These two books, both from Princeton University Press, struck me as taking very different views about the prospects for policy interventions to affect cognitive ability.
1. Jonathan Rothwell, an economist, wrote Republic of Equals. p. 291:
As I’ve argued, there is no logical or scientific basis for hereditarianism. The best available genetic research shows that the heritability of IQ has been greatly exaggerated, even within the same age group and by greater amounts across generations or societies at different levels of economic development. The environmental origins of the massive two-standard-deviation increase in cognitive ability over the past 100 years are themselves powerful evidence that genes cannot account for more than a modest share of variation across individuals.
p. 285:
At the heart of this book is the claim that equal access to education would have profoundly pro-egalitarian effects on the distribution of IQ, income, and wealth, among other good things.
2. Kevin Mitchell, a neuropsychologist, wrote Innate. He emphasizes that variation in brain structure occurs during gestation, meaning between conception and birth. p. 9:
In sum, the way our individual brains get wired depends not just on our genetic makeup, but also on how the program of development happens to play out. This is a key point. It means that even if the variation in many of our traits is only partly genetic, this does not necessarily imply that the rest of the variation is environmental in origin or attributable to nurture—much of it may be developmental. Variation in our individual behavioral tendencies and capacities may be even more innate than genetic effects alone would suggest.
and p. 53:
Many studies have looked for systematic associations between specific environmental factors or experiences that differ between siblings and specific behavioral outcomes. These typically fall under a number of categories including differential parenting, peer relationships, sibling interaction, teacher relationships, and what is known as “family constellation” (birth order, age difference between siblings, whether or not they are the same gender, etc.). The results from these studies are very clear. They have failed to identify any robust, consistent, or substantial effects on any of a variety of outcomes including adjustment, personality measures, or cognitive ability.
This struck me as a generalized form ot the Null Hypothesis.
Try climate and geography . How does cognitive character vary with the weather of the geography. I would speculate that in an extreme 40 below blizzard winter one needs all three of those theories to survive.
There is a podcast with Kevin Mitchell on Innate at Brain Science Podcast—https://brainsciencepodcast.com/bsp/2019/159-mitchelll—worth listening to.
This is not so hard people. Look around do groups of people exist who are really big, and really small? Yes, these groups have a genetic differences from one another, however those differences are outliers. Think Pygmies very short, and the Dinka very tall. These groups are small. When you are more than a SD away from the mean genetically, as a group, it’s hard for a group to keep this up when there is gene-flow from neighboring groups. Too much gene flow between Europe, Asia and northern Africa for the groups there to be too far apart either in height or cognitive ability. If you find a group that is far outside of the mean, they likely have been endogamous for a very long time, or have an environment that a maintains a strong selection bias overtime.
So if someone tells you a difference between groups is genetic. Think, what was gene flow between those groups. It takes very little to keep groups broadly similar. I would be really surprised if any continental scale group is more than .5SDs away in height or IQ from a genetic standpoint.
Actually, geographic features like the Sahara Desert and the Himalayas divide the Eastern Hemisphere fairly neatly into thirds: Sub-Saharan Africa, West Eurasia, and East Eurasia. There has not been a lot of gene flow between the three.
There has been significant gene flow between East Eurasia and West Eurasia, sometimes in the form of Mongol hordes (which conquered everything from Turkey to China). Sub-Saharan Africa less so until relatively recently, AFAIK.
Gene flow between continents was actually very low for most of the last few tens of thousands of years and has been low enough for significant genetic adaptations to occur even in the span of a few millennia, e.g. lower platelet counts in residents of the Tibetan plateau (who’ve only lived there ~7000 years) or the sickle cell anemia gene in parts of subsaharan Africa. David Reich mentions in his book evidence of extreme endogamy (to the extent of possibly essentially no interbreeding for thousands of years) between ethnic-caste groups in India that live basically side by side. I think you severely overestimate the extent to which gene flow could have homogenized human genetic variation historically.
The tibetans prove my point. They have an adaption. You’ll probably find those genes in the surrounding communities, but they have no advantage there, so they never hit fixation in other populations. The fact that they exist means there was continual gene flow. So you have a small outlier group, who stays that way because of their environment. Other groups have the same genes, but without constant gene flow, the gene which carries no advantage would eventually be eliminated outside the area it’s advantaged.
So for sub-Saharan Africa the fact that 5% of the population has lactose tolerance, means not only was there gene-flow, but continuous gene-flow. Otherwise it would be even lower. The real question is whether IQ or height offer a reproductive advantage. If they do, then the smallest gene flow will make sure those genes spread far and wide. If they don’t, the answer is groups can be different on features that offer no reproductive advantage, which is kind of trivially true, because there are plenty of genetic group differences about things no one cares about.
Also plenty of gene flow between Africa and Europe. Too much sequencing data exists, in the last 4k years there were a number of intermixing events, probably a good data set to see what genes survived and provided advantage and what they do.
No, the Tibetan example doesn’t prove anything about gene flow; it shows the ease with which genetic variation can arise even between nearby groups with even modest differences in the fitness landscape.
And it’s not true that a fitness advantage to high IQ would cause it to spread far and wide. For one thing, the selective advantage of a trait can vary greatly from one region to another; and in the case of traits related to cognition the advantage arose fairly recently, e.g. genetic predisposition for literacy may not be very useful in pre-literate populations. If your premise is correct about any selective advantage rendering genetic diversity across the continents improbable, then we should observe no serious diversity in height, muscle mass, predisposition to genetic diseases with no associated upside, etc. And yet, we observe strong founder effects in populations like Ashkenazi Jews (e.g. Tay-Sachs) which has far greater gene flow with Europeans than people on other continents did.
If levels of gene flow are sufficient to homenize useful traits, why did the sickle cell anemia gene not spread beyond Subsaharan Africa to any of the many other regions where it would’ve been adaptive? I’m curious what math you’re (or someone else) is doing to conclude that the number of migrants per generation is high enough to preclude genetic differentiation relevant to positively selective traits.
Lactase persistence in Africa is only found in a few populations, not surprisingly those who herd animals that can be milked, which renders the geographic distribution “patchy”. While the Fulani have the same lactase persistence alleles as Europeans (indicating possible gene flow), other African populations have a number of different alleles that almost certainly have arisen independently.
Doesn’t the fact that we can tell Africans, Europeans and Asians apart visually basically disprove your theory that the gene flow across continents is sufficient to keep groups broadly similar?
Perhaps genes for appearance require a lot more interbreeding to prevent drift than genes for intelligence do. But that’s a much more complex point than your “this is not so hard people” comment implies.
It’s about Sex. Does looking asian or European make you have more surviving children? If not then even with gene flow lots of variation will exist between groups. A dark skinned population that moves north and has minimal inflow from a lighter skinned population will get lighter skinned, because it’s advantageous. If it confers no advantage, the groups will stay different unless they fully merge.
When you’re talking about skin tone, it’s mostly about sunburn (if you’re too pale for the environment) and rickets (if you’re too dark for the environment, and don’t happen to have an Eskimo-style diet that’s sufficiently rich in vitamin D). For indigenous populations, skin tone tracks incoming UV radiation surprisingly well.
Rothwell’s Flynn effect argument has been made many times before and is as unconvincing as ever. X having a significant affect on Y does not mean that Z does not have a significant affect on Y. In fact, the Flynn affect has seen almost everyone’s IQ increase in affected regions; it has not eliminated most of the variation in IQ. Nor is its failure to do so the result of there being ‘a long way to go’ as the effect has stopped and is now being reversed. Moreover, it seems most of the effect is observed by preschool or even infant IQ tests, so it’s unlikely access to education is the primary mechanism. I think improved nutrition and physical environment (e.g. reduced lead exposure, etc.) are the most plausible explanations. But such explanations don’t allow for much explanation of variation between individuals in the same society.
I always assumed the primary reason for the Flynn effect the last several generations is that society has figured out how to get more people to reach their potential IQs on standardize test. So we figured that basic food, health and early education solves a lot of these issues and the effect will decline as in the future.
Now, I do believe a big deal on IQ is doing well on standardize test in which I have seniors who improve the last several times they took the PSAT & SAT. (Yes I know IQ is not as standardized as SAT.) So that is probably the main reason why African-Americans and Hispanic-Americans have jumped up from 85 in 1980s to 93 -95 today.
Hmmm… I thought that the Flynn Effect was isolated to a small set of question types that represented advanced techniques that are learned rather than the generalized intelligence questions. If this is true, then it would only occur after pre-school. So Mark Z, Collin, and I have three very different ideas about the underlying causation. I’m guessing that the Flynn Effect is still a puzzle waiting for a solution. Maybe I should read the Wikipedia page.
I think the Flynn effect is mostly in non-G-associated tests/questions, which may be indicative of teaching people to do better on IQ tests, as some argue, but the effect seems to be detected pretty broadly, including (in fact slightly more so I believe) in populations that I’d expect aren’t being taught so much how to do better on IQ tests; and one thing mentioned in the wiki article is that it’s detected very early, which makes me doubt educational explanations.
Flynn can be divided into:
1) Nutrition (etc) increasing ‘g’
2) Increases in non-‘g’ IQ test questions.
#1 is real, but basically played out anywhere that isn’t a third world hellhole. Non-starvation African IQ probably isn’t as low as 60, but I’d be surprised if it wasn’t less then 85 (African American, which has 20% white admixture).
#2 is like a piece of trivia. Who really cares about that kind of IQ? It’s not the kind of thing we normally think of as “intelligence”. I.E. the kind of intelligence that helps you build nuclear reactors or otherwise make a first world society work.
1. Nutrition: shouldn’t the contribution of nutrition levelled off quicker? Maybe health improvements during development also fits in here.
2. non-G IQ: it was my understanding that this part of the Flynn Effect was exactly the kind of thing that helps build nuclear reactors; heuristics and other system-level understanding that comes from learning and experience.
1) Not sure the question, but in the first world for a long time now nutrition, even in the poorest parts, is good enough that it can’t be the cause of The Gap.
2) “The Flynn Effect” is a term that The Bell Curve came up with, so its not a finding that contradicts it.
Let’s take a simple example (quote below):
Flynn hypothesises that people nowadays have far more use for this sort of abstraction. For instance, he quotes a study in which rural peasants in Russia – people who lived lifestyles much more like those of our ancestors – were asked logical, IQ-style questions. The responses were interesting. If the questioner asked a modern, Western person “All bears are white where there is always snow; in Novaya Zemlya there is always snow; what color are the bears there?”, the respondent would probably complete the syllogism and work out that the Novaya Zemlya bears are white. But a typical answer was: “I have seen only black bears and I do not talk of what I have not seen.”
When the questioner pushed further, asking “But what do my words imply?”, the respondent said: “If a person has not been there he can not say anything on the basis of words. If a man was 60 or 80 and had seen a white bear there and told me about it, he could be believed.”
If that peasant grows up being drilled into him to respond to the questioner that bears are white, is he “smarter”. Is he better at abstract thinking? Is he more able to build a nuclear reactor?
It seems the answer is no. It also makes sense now why other IQ subtests that more directly test ‘g’ aren’t affected by this sort of nonsense.
My badly worded first question was meant to ask why nutrition effects, captured in pre-school IQ tests, are not more of a one-time increase instead of the steady multi-generational increase that we call the Flynn Effect. If it is true that the Flynn Effect has stopped then its was just the amount of time for these benefits to spread throughout the population. If the Flynn Effect is mostly nutritional, I would expect the impact to be mostly confined to the lowest part of the distribution.
I’m not sure what “The Gap” is you mention. I was focused on the causative factors of The Flynn Effect.
The quote you give seems to anecdotally support collin’s claim that populations are just getting better at answering IQ-style questions. I agree that this style of non-G improvement is irrelevant in terms of the greater good of society.
I don’t remember where I read about the Flynn Effect being isolated to a small subset of IQ questions that have a similar characteristic; learned cognitive techniques that are relatively advanced and not intuitive. At the time I thought, “ahhhh, that explains it”. In this comment thread, we are talking about three plausible mechanisms. The “cognitive toolkit” component I was talking about is also non-G and it is important to society. It can be thought of as better STEM-style problem solving. IQ was supposed to capture G so these types of questions should not have been included but since they were included in historical data it is best to keep them and quantify their contribution moving forward.
I wouldn’t assume that the improvements in nutrition wouldn’t have broad effects across the distribution, as improvements in quality, storage, and production of foods even out of season have done more than make food more affordable; many nutrients for example may have been nearly impossible to get much of the year no matter how rich you were 100 years ago. Just a thought.
If you actually made it 291 pages into a book filled with sentences like that, Arnold, you must be a very patient person.
The environmental origins of the massive two-standard-deviation increase in cognitive ability over the past 100 years are themselves powerful evidence that genes cannot account for more than a modest share of variation across individuals.
It’s likely that most of the (assumed true) 2-SD increase in “cognitive ability” was purely education for-the-test. IQ test taking is a skill that can be learned, and those who learn that skill get higher IQ scores.
J. Diamond’s New Guinea guide (from Guns, Germs, and Steel) makes him sure, like Rothwell, that IQ is not the main driver of the differences. I am sure there are high IQ folk in all regions. But once an “almost optimal” environment is achieved for two groups, it’s no surprise that group avg IQ differences remain.
It’s also pretty well established that married husband-wife pairs are the optimal form for raising children in all developed countries — no other raising environment has been shown to be better, and many are worse. I wonder if Rothwell would support “limiting” Harvard admissions to no more than 10% students from the top 10% income, no more than 40% in the top 40%, or some other way to “equalize” access to top colleges. Where it’s already too late to change cognitive ability much, tho not to late to get great contacts & signals.
Mitchell’s idea of “development” being different from genetics might be very true, and important, but is functionally the same with respect to “choosing one’s parents”. Unless he has suggestions on increasing the quality of the pre-birth development, his contribution seems to make it Genes + Dev (= Nature) vs Nurture, so that genetic studies which reduce the Genes side, don’t automatically increase the Nurture side. So, less hope for policy to improve cognition.
Jewish selection has seemed to work to raise Jewish IQ over the last 2-3 thousand years, tho it’s not fully clear even in that case what are the proportions of influence.
In all cases, by the time a child gets into school, their behavior has a big influence on what they actually learn. Yet pushing optimal individual behavior doesn’t seem to be the goal of most such studies.
Better prenatal care and education is something that you will sometimes hear progressives talk about, as well as reducing environmental toxins and air pollutants.
I’m confident that education policy is independent of IQ manipulation. One does not have to resort to IQ statistics to recognize the positive effect of increasing literacy and numeracy in a population. Even without the low-hanging fruit there is probably unexplored techniques for improving diligence/character. The question is whether modern education is about as efficient as it is going to get.
In the 1960s, American cars were full of defects. Making a low cost/high quality car seemed an impossibility. Then Toyota and Datsun (now Nissan) came along and showed that approaching things in a different way, it was indeed possible.
I have seen no comparable different way coming out of East Asia in education.
One small bit of evidence that American schools are about as good as they can be.
Better ed outcomes ARE possible, but opposed by Teacher Unions:
https://www.ocregister.com/2019/10/19/parents-stand-up-to-the-failing-education-establishment-and-win-will-swaim/
Rather than focus on Genes, or Genes+Development, vs Nurture, it would be better to focus more clearly on optimized Nurture. What ed systems promote the best outcomes.
And of course some focus on what “outcomes” is would be good, too. Like learning to read at grade level. Many studies seem to show that learning phonetics is better for learning to read than using the “whole word” model.
I’d be happy to see Reps pushing for more vouchers in more states, and especially for parent-guided schools.
This particular issue, the heritability of IQ, turns many otherwise reasonable, rational people into emotional, unreasonable, irrational people.
It’s not that they aren’t reading the same research, it’s that many people with reputations for rational objective scientific inquiry turn entirely irrational and emotional.