala’s nonsense is if anything another piece of evidence against the “it’s all environment” thesis.

        Because we have a de facto state religion progressivism with chief tent biological egalitarianism, a lot of very sharp minds are toiling away for social prestige to come up with sophisticated explanations for the obvious that do not involve genes.

        So you’d expect to at least have some interesting theories, counter examples (no matter how convoluted), etc.

        That the best they seem to come up with is Chandalas hilarious scrabble example and some Turkheimer nonsense at Vox if anything suggests how dire the state of the oficial “denialist” position is.

        • gcochran9 says:

          “a lot of very sharp minds”

          Like who?

          • j says:

            “Swiss banking huge Credit Suisse has named Tidjane Thiam as its new chief” a year ago, and he seems to be turning around that losing dinosaur. He is not Igbo.

            • j says:

              PS.: Uber-reactionary Swiss bankers are known for their policy of NOT advancing underrepresented minorities. They were desperate.

          • jb says:

            “A lot of very sharp minds” being basically the entire academic establishment! Not necessarily sharp in terms of rigorous thinking, but sharp in terms of coming up with sophistic arguments to advance the causes they care about, which is really what we are talking about here.

            Academia is full of leftists who make a living as clever wordsmiths. They can churn out vast quantities of memes like “white privilege” and “we are all Africans”, polish them, do a lot of what is essentially A/B testing, and promote whatever works best (which they are in an excellent position to do). The Internet now allows conservatives to do something similar, but because anyone can play the quality tends to be lower, and you often end up with nonsense like Pizzagate, because there is no way to keep everyone on message.

            This is yet another of the ways that controlling the commanding heights of the academy gives the Left an advantage over the Right.

          • magusjanus says:

            That’s kind of my point. They’re not coming up with much anymore that’s interesting or doesn’t fall apart under basic scrutiny. (The pessimistic hypothesis is they don’t really have to, people will buy into it cuz they want to, as they have so much other nonsense.)

            But yeah, if you’re looking for some mainstream social prestige, just come up with a new theory to explain black persistent underperformance (or Jewish outperformance) that does not reference biology. There’s a lot of competition in the field, but I’m sure you could come up with something creative that Nisbett/Marcks/Lewontin and their ilk haven’t tried yet and that Gladwell can then write a ‘for dummies’ explanation of.

            Money and fame await!

    • dearieme says:

      Could be. They were the “Ibo” when I was young, so they’ve put on 33% more letters in a couple of generations.

  2. cargocultist says:

    Are there any signs of reduced fertility when Bushmen or Pygmies mate with other humans? And what about their progeny, do we see reduced fertility there?

    • Michael Daxhammer says:

      Or even some sign of hybrid vigor? Would be cool if some pygmy and a aborigine had some eg. physically superhuman kids. Think Dr. Cochran talked about it that there were very weak hints that mixed Asians did somewhat better on some cognitive tests.

      • If you’re talking about Europeans-East Asian hybrids then the genetic distance between them is rather small. If any such advantage existed it would have been noticed and advertised by antiracist crowd.

        • Michael Daxhammer says:

          Yeah, European kids. Just found it: “This suggests that crosses between particular human populations might produce superkids. Clearly, there is no dramatic positive effect in any of the common crosses – we would have noticed. I’ve seen a study that observed a moderate IQ advantage in Japanese-European kids in Hawaii, something like a quarter of a standard deviation, but I don’t think it’s been replicated, and that’s not enough change to get excited about in any event.”

        • Michael H says:

          Brazenly anecdotal, and not about IQ… But i have Asian friends who know a lot of White-Asian couples, and they tell me that those couples’ kids tend to be unusually attractive.

          • Dave Pinsen says:

            Steve Sailer posted female composites of different nationalities he got from another blog once, and pointed out (IIRC) that the composites tend to be attractive because they represent a sort of smoothing-out of features: maybe some X have big noses and some have tiny noses, but the composite will likely have a medium nose. Maybe there’s a similar dynamic with Asian-white mixes.

      • savantissimo says:

        The thing about hybrid vigor which people forget is that it only applies to the first generation cross between two different pure strains. Breeding hybrids with hybrids usually gives worse results on the whole than either pure parent strain, but a great deal more variability in the results — which might be useful in developing new strains, but is undesirable in crops or livestock that are to be sold at market. Even if some races’ progeny had hybrid vigor, they could not continue to be born without the continued existence of the pure strains, an existence which is incompatible with the current elite campaign of multi-culti-genocide which claims that failure to mix all races into one is racist.

        • j says:

          Just mentioning “pure strains” makes you suspect. You must know that there is no such thing.

          • melendwyr says:

            There are, actually – artificially bred populations that have been rendered as uniform as possible.
            There’s no inherent rule that the descendants of hybrids must be worse. That’s only the case when we make artificial F1 hybrids out of severely inbred populations – the F2 and later generations tend to manifest the problems of their grandparent populations. If the relevant strains were natural, the only thing you can say about the F2+ generations is that they’ll be highly variable.

          • Frank says:

            Pure strains exist. They are just doubled haploids. As long as their population is closed, they are basically clones. In the lab and farming test fields, this is common. Zero diversity.

            In humans, Khoisan and African Pygmies are the furthest thing from pure strains. They have the highest diversity, not the least.

            But, if you wanted to select for ANY particular trait, they would be the ideal population to start with.

            All of the positive stuff is in there, it is just balanced out by negative stuff. They have had the least amount of genetic selection on a genome wide scale of any living humans. Of course they also have had a few selective sweeps, but not so many bottlenecks.

            are the furthest

  3. JayMan says:

    “Occasionally I hear people talk about Africa’s great genetic variety. It exists”

    The problem is (as I often encounter) that people hear that and they think (overall) genetic diversity corresponds to phenotypic diversity on a near 1:1 scale, particularly in whatever trait of interest. Which of course is silly.

    • Peripatetic commenter says:

      Heh. Yeah. I have heard otherwise intelligent people claim that because genetic diversity is so great in Africa there must be a population with average IQ of 150 in Africa somewhere!

      • Ironically, the Ashkenazim singer/songwriter (and damn good fencer from NYU) Neil Diamond, who likely has a pretty decent IQ, wrote that song.

      • magusjanus says:

        D Mcloskey had a part in his/her latest book on this, and was (rightfully) torn to shreds here. Even showed up in the comments section to take more punishment.

        Which is a shame cuz Second book in the series, Bourgeoise Dignity, was decent, other than the total inability to understand regression to mean which Greg Clark had to write a retort with basic math to (Google it).

        Speaking of which, I have no idea how GClark gets away with stuff he does. It’s gotta be the accent…

  4. harpersnotes says:

    Warm latitudes mean greater disease prevalence, all other things equal. Also, population densities that have been around for a long time (ancient Egypt?) Parasite selection pressure drives diversification of gene alleles for bodily tissues as a resistance strategy, but from the perspective of other adaptations can act as a kind of noise. (i.e. Antagonistic pleiotropy. See also life history evolution generally.) So perhaps along with certain benefits of genetic diversity comes potential costs in the developmental integrity of the most extremely complex and recently-evolved adaptations, such as seem to be associated with some types of cognition. Or perhaps not.

    • dave chamberlin says:

      Conversely colder climates allow more children to reach adulthood because of decreased mortality rates. More children reaching adulthood leads to increased malthusian pressure which over enough generations results in exactly what intelligence tests tell us. Average IQ goes up the farther you move away from the equator. It is really quite simple. What confuses this issue is people not wanting to believe it.

  5. anonymous says:

    Is it true Nigerians have more potential than the local economy provides?

    • whyteablog says:

      Does the terrible local economy dampen potential, or does the terrible potential dampen the local economy?

      • Not high enough IQ says:

        Does the body rule the mind or does the mind rule the body? I don’t know

      • Ursiform says:

        Both are true. But the latter determines whether you can establish a positive feedback situation and advance.

        • whyteablog says:

          I was being cryptic in a silly attempt to guide the original commenter’s thoughts.

          Original commenter, this guy is right.

          • anonymous says:

            What I meant was in another discussion somewhere it was claimed the government over there represses the population. Perhaps to prevent revolution? They didn’t go into details…

            • Ursiform says:

              It’s more vastly corrupt and incompetent then repressive for the sake of repression. Although they do repress areas with separatist tendencies.

            • whyteablog says:

              At that point you have to ask bigger/deeper questions. Why would all African countries be like that then- would all their governments be like that? If so, why?

              The usual handwave is colonialism. But why were these people so vulnerable to conquest that their entire continent got rolled over? Up until Jared Diamond wrote “Guns, Germs, and Steel” people typically couldn’t explain why Africans were ripe for getting their asses handed to them by Europeans. I think a lot of holes could be and have been poked in Diamond’s book, but that’s another topic.

              Look for explanations that aren’t “turtles all the way down” and can’t be demolished by having their most important assertions and assumptions refuted.

              • anonymous says:

                I wasn’t expecting an army of rocket scientist or Wakanda… There were stories of Facebook snatching some developers though. A comment on another blog claimed “barbarians” can be more dynamc and Nigeria would become a powerhouse if average IQ increased to 100 (not that it would happen). Those Nigerian scammers do seem enterprising at least. Just wondered about not reaching potential, at whatever level it was.

    • anonymous says:

      Looks like Nigeria isn’t particularly promising. The Ravens IQ test cited gives more points per standard deviation BTW but still. You would think specific tribes (or caste within tribes, like goldsmiths, merchants, or royal families) would do better…

  6. Insightful says:

    What i find incredible is the genetic difference between TWO unrelated Bushmen, say, a couple hundred miles apart, is greater than that between a Han Chinese and a French person who are a continent apart!

  7. says:

    It seems that many people, including me, have not estimated the amplification factor for self and migrant host country selections. By using just two simple assumptions, 1) national and migrant group IQ are normally distributed; 2) percentage of university graduates is meritocratically determined, the amplification factors for various migrant groups in some host countries can be determined. And for Nigerian in UK the migrant group IQ is about 2 SD above the mean national IQ value from Lynn. The selection factor is dominant and the mean national IQ value appear to be secondary.

    Step 1 is the determination of the Expected Nominal IQ of Graduate from Multiple Attributes (ENIGMA) from the given IQLynn and the percentage of graduate in the country data from OECD or World Bank. Step 2 is to pivot that ENIGMA value for the migrant group in host country with the known percentage of graduate to calculate the Mean Average Group IQ (MAGIQ) for the migrants. (This is in contrast to the ‘magic dirt’ theory for migrant success.) In the UK case the quality of the migrant groups is reflected from the performance of their children in GCSE and the empirical CAT3 data collected.

    It is interesting that the results show that the Nigerian children should perform much better than those from India, Pakistan, Chinese from HK and Singapore, they did not send the best to UK. It also agrees with the finer trend that India’s lower IQLynn, Indian children should be performing better than the Pakistanis.

    A real surprise is the predicted performance of the Mexican due to the high UK selection criteria. Comparatively, the equivalent Mexican MAGIQ score for the very lax US is 86.74 .

    The model is also able to explain the high performance of sub-continent Indian in US dispite the low IQLynn for India.

    The results also show the popular assumption of the IQ of graduate to be 1 SD above the mean, i.e. 115 is mostly not true.

    • says:

      NG 84 106.1 115.0
      IN 82 105.8 104.1 99.8
      PK 84 107.8 101.3 93.5
      BD 82 111.9 100.8 94.7
      JM 71 91.7 82.9 94.0
      HK 100 114.2 110.9 107.5
      SG 108 115.6 116.4 107.5
      MX 88 102.8 117.3
      GR 92 105.8 115.5
      IL 95 102.9 107.8
      TR 90 106.2 100.8
      IE 92 103.4 96.7 100.4
      IE travellers – – – 87.4
      Roma – – – 87.9
      BlackAfrican – – – 93.4

    • says:

      A better model will be to use the time series data for at least 3 generation instead of the current model of using only a single year data points. However data are hard to come by.

    • says:

      Using the MAGIQ scores, an interesting brain gain/drain quality model for the developed nations. Germany and France are in serious problem.

      Host Source Out In IQDiff
      DE AU 113.10 107.25 -5.85
      DE CA 124.61 104.50 -20.11
      DE FR 110.40 112.28 1.88
      DE UK 113.50 107.07 -6.43
      DE US 116.13 105.48 -10.64

      FR AU 118.18 113.22 -4.96
      FR CA 127.14 115.05 -12.10
      FR DE 112.28 110.40 -1.88
      FR UK 120.62 107.88 -12.74
      FR US 121.68 117.50 -4.18

    • Cecil B. Whiffenpuffen says:

      There is one major problem with this: the GCSE is not an accurate measurement of anything. For one, the exams are taken at 16 and not 18. More importantly they replaced a much more rigorous system called O-levels that is still in place in the parts of the Commonwealth that value education like Singapore. GCSEs were explicitly designed to allow girls to exceed boys in educational attainment. 80% – 25% of the mark is awarded for coursework (subject dependent). The exam component has also been extremely heavily dumbed down. You can confirm this for yourself by looking at GCSE Maths papers and comparing them to O-level Maths from the late 80s. There is also very wide variability in difficulty between subjects, and the overall GCSE results do not take this into account. An A* in Physics is not the same as an A* in Media Studies. London schools with high numbers of immigrant children push those students to take easier options at GCSE, thus artificially inflating their scores. There are a whole raft of subsidies, promotions, awards, and sanctions associated with a school’s overall GCSE grades and their place in national “league tables” so teachers are very heavily incentivised to game the system. I went through GCSEs in the late 90s and teacher cheating was obvious to me even at 16.

      GCSEs are highly politicized garbage and their results do not indicate anything other than how far the educational establishment in the UK has fallen.

      • says:

        Currently most people used the native source average IQ as the proxy for the migrant group average IQ. My investigation is to determine how much the self and host countries selection processes can skew the resulting migrant group IQ. Some people questioned the estimated values of the source native IQ. My investigation shows that the variations from the selection processes are more dominant. The empirical CAT3 results is just a check to see if the results make sense, and whether the source native IQ or the MAGIQ score is a better proxy.

        For example for the Nigerian, IQLynn=84, from World Bank data FracGrad=0.07, from UK census via IAB dataset Nigerian migrants with university degree FracHi=0.72. Very strict UK selection process for Nigerian. That give the Nigerian in UK MAGIQ=115.01, an increase of 31.01 IQ points, slightly more than 2 SD IQ points.

        In contrast for HongKong which is the main source of Chinese migrants in UK, IQLynn=100, from HK census FracGrad=0.1721, from UK census via IAB dataset FracHi=0.41, that gives MAGIQ=110.94, an increase of only 10.94 IQ points. Compare Nigeria’s FracHi=0.72 with HK’s FracHi=0.41. HK did not send their very best to UK. The associated empirical CAT3 score is 107.5 .

        For India, IQLynn=82, from Indian census FracGrad=0.0564, from IAB data FracHi=0.46. That gives MAGIQ=104.12. An increase of 22.12 IQ points. Compare to CAT3 of 99.8 .

        Furthermore, the MAGIQ model can estimate the migrant group IQ to be below the s
        ource native IQ, e.g. in the case of US host, Mexican native IQLynn=88, FracGrad
        =0.1622, FracHi=0.142, FracHi < FracGrad, thus MAGIQ=86.74 is less than the s
        ource IQLynn=88.

        • Cecil B. Whiffenpuffen says:

          I understood what you were trying to do the first time and was offering a critique of this statement: “In the UK case the quality of the migrant groups is reflected from the performance of their children in GCSE and the empirical CAT3 data collected.”

          The CAT3 is even worse than GCSE. How old are children who normally take these tests? 11 – 14?

          This prior: “2) percentage of university graduates is meritocratically determined” is also nonsense.

          Please feel free to engage with this critique or simply restate, again, what you are trying to achieve.

          • says:

            Re: “””a critique of this statement: “In the UK case the quality of the migrant groups is reflected from the performance of their children in GCSE and the empirical CAT3 data collected.” The CAT3 is even worse than GCSE. How old are children who normally take these tests? 11 – 14?”””

            The correlation of children’s performance with respect to that of their parent is well established, e.g.

            “””Long-term Effects of Parents’ Education on Children’s Educational and Occupational Success: Mediation by Family Interactions, Child Aggression, and Teenage Aspirations”””

            “””Parental educational level is an important predictor of children’s educational and behavioral outcomes (Davis-Kean, 2005; Dearing, McCartney, & Taylor, 2002; Duncan, Brooks-Gunn, & Klebanov, 1994; Haveman & Wolfe, 1995; Nagin & Tremblay, 2001; Smith, Brooks-Gunn, & Klebanov, 1997).”””

            “””third graders in a semi-rural county in New York State were interviewed along with their parents; participants were reinterviewed at ages 19, 30, and 48 (Eron et al, 1971; Huesmann et al., 2002). Parents’ educational level when the child was 8 years old significantly predicted educational and occupational success for the child 40 years later.“””

            I leave it for others to judge your evaluation.

            Re: “””This prior: “2) percentage of university graduates is meritocratically determined” is also nonsense.”””

            You might question the effectiveness of the selection processes, but the university graduates had to go through selection examinations before enrolment, the performance evaluated throughout the course before graduation. If you think that is nonsense are you implying that any random person from the street can walk in and get a degree without efforts or as capable as a graduate?

            Again I leave it for others to judge that.

          • Cecil B. Whiffenpuffen says:

            Is English your second language? Because I am struggling to think how I could have made it more clear that I am criticising the GCSE and the CAT3 as reliable measures of the childrens’ performance (and as proxies for adult performance). The data that you have chosen to use is very heavily distorted by a number of political realities that I am attempting to explain to you with little success.

            I am not disputing that there is close correlation between parents’ and childrens’ outcomes and achievement in education. That is obvious. What I am disputing is that the GCSE and CAT3 are a good proxy for that. If your data is garbage and your assumptions are garbage then the conclusions that you reach will also be garbage.

            GCSE has been very heavily distorted as I have already explained. CAT3 is an intelligence test administered between 11 and 14, when environmental factors have more influence than at 18. Do you agree with this statement? Do you also agree that racial IQ gaps are generally smaller in younger children, growing larger at adulthood? Or is that too controversial?

            There is a very strong environmental factor that heavily favours the BAME (the current “correct” term over here – black and minority ethnic) children over white children. They receive very much more funding per pupil than native children do. This is rarely discussed. The funding mechanism is called the Direct Schools Grant. Here are the figures for 2013: ]…

            You clearly have no familiarity with the UK at all so let me highlight some facts from this document. Majority BAME Tower Hamlets received £8,051 per pupil. Majority native Cheshire East received £4,695 per pupil. Majority native Cornwall £4,663 per pupil and majority BAME received £7,811 per pupil. The city of Leicester with its massive BAME population received £5,290 per pupil. Rural Leicestershire, right next door but now overwhelmingly native, received £4,428 per pupil. Do you think that this might be distorting the CAT3 data? Because I certainly do.

            The willingness of strangers on the internet to argue passionately about subjects of which they are totally ignorant never ceases to amaze me. A short lesson on the history and structure of higher education in the United Kingdom:

            In 1992 the Further and Higher Education Act was passed by Parliament. It allowed 35 institutions that had previously been called Polytechnics to become universities. The function of a Polytechnic was the same as a Community College in the US and they did not award degrees. That changed. The sector went from having c. 60 universities to more than 100.

            There are currently 78 “new universities” in the UK. They have the same status as the old universities in law but are not respected, and quite rightly too given the quality of education they offer. Native middle class children almost never attend them, and only native working class children who don’t know any better do. These schools are attended by BAME students who don’t know they are being conned in disproportionate numbers: 49% of students at the University of Westminster, 54% at the University of East London, 54% at Middlesex University.

            No, these new universities do not have selection examinations for enrolment. And their performance evaluations are absolutely nothing like the evaluations you will find at the pre-1992 universities. Let me state it clearly and unequivocally: any random person from the street can walk into any number of post-1992 universities in the UK and get a degree without effort or as a capable graduate. I employ around 10 technical graduates a year. I interact with probably thousands of graduates a year as I tour universities in the UK looking for them. Long experience has taught me to throw any CVs from a 1992 university straight in the bin.

            Your data is garbage, your assumptions are garbage and your conclusions are garbage. All you’ve demonstrated is that you probably have tenure at Harvard.

            • says:

              Rather than just hand waving I am interested in getting empirical data to calibrate the theory and CAT3 seems to be the only one available. If you have more reliable migrant IQ sources I will be interested.

              Just an aside. The minimum IQ for (reputable) university graduate is traditionally assumed to be 115. Just to use my assumption and only two publicly available numbers,
              the average IQ of USA IQLynn=98 and the fraction of graduate data Frac=0.38 from OECD, the assumption shows that the minimum IQ for USA graduate is 105.5. Somebody has estimated the minimum IQ differently by using the SAT data and found that the minimum IQ needed (to enrol, not to complete the course) was 103
              Compare to the traditional assumed value of 115 that estimation method is pretty useful.

              As to the sorry state of GCSE and the standard of UK graduate you described, that is exactly what I am trying to find with the model what are the causes.

            • says:

              Nial Ferguson mentioned that since after WW2 there was steady emigration out of Britain
     at the 30 min mark. I am trying to find if the British talent bank were/are being drained out.

              With the average IQ of British at 100, some of the estimation from the model for the IQ of British emigrant groups to various countries in 2010,

              Source Destination MeanIQ
              UK AU 107.41
              UK CA 116.73
              UK CH 109.24
              UK DE 107.07
              UK DK 101.81
              UK ES 98.01
              UK FR 107.88
              UK GR 110.31
              UK IE 106.91
              UK LU 108.60
              UK NZ 105.91
              UK SE 108.99
              UK US 114.06

              If the UK situation is as bad as you described, may be there is some merit to the above result. The IQs of most British emigrant groups appeared to be above 100.

              These sort of data are non PC. Definately not Harvard materials.

            • Scammed from the pubs says:

              That was a very insightful look on the state of higher education in the UK. A few years ago I did a 1-year MSc course in one of those “new universities” as an EU student. There were virtually zero native postgrad students, a handful of BAME ones, and the rest of the class was made up of roughly 60% South Asians, 15% Chinese, 15% Africans, and 10% Euros. All of them were either not smart enough to do a course in a proper university, or they were too lazy to put the work needed.

              The quality was so low that I handed a 5k word dissertation (where the absolute minimum was 10k words) and not only did I not get cut, but got a good grade, just because I could string some spelling and grammar error free sentences together that actually made sense. There’s countless other examples I could bring up to show that it resembled more a 3rd world university than a proper UK institution.
              Getting such a degree is like being scammed into buying a fake mobile phone that looks like a real one on the outside and on the inside, but never works. You’re just going to get it if you pay full tuition, unless you voluntarily drop out.

              None of those stats that lump all universities together should be taken into account for any calculation. Is there anyone who’s published such data broken down by university type?

              • says:

                The ENIMA score in my model is the estimation of the weakest ability of the graduate. The result for UK is not that good. With mean national IQ of 100, the fraction of graduate with minimum IQ of 115 should be 0.159 but UK has graduate rate at 0.286, something got to give. The ranking from the model

                Rank ENIGMA IQLynn FracGrad Frac115 SRatio Country
                3 117.6 100 0.12 0.159 1.32 AT
                4 114.84 106 0.278 0.274 0.99 KR
                5 114.51 105 0.263 0.252 0.96 JP
                6 114.49 99 0.151 0.143 0.95 DE

                24 108.47 100 0.286 0.159 0.55 UK
                33 105.54 98 0.308 0.129 0.42 US

                As I said above, the score of 105.5 for US is close enough to another independent estimation of 103 using a completely different method. Not that high from the random average person from the street. The details are in the numbers. Not PC to directly talk much about them.

                Nevertheless, instead of the university type data you asked for there are data for scores and graduation rates for individual US universities/colleges who participated in a CLA+ survey but the some of the ivy leagues are not there. The results are published in the WSJ,

                From the data it can be determined the best practice graduation rates for the universities with respect to the students’ scores and less demanding universities can be noted.

                GradRate=+0.124027*CLASenior-86.0292; # n=67; Rsq=0.4384; p=1.05e-09

                With such low p value the result is highly statistically significant. The universites with the largest positive deviation from the respective expected graduation rates,

                zscore ExpectedRate GradRate CLASenior
                1.5 69.62 84.0 1255 University of Georgia
                1.78 42.96 60.0 1040 Eastern Illinois University
                1.79 45.94 63.0 1064 Keene State College
                1.81 55.73 73.0 1143 Ramapo College of New Jersey
                2.73 42.96 69.0 1040 Citadel

          • Cecil B. Whiffenpuffen says:

            The link doesn’t seem to be working. Here is the data for DSG allocations 2012-13:

  8. ” […] there’s no law of physics that would make it impossible. If local selection pushed in that direction, it could happen.”
    That also applies to chimpanzees and other apes. It is true that if the standard of living in Africa were raised, the IQ would probably also increase because of the Lynn-Flynn effect, but it is pure optimism to think that natural selection can do that today. Natural selection no longer favors the most intelligent, resourceful and adaptable individuals, but those with more children (as it always was), who are unfortunately low-class, low-IQ individuals. Not to mention that countries with a higher IQ suffers the addition of migrations of races with lower IQ.

    Some say that increasing the Lynn-Flynn effect in developing countries operating in conjunction with fertility and dysgenic migration in the developed world could virtually wipe out the current IQ breach.

    • magusjanus says:

      Blacks in Us provide a good “ceiling” for ss-African black iq. And US blacks have avg white admixture of 30-50% from what I recall, so if anything all black avg genetic iq post Flynn effect gains probably caps out at Low 80s at best.

      Probably the biggest population group iq gains to be had in the world would be dressing up as the Angel Gabriel with a light show in Mecca and saying “oh, And one more thing I almost forgot, don’t marry your cousins”.

      Any volunteers?

      • gcochran9 says:

        Average white admixture in US blacks is 17%.

        • syonredux says:

          Here are the results from a study that used data derived from 23andMe customers:

          “Broadly, the genomic analysis found that on an average the African American genome was 73.2 percent African, 24 percent European and 0.8 percent Native American. Latinos as expected had significantly more Native American ancestry with the average Latino genome being 18 percent Native American, 65.1 percent European and 6.2 percent African.”

          Gotta say, though, I’m not sure how representative people who pay for a 23andMe test are…..

          “It is important to note that the study has several potential biases that should be taken into account such as the socio economic status of those who might have purchased the test and as Carl Zimmer reported in his piece, the fact that people of mixed race are more likely to get the test out of curiosity. Nevertheless, the large sample size lends unique credibility and the trove of data will no doubt continue to yield very interesting results about how the ethnicity of the current American population came to be.”

          • 24% is too much in my opinion. Although I am willing to accept that conclusion. But I think the percentage of Latinos is not representative, look at a common Latino, it’s mostly American Indian. Moreover, what is a ” Latino ”? A Latino is not a racial category because in Latin America there is a great racial diversity, especially in Brazil and in the Southern Cone. It would be like saying that in North America inhabits the American race, Americans are a race, it does not make sense. Racial censuses in the United States began with good standing trying to classify races (understood as different human morphological lineages), but I think that at some point that was corrupted and there are currently categories that have no relation to reality.

          • DataExplorer says:

            “Latino genome being 18 percent Native American, 65.1 percent European and 6.2 percent African.”

            I guess Trump was wrong when he said they are not sending their best. The Mexican genome is evenly split at 50/50 European-Native American. The Average USA “Latino” seems to reflect Colombian rather the Mexican averages.

            • JerryC says:

              Again, this data isn’t on “The Average USA “Latino”, it’s on The Average USA “Latino” who shelled out for a 23 and Me test.

  9. Rum says:

    Speaking of hybridization: If one starts with the assumptions that ancient proto-human development started in Africa and that successful interbreeding across sub-groups is possible (I am thinking of canines, for example) could not one explanation for diversity in Africa be that Lucys genes are still in the mix? In other words, the continuity of genetic humanity is simply deeper into the past there.

  10. Rum says:

    The other half is the notion that, regardless of where the Neanderthals started out from ~300,00 – 500,000 years ago, they ended up in parts of the world where the option for breeding with older sub-groups did not exist.

  11. kot says:

    A while ago greg said this:

    For example, suppose we get LIGO working, and detect the signature of two neutron stars spiraling in towards collision – two relatively nearby neutron stars. Near enough that the gamma-ray pulse would kill everyone on one side of the Earth. There are things you could do, but probably we wouldn’t do anything. We’d just die.

    How accurately could we time the collision? To the minute, day, year?

    If the answer is >24 hours… I’m picturing the British royal family in a continuously refueled jet (blimp?) hanging out at the antipode… every time they fly back into Britain’s hemisphere they breathe a sigh of relief, etc..

  12. Halvorson says:

    Two weeks ago Schlebusch et. al revealed that West Africans trace 31% of their ancestry to a group they call “Basal Humans” who split apart from the modern human tree at some unknown point much earlier than the Bushman divergence. A result like this had been predicted by Hammer and company a long time ago.

    This is kind of a big deal.

    • dearieme says:

      Presumably quite soon it’ll be forbidden in polite society to discuss ancient African DNA and its lessons.

      • dave chambelrin says:

        I have a guess on this from what we know now. Ancient African DNA doesn’t dent African IQ any more than neanderthal DNA dents (or helps) out of African DNA. IQ via genes is a vast milieu that bobs up and down not by good new genes or bad old genes but by evolutionary pressure that in some cases rewards higher IQ and sometimes are almost indifferent to higher IQ so in those cases it slowly decreases IQ.

        All this will be huff puffery until genetics progresses a bit farther which is coming in the lifetime of young people so stay tuned to the few folk like Cochran that aren’t full of shit in reporting on this subject.

        • dearieme says:

          The lessons from ancient African DNA might have nothing to do with IQ. I wouldn’t be surprised if they were much more interesting than that.

        • Halvorson says:

          I highly suspect Neanderthal DNA does dent IQ, given what we know about their tool complexity and the many Neanderthal genes that have been under negative selection in the last 50,000 years. But it’s only 2% of all Eurasian DNA, not high enough to have a visible impact. The amount of archaic ancestry in Africans, at least West Africans, is massively higher and could be much more recent if it came from a group like the 13,000 year old Iwo Eleru skull. This would leave less time for IQ-lowering alleles to be purged.

          • says:

            From Pearce’s result, Neanderthal had large brain and large eye sockets. Thus significant portion of their brains might be devoted to image processing. Brain cells are plastic and can be adapted for other function like cognition.

            A plot of % of Neanderthal introgression in sapien from Sankararaman shows statistically significant linear increases with OECD PISA scores.

            PisaM12 = +169.731*PctNeand +288.75; # n=18; Rsq=0.2911; p=0.02084

            Thus there could be direct or colleteral factors but nobody wants to study that.

  13. Jim says:

    Regarding the split of Bushmen from the rest of the human race it is interesting that the Khoisan languages don’t seem to be fundamentally different from other human languages though they are notable for a very large inventory of consonant phonemes and an abundance of “click” (velar suction) phonemes.

    • gcochran9 says:

      If there were interesting qualitative differences in language, i’m not sure that anyone would admit it. The psychometrics people (and the track record) see big differences in intelligence between populations, but anthropology (with a few exceptions) doesn’t admit any such thing. They talk about the “psychic unity of mankind”, but I’ve never heard of any of them making measurements to see if was actually so.

      • AllenM says:

        The really interesting question is how much should researchers be testing in all of Africa to find ancient lineages before they disappear. After all, the Bantu replacement will be essentially complete when they finish crowding out the original San populations. Some of those ancestral lineages will be preserved through admixture, but that will be ultimately selected through fitness and sweeps.

        The interesting question is how people will feel about these ancestral lineages, especially if they are not able to compete with newer lineages. The intense pressures on the San and the Pygmy populations over the last five centuries have continued unabated, and the end of them as separate identifiable populations will happen unless a preserve system is implemented by African polities.

        After all, even though Northern Europeans have large Neanderthal genetics in their background, no Y or mtDNA from them exists, and most likely even the X chromosome contribution is minimal. Given this, what are your thoughts on this paper:

        My thinking is that humanity has diverged significantly, and these divergences will have real implications in dealing with newly mixing populations in modern society. One of the most interesting hints seem to be the expression of some of the Neanderthal genes in brains.

        I guess that one has to be flexible in reading the research, because the conclusions are going to be very unsettling to the modern sensibilities, lol.

  14. Isn’t there also the question of on what dimensions the genetic distance exists? There could be much larger genetic variation between subsets of African groups on dimensions we don’t find explanatory towards social/political phenomena, but a closeness on the dimensions that we do. Is this possible?

    • gcochran9 says:

      Generally people talk about genetic distance in the genome as a whole ( if they have that data), which is mostly neutral. That’s pretty much a function of separation time and population size. You could have two species that were fairly genetically distant and functionally nearly identical.

  15. M. M. says:

    “O’Keefe gave a TEDx Talk in October 2016, on the natural origin of homosexuality. In this talk, he proposed a theory about the evolutionary basis of homosexuality mediated by epigenetics.”

    I have no idea. Any thoughts?

    • melendwyr says:

      Humans are far less sexually dimorphic than most of the most closely related species of apes, and there are some indications this extends beyond gross traits like size to brain structure and function. If part of the reason that’s the case is that we have a trait that randomizes the expression of the male- and female-specific programming, so that the original design with a rigid and consistent distinction between maleness and femaleness is blurred, that randomization might occasionally result in people whose sexual orientations strongly mismatch their biological sex.
      It’s at least conceivable that the reason we can’t find a ‘gay gene’ or even a complex of genes linked to homosexuality is that everyone already has them – and expression leading to homosexuality is determined by chance combined with physiological insults to the fetus.
      Handedness is something like this – being left-handed is correlated with having a wide variety of physiological problems, even though most left-handed people are fine, because conditions arising from defects in fetal development are even more strongly associated with non-standard hand dominance. Homosexuality may exist because it’s not worth evolution’s time to develop a way to shield the parts of the brain responsible for sexual preference from a process which makes men and women more like each other.

      • gcochran9 says:

        You know, you say a lot of stupid things. Just to make it perfectly clear, everything you’ve ever said about biology and genetics is wrong, and as far as I can see, that’s always going to be the case.

        For example, human skeletal size dimorphism is intermediate between that of chimpanzees and gorillas. Not “far less than”. You’re wrong.

        Men are about 20% heaver than women. However, humans are more sexually dimorphic than overall body mass along suggests, mainly because human females have a lot of body fat. Fat-free body mass is 31-43% greater in men than women.

        Men have about 90% greater upper-body strength, so the average man is stronger than 99.9% of women. The sex differences in muscularity are comparable to those of gorillas. Again, you’re utterly wrong.

        If you were talking about some obscure South American rodent, your errors would be excusable. But as it is, I have to wonder if you’ve ever had much experience with human beings.

        • M. M. says:

          Err, Greg, I have the suspicion that you don’t favour Mr O’Keefe’s theory but I’d like to know for sure if possible, some people want to know, me too.

        • melendwyr says:

          Combined strawman and exaggerated claims, your standard response.
          I know you like to insist that pathogens are the only possible explanation for the existence of all forms of homosexuality… but as is often the case, you’re far more confident than the evidence justifies.

          • gcochran9 says:

            “Homosexuality may exist because it’s not worth evolution’s time to develop a way to shield the parts of the brain responsible for sexual preference from a process which makes men and women more like each other.”

            I would have to come up with new words to completely express how wrong this is.

      • Actual Anthropological Factual Information says:

        “Human sexual body size dimorphism (male/female ratio) is on average 1.15, though depending on the location values range from 1.09-1.28 … Gorillas, chimpanzees, bonobos and orangutans all exhibit sexual body size dimorphism, but to different extents and for different ontogenetic reasons. Lowland gorillas show the greatest dimorphism, having a male/female bodyweight ratio of 2.37. Orangutans also show large dimorphism (male/female ratio = 2.23), while the ratio for bonobos and chimpanzees is more moderate (1.36 and 1.29 respectively). “

        Chimpanzees: Muscle in females averages 37.4% of total body mass, ranging from 30.1–44.1%. Males have more than half of their body mass as muscle, 51.6% on average, ranging from 48.0–56.1%. There is no overlap with females.

        Humans: Men had significantly (P < 0.001) more SM in comparison to women in both absolute terms (33.0 vs. 21.0 kg) and relative to body mass (38.4 vs. 30.6%).

        Chimps more dimorphic on size and muscle mass.

        • Garr says:

          In my neighborhood the women out-obese the men (women get fat more quickly; both men and women here are fat but the women more so) so general fatness might tend to skew body-weight results in favor of women.
          About male-female difference in upper-body strength: I’ve noticed that the difference for shoulder-pressing, although dramatic, isn’t as dramatic as it is for bench-pressing or upper-back (rowing/pullup) lifts. I have a Darwinian theory about why this is the case. Note that shoulder-muscles would come into play in hunting and fighting to a greater extent than chest and upper back muscles do (punch a heavy bag and you’ll see; it also seems to me that spear thrusts would be primarily anterior-deltoid moves. So chest muscles (gripping something in front of you tightly from either side) and upper back muscles (pulling it back in towards you as it’s trying to wriggle away) would have another sort of reproductive advantage for ancestral male hominids …

  16. Neocolonial says:

    How hard would it be to knit a whole genome from scratch? Seems to me like a chemical engineering problem that requires a degree of attention.

    Process, speed and accuracy. Simple feedstock. Conceptually this appears not difficult – is there unexpected complexity down the chain?

    • Frank says:

      It is hard, but progress is rapid. We can already make complete bacterial genomes. Probably by next year they will have made a few synthetic yeast genomes.

      But humans are much more complicated than single celled organisms. The epigenetic interactions between the different cell types and alternative splicing and DNA modification and transcriptional effects are hard to simplify at a DNA level.

      Something like a mammoth will be first, because of public interest, but it will just be coding region changes. Not a ‘true’ mammoth, but good enough to look like one.

    • dearieme says:

      Never say never, but biological organisms are infinitely more complicated than the sorts of things that are the chemical engineer’s stock in trade.

    • dearieme says:

      I can never read this kind of stuff without being distracted by wondering whether or how it applies to me. It takes a conscious effort to adopt the scientific attitude instead. Is that common? Unavoidable?

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 )

Google+ photo

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

Connecting to %s