I have been a big fan of your work for many years now. Not many academics I like are posting on substack: I guess they are too busy publishing academic papers. So I was so excited when I found your substack! And you're communicating your work in a manner that is understandable to the general population, which is not the case with academic articles. Not only that, but this is like a small book and it was for free! I shall return to this article many times and use it as a base for much further reading. Thank you so much for taking the time to do this for us!
You are a wonderful writer, Peter. When it comes to exposition no one is better.
That said, I notice that the word "conquest" occurs nowhere in this article. Military conquest (as opposed to warfare in general) is a distinct human institutional innovation that no doubt could not get started until the agricultural revolution was widely established, making it (a) possible for armed groups to seize food storages and (b) impossible for weaker people to run away and live off the land (as was possible in hunter/gatherer societies). In other words, it became possible for some groups to physically subdue others and put them to work. We see a new kind of society based on class rule and agricultural servitude.
Which is another way of saying that we shouldn't assume that the first food surpluses just somehow magically appeared via some kind of voluntary process, as is often implied when scholars write about the rise of civilizations. Based on the absence of any evidence to the contrary, they were more likely compelled as a result of military conquest. And yet we almost never read about the central role that this kind of violence has played in the historical process: how it would inevitably lead, first, to the organization of political states; secondly, to the appearance of capital cities; and then, eventually, to the establishment of geographical empires.
Indeed, once conquest became widely known as a political possibility, it is not too much to say that history becomes little more than a story of warring states in a relentless competition for power—a competition that has continued right on up into modern times.
And yet when you go to any modern encyclopedia you will find no entry under the headings "conquest" or "military conquest" as a distinct human institution. Why is that?
Thank you for your kind words. I see conquest, and imperialism in general, as playing a negative role in cognitive evolution. An empire will eventually liquidate its own founding people and the mental qualities that made their success possible (see https://evoandproud.blogspot.com/2020/12/frank-salter-and-national-question.html).
In the case of the Roman Empire, imperialism led to a steady influx of foreign slaves, initially prisoners of war and then slaves purchased from outside the Empire. Since lower class niches were now permanently filled by slaves, this put an end to the "Clarkian cycle", i.e., the upper classes were no longer replacing the lower classes through higher fertility and lower mortality,
As for your point on food surpluses and the rise of the State, we see dominant families controlling the supply of stored food even in pre-State societies. So this was a cause, and not a consequence, of State power, which then made military conquest possible.
Great piece as usual. One thing puzzles me though. If cold climate had an effect on cognitive ability, why did the early hunter-gatherers in Europe have an IQ equivalent of 55 (3 SD lower than today), which is lower than that of Sub-Saharan Africans?
1. Please provide a reference for this estimate of 55 IQ points. I have never encountered it in the literature.
2. We don't have adequate samples of DNA from early European hunter-gatherers. Any estimate of Edu PGS from that period is unreliable.
3. European hunter-gatherers were not a uniform group. There were probably regional differences.
4. There is no accepted methodology for translating Edu PGS scores into IQ. The relationship between the two seems to be non-linear, especially at lower IQ values. In other words, educational attainment also reflects non-cognitive factors at lower levels of academic success (ability to sit still in a classroom, rule following, susceptibility to boredom and distraction, etc.)
In figure 3 of your post, Western Hunter-Gatherers (orange) have an "Intelligence" PGS 3 SD below that of modern Europeans, which would be equivalent to an IQ of 55.
So I guess then Western Hunter-Gatherers may not actually have had an IQ 3 SD lower than today, but probably lower by a small amount, say 1.5 SD or so?
BTW, I pointed out a problem with the rise in Mongolian PGS in a comment on your eastern Eurasia article if you don't mind checking. TL;DR you attributed the rise in Mongolian PGS to increase in northern Han admixture, but the ancient Mongolian PGS was already the same as that of northern Han, and then increased beyond that of northern Han.
Maybe. This is also a problem with IQ results at very low levels (less than 85). At such levels, you're probably dealing with a population that has low thresholds for distraction and boredom. This is particularly true for hunter-gatherers who need to forage over a large area and must therefore be continually on the lookout for potential food sources.
Farmer have higher thresholds for distraction and boredom because they need to stay put in one location and take care of their food source.
Off-topic, but any comments on this new report in the Daily Mail that "Britons had black skin" until 3000 years ago? It makes this claim: "almost all Europeans had dark skin until around 3,000 years ago." Is this accurate? Sounds way too recent to me. I would have assumed light skin emerged much earlier, like 8,000-10,000 years ago.
"Almost all Europeans had dark skin until around 3,000 years ago."
What?? Europeans are portrayed in Egyptian tombs and on Minoan frescos that are dated to over 4,000 years ago. They’re shown as being pale-skinned.
Yes, ancient DNA indicates that dark skin prevailed until c. 7,000 years ago in western and southern Europe, perhaps as late as 5,000 to 4,000 years ago in parts of Britain (Brace et al., 2019; Lazaridis et al., 2014; Olalde et al., 2014). These relict populations may be recalled in mythology. An ancient Norse poem, the Rigsthula, describes how the god Rig created a class of black-haired, swarthy, and flat-nosed thralls (Jonassen, 1951; Karras, 1988).
But ancient DNA also shows that Europeans had pale skin much earlier in Scandinavia and the eastern Baltic (Günther, et al. 2018; Mittnik et al., 2018). Using inferential methods, three research teams have estimated the time when pale skin first appeared among Europeans: 19,200 to 7,600 years ago (Canfield et al., 2014); 19,000 to 11,000 years ago (Beleza et al., 2013); and 12,000 to 3,000 years ago (Norton & Hammer, 2007).
Western and southern Europeans became lighter-skinned at a later date, either through evolution in situ (especially through selection for lighter-skinned women) or through northeast Europeans expanding westward and southward, such as during the Yamnaya expansion that began some 5,000-6,000 years ago.
Yes, the reasons for having more children were not only economic but also cultural. It was considered good to have children — in and of itself.
But culture is influenced by economic priorities. Today, many people have the means to have large families, yet most don't (Elon Musk being a notable exception). Our culture no longer values large families, and even views them as a hindrance to the pursuit of other goals.
This was not the case in the late medieval and post-medieval period. For one thing, culture was much more constrained by the family economy. The family was the main unit of economic production, and it was economically impossible to move toward a child-free culture. That kind of cultural evolution was not yet possible.
An outstanding article. I'm wondering what your view is of the cognitive effects of the 14th century Black Death might have been. Phillippe Rushton once hypothesized that since the the poor were more likely to die from the plague which killed up to 50% of the population of NW Europe, this likely had a positive effect on average European cognitive abilities. I believe he even suggested that European IQ suddenly leapt ahead of China's as a result of this mass culling of the lower strata of the population. It's an interesting idea although I have no way to evaluate the claims. Aside from intelligence, it may have also had significant effects on social structure due to the sudden depopulation. Besides the increased value of labour, could this has led to a greater openness to non-kin and a greater willingness to cooperate?
I'm skeptical about claims that the Black Death selected for smarter people. First, natural selection has a greater impact through minor mortality over many generations than through major mortality over one generation. This is especially true in the case of recessive alleles.
Second, at the time of the Black Death, selection for cognitive ability was strongest among members of the middle class (merchants, yeoman farmers, independent artisans, etc.). This group tended to have more social interactions with strangers and thus were more exposed to the risk of infection. We do know that Jews suffered less mortality, apparently because they regularly washed their hands after social interactions. But was this true for Christian merchants?
It looks like Chinese IQ overtook European IQ long before the Black Death. The following graph is a bit rough, but it gives the general idea.
Thanks for the Piffer link. The historical dimension of intelligence is interesting. Of course, it is harder to measure. I can't find the Rushton reference just now, but I think he acknowledged Chinese IQ superiority well before the Black Death. He thought the Europeans only enjoyed a temporary advantage due to selection pressures after 1348 which eventually were lost. Most likely he would have projected Chinese dominance going forward due to its higher IQ compared with Europeans.
Your point is well-taken about the disease vulnerability about high social interaction professions.
I suspect that parts of Europe surpassed China in the 19th century (in terms of mean cognitive ability), but we need more data to answer that question.
If hunter gatherers entering farmer communities was partially the cause of their cognitive evolution, why is the WHG IQ PGS three standard deviations below the EEF one? Simplest explanation for the data -> hunter gatherer lifestyles cause little to no selection for intelligence, independent of climate. Cold-weather farming causes intense selection for intelligence. Thats why the curve was flat for ~50,000 years until farmers began to move into colder climes.
First, I wasn't referring to Western Hunter-Gatherers. I was referring to the hunter-gatherers of northern and eastern Europe, essentially the Great European Plain. This was where hunting distances were at a maximum and where hunters had to store, process, and manipulate large volumes of spatiotemporal data.
Second, we don't really have enough DNA from European hunter-gatherers to estimate their Edu or IQ PGS with confidence.
If you are referring to the EHGs with heavy ancient north eurasian admixture, there is little to no evidence of EHG admixture into farmer communities iirc. HG admixture in general is not high in Europe, and was especially not high in the earliest regions that displayed high PGSs for intelligence – Italy and Greece. You yourself acknowledge that the N European primacy in IQ is of a relatively recent vintage, long after the genetic composition of europeans had settled.
Amerindians, who were ancestrally exactly the sort of big game hunters you describe, are not noted for their intellectual prowess.
Then we'll have to agree to disagree. There is a fundamental difference of opinion between the two of us.
I lean toward estimates that place hunter-gatherer admixture in the range of 60 to 75% of the current European gene pool, the proportion being higher in the North and lower in the South. Such estimates are difficult to make for the reasons given in my essay: low estimates are based on the assumption that ALL genetic change across the hunter-gatherer/farmer time boundary is due to population replacement of hunter-gatherers by farmers. Some of the genetic change is, but some is not. The latter resembles farmer ancestry, but is actually due to:
- alleles that originate among farmers and then increase through natural selection, regardless of the degree of hunter-gatherer admixture
- convergent evolution (similar selection pressures produce similar results, even if the farmers are simply hunter-gatherers who have adopted farming).
- founder effects among those hunter-gatherers who join farming communities. These effects are random but a certain number will be "false positives" that wrongly indicate farmer ancestry.
I have only an undergraduate level understanding of the mathematics that goes into modeling drift vs. selection vs. admixture in ancient populations, so I will defer to you if you believe current models are not robust enough to differentiate the above factors and give us accurate models of HG ancestry in Europeans.
The low range of your estimate of 60% isn't actually too far off; northern europeans are basically 50% Globular Amphora and 50% Yamnaya; Globular Amphora farmers were themselves ~25% WHG in ancestry, while Yamnaya were half-EHG and half-CHG.
Perhaps the farmers were higher in verbal intelligence and the hunter gatherers in spatial intelligence? Might this explain Nordic AUTism and Med ARTism :-) ?
Alleles associated with autism have increased over time in Europeans (Piffer & Kirkegaard, 2024). So the higher incidence of autism in Western Europe may be due to the late medieval/post-medieval increase in cognitive ability, which seems to have begun earlier in England and then spread to Western Europe and then the rest of Europe.
One paragraph that really piqued my interest was 'Alleles associated with educational attainment." Especially these two sentences. 'It is constructed from the human genome, specifically from alleles associated with differences in educational attainment (EA) — a good proxy for cognitive ability. Such alleles have been identified at 1,271 loci in over one million people.' I was very surprised to see the relatively large number of loci where alleles associated with cognitive ability are found. What is the current number of alleles that are associated with cognitive ability? I realize that the calculated EA polygenic score is unreliable in predicting individual IQs, but it seems to be a good beginning toward that goal. This seems to me to be a great subject for a future article.
To date, 3,952 loci have been identified, specifically by Okbay et al., 2022. A locus can have more than one allele, so the number of alleles is much greater.
I should have mentioned Okbay et al. (2022), since I refer to that study elsewhere in my essay. I've corrected the paragraph to read as follows:
"In an initial study, such alleles were identified at 1,271 loci in over one million people. Using them, we can calculate an "EA polygenic score" that explains 11-13% of the differences in educational attainment among individuals (Lee et al., 2018). In a more recent study, EA-associated alleles were identified at 3,952 loci in about three million people. The resulting EA polygenic score explains 12-16% of the differences in educational attainment among individuals (Okbay et al., 2022)."
"To date, 3,952 loci have been identified, specifically by Okbay et al., 2022. A locus can have more than one allele, so the number of alleles is much greater."
I knew that cognitive ability was polygenic since intelligence is a spectrum trait. Thanks for the update. At least progress is being made.
The usual and surely most important effect of culturally-initiated natural selection is civilization-enhancing. Organismal biologists would likely term this process "niche construction." However, it is also conceivable that some cultural niche construction (sensu lato) works against civilization. Thuggees of the past and gangbangers of today likely discourage the kind of coherent civil life that would prevent them from plying their trades, i.e., a niche deconstruction effect that harms the larger society while the criminals enhance the social chaos in which they can thrive. The extent to which other groups that practice anti-social behaviors are deconstructive of the general society's niche while they construct their own niche is an open question.
This is why subcultures have historically been viewed with suspicion, if not worse. A society works best when there is a consensus on the rules to be followed, even though a different set of rules would work just as well.
A subculture, simply by existing, tends to delegitimize the majority culture. Either it feels that the majority's rules lack moral authority or the majority comes to doubt its moral authority through exposure to different rules. This problem is now arising all over the world through globalization. People are doubting the legitimacy of their own culture because they see a different set of rules (or an absence of rules) when they travel or turn on the TV.
Thanks for putting this together. I’m not sure that I’m fully sold on some of the details, such as the decline in cognitive ability during the Roman Empire, but you may be right. Two questions, if you don’t mind. First, what do you think about Lyman Stone’s recent attack on Greg Clark’s argument?
Second, do you think that the decline in genetic potential will continue or be reversed? I’ve read that, with the collapse in fertility including among the low income, the old gradients are beginning to re-emerge.
Lyman Stone has convinced himself that the graphs in Akbari et al. (2024) refute the comparison that Piffer & Kirkegaard (2024) made between medieval and contemporary genomes. Since Akbari et al. (2024) use a larger dataset (a little over 3 times larger), he sees this as an open-and-shut case.
There are several problems with his reasoning:
- A three-fold increase in sample size should not dramatically change the results, unless the initial dataset was biased in some way.
- Akbari et al. (2024) did not compare medieval genomes with contemporary genomes. Lyman Stone has simply made that comparison by looking at the lines on the graphs in Akbari et al. (2024). This is not a valid method. A line through a scatter plot will necessarily reduce the ups and downs in the actual data, on the assumption that these ups and downs are mostly noise. As a result, a genuine rise and fall will be reduced or even eliminated.
- For instance, we have good evidence from several studies that Edu PGS declined throughout the 20th century, but this decline is absent from the graphs of Akbari et al. (2024). It is simply treated as noise. As a result, the previous rise — including the late medieval/post-medieval surge in cognitive ability — is greatly reduced.
- Another example is the cognitive decline during the Imperial Era of Rome. It is visible on the graphs, but the magnitude has been reduced, with the result that the subsequent rise is likewise reduced.
I'm glad that Lyman Stone is interested in this question, but he repeatedly responds with anger and insults whenever I point out the flaws in his reasoning. Perhaps I'm wrong, but that still isn't an excuse for acting like a goon in a seedy bar.
The only solution, as I see it, is to go back to the raw data in Akbari et al. (2024) and redo the same comparison that Piffer and Kirkegaard (2024) did.
For the former: this looks like an academic tiff. Clark is more of a conceptual thinker than Stone who hews more closely to the verities of more abstract methods. For the latter, my understanding is in the USA it is only the very wealthiest who are having more than 2 kids. These are a tiny proportion of the population whose reproduction is way overwhelmed by relatively high birthrates among the poor.
The big difference is between the non-religious and the religious, especially those who try to insulate themselves from the surrounding culture, e.g., Amish, Mormons, Hassidim, etc.
NZ is a little different (low levels of formal religious practice) as TFR remained quite high in nonreligious communities.
NZ Maori in (sub)urban areas have a (somewhat) higher birthrate than (sub)urban NZ Caucasians who share essentially the same culture. Since the 1960's when Maori moved into the towns and cities there has been a great deal of intermarriage so that now describing yourself as Maori (for many) is a political act and having more kids is a political demographic move.
Up to about 2010 NZ TFR had stayed around 2.1 for 30 years. As a rule the poor have more kids when they receive government assistance and in rural areas. Also migrants from the Pacific Islands (mostly Polynesians but also Indians from Fiji and fewer Melanesians) tended to have relatively high TFR, though this has dropped in line with the drop in TFR of the general NZ population over the past 15 years.
We have an insular high TFR religious community of a few hundred in our (remote) South Island west coast, though it was started by a charismatic Pastor in the recent past ('60s or 1970s). It has been under systematic 'progressive' attack (by CPS etc) for about the past 15 years. Otherwise we have closed Brethren who tend to live among other Kiwis in small towns. Their TFR is about the same as the people they live among. The Polynesian communities are comparable.
Some comments regarding the correlation between brain size and latitude. While brain size clearly correlates with latitude, there appears to be strong evidence against the notion that this correlation is due to cognitive ability.
In the Smith & Beals (1990) data, Native Americans in Canada have almost the same cranial capacity as Europeans (~1356cc). The CC in equatorial South America, on the other hand, is much smaller; the Goajiro, Carib, Quechua, Botocudo have CCs of 1263-1350, average=1306. But once you go south from there, to southern South America (near Antarctica), the climate gets cold again and the CC once again increases significantly; the Araucanians, Yahgan and Ona, which are indigenous to southern South America, have CCs of 1363-1391, mean=1380. This is higher than the European CC of 1365 (the average of nine west/central European groups). In other words, CCs of Native Americans are largest in the cold northernmost and southernmost parts of the Americas, surpassing Europeans, and smallest in the warm equator region. In fact, the difference between Native Americans in equatorial and southern South America is 74cc, almost as large as the European-African difference of 83cc. Yet, I am unaware of any significant IQ differences between different groups of Native Americans. Apparently, holding IQ and race constant, brain size still differs as much as ever with distance from the equator.
This is also true within Northeast Asia; the Japanese and Chinese (1318cc-1418cc) have much smaller CCs than the more northern Mongols and Yakuts (1478cc-1489cc), despite the lower IQs of the latter.
Thus, couldn't the climate-brain size correlation be entirely due to something else, like lower surface area-to-volume ratio due to heat conservation?
References:
Smith, C. L., & Beals, K. L. (1990). Cultural Correlates with Cranial Capacity. American Anthropologist, 92(1), 193–200. doi:10.1525/aa.1990.92.1.02a00150
I agree that brain size correlates only in part with cognitive ability. The increase in brain size in Arctic environments probably reflects the need to store, process, and manipulate large volumes of spatiotemporal data for hunting over long distances.
Inuits do not score better than whites on IQ tests. In the article you linked, you state: "Inuit children in Arctic Quebec perform as well as white children from southern Quebec and even exceed U.S. norms." It appears you were referring to Wright et al. (1996). First of all, this is only one single study of Inuit IQ out of 18 others, and second of all, the problem with that study is that it did not correct for the Flynn effect and the American norms included a significant number of non-whites. Lynn looked at this study in his 2015 book, quote: "The authors claim that the Inuit children scored higher than Americans, but this is because American norms were depressed by the inclusion of ethnic minorities, [and] they made no allowance for the secular increase of scores." Correcting for the Flynn effect and comparing to American white/British norms, the actual IQ of this sample is 92. Lynn summarizes 18 IQ studies of Arctic Peoples in his book, all giving results in the range 74-96 with a median of 91. There's even a recent study from 2015 of 282 Inuits aged 8-14 in Canada tested with the (American) WISC-IV. They scored highest on Perceptual Reasoning (94), but adjusting for the Flynn effect and for the fact that whites score 3 points higher than the overall WISC-IV norms (due to inclusion of minorities) brings the IQ further down to 89. So I think it is safe to say that Inuits definitely don't score the same as Europeans, and even their EDU PGS seem to be lower according to Piffer's recent Eastern Eurasia study.
In any case, as you noted, Inuits do have higher IQ than Native Americans. In Lynn's book, the median of 23 studies give Native Americans an IQ of 86, lower than the Inuit IQ of 91. However, almost all studies of the Inuit are with non-verbal tests, whereas a lot of the Native American results are from verbal tests, which they are handicapped on since many do not speak English as first language. So the difference between Inuits and Native Americans is probably simply an artifact of inclusion of more verbal tests among the Native Americans. Strictly taking the nonverbal medians of the Native American and Inuit results respectively, they both end up with similar IQs around 90.
In fact, I'd guess most racial groups actually have around the same IQ of 90; when raised in Western societies, Southeast Asians, Middle Easterners & North Africans, Pacific Islanders, Inuits and Native Americans all obtain similar IQs of about 90. E.g. the 2nd generation Filipinos in Hawaii have an IQ around 90, and these immigrants were rather unselected unlike those who have later migrated to the mainland US. 2nd generation Turks & Moroccans in the Netherlands also have an IQ of 90. Native Hawaiians and Maoris also have an IQ of around 90. Same with Native Americans. So I guess the Inuits don't necessarily have a higher IQ than most other racial groups, despite occupying the harshest climate. On the other hand, maybe one could argue that their IQ is high considering that they never developed agriculture etc., unlike most other racial groups.
References
Jacobson, J. L., Muckle, G., Ayotte, P., Dewailly, É., & Jacobson, S. W. (2015). Relation of prenatal methylmercury exposure from environmental sources to childhood IQ. Environmental Health Perspectives, 123, 827-833.
Adjusting for the Flynn effect would actually increase Inuit IQ relative to European IQ, since the Inuit are closer to the beginning of the Flynn effect.
I agree that their IQ seems to be qualitatively different. The Inuit have an incredible ability to navigate over extensive territory and remember the spatial coordinates long after.
The mean age was 5.5 and the mean raw score was 14.06. Using British 2007 CPM norms gives them an IQ of 87. They were tested in ~1991 and the norms are from 2007, so a Flynn effect of 0.3 points per year raises the IQ upwards by 4.8 points to ~91 (the FE did continue in Britain among young children at least until 2007, though there's been no FE for those aged over 14). The same study also did show them scoring lower than Canadian norms, just as they scored lower than British norms.
Well, I'm not sure if the Flynn effect can expect to increase Inuit IQ relative to European IQ if that's what you mean. The Flynn effect appears to have been continuing at the same rate among Inuits as Europeans for the last 80 years; the Inuit-European gap hasn't changed from the first study in the 1930s to the most recent study from 2015.
It is true though that their visual memory is quite good and even better than Europeans. The same is true of Northeast Asians and Native Americans.
Let me quote from the two studies on Inuit in Canada:
"There were no significant differences between Inuit and White Ss in spatial, verbal-educational, or inductive reasoning abilities." https://psycnet.apa.org/doi/10.1037/h0081930
"Inuit children's CPM scores were consistently higher than age-appropriate U.S. norms and were comparable with data for White children in southern Quebec. In addition, the scores of children with two Inuit parents did not differ significantly from those of children with mixed Inuit/White heritage."
For Inuit in the Canadian Arctic, public education began to be established in the 1950s and was not fully available until the early 1960s. Given that the above studies were conducted in the mid-1970s and mid-1990s, these children were at most 20 to 40 years into the Flynn effect — in practice, even less because their parents were either illiterate or partially illiterate.
Adjusting for the Flynn effect should therefore increase Inuit IQ relative to European IQ.
On another note, I agree that the term "age-appropriate U.S. norms" doesn't mean a lot, but there was also a comparison with "White children in southern Quebec" as well as children of mixed ancestry.
The first study (Taylor & Skanes, 1976): This study showed Inuits on the Labrador coast scoring lower than one sample of whites and higher than two other samples of whites. However, it was a small sample of only 22 Inuits and the three white samples only numbered 14-22 each. A subsequent study by the same authors (Taylor & Skanes, 1977, https://doi.org/10.1037/h0081636) tested more Inuits, numbering 63 (presumably they expanded the initial sample by adding more subjects?). It also included three larger samples of whites consisting of "148 middle-class urban children, 173 low socio-economic urban children, and 175 White children living in coastal communities of Labrador", the last of which lived in similar communities as the Inuit. I've calculated the IQs (and corrected for the FE) as follows:
Inuit: 87
Coastal white: 78
Low SES white: 93
Middle-class white: 100
So the whites in the same community as Inuits (coastal whites) did in fact score lower than the Inuits. These may be a substandard sample of whites and thus not representative, but on the other hand this alone probably can't explain their very low IQ of 78. This could indeed suggest that Inuits obtain similar IQs to whites when raised in similar environments. However, another study, albeit smaller, tested 87 Inuits and 33 whites "attending school in integrated classes" in northern Canada, and Inuits scored 0.6d lower than whites on the Raven's, equivalent to an IQ of 91, exactly the same difference as when they aren't living in the same communities (MacArthur, 1969: https://doi.org/10.1037/h0082686)
The second study (Wright et al., 1996): The Canadian norms for white Quebec cited in this study don't go below age 6, whereas the three age groups of Inuits tested were aged 5, 5.5, and 6. Only those aged 6 could be compared to Canadian norms. They numbered no more than 14 but they scored 16.3, slightly lower than 17 for the Canadian norms. Britain on the other hand does have norms for those aged 5 and 5.5, and Inuits did in fact score lower than those. The average of the three age groups is equivalent to an IQ 91 on British norms. The authors did note that mixed-race Inuits didn't differ from full-blood Inuits, possibly suggesting that European admixture doesn't affect Inuit IQ, although the mixed-race ones only numbered around 25.
The Jacobson et al. (2015) study I cited earlier tested 282 11-year-olds from 14 Inuit villages around 2010. They would've been born around 1999, and their mothers (who were aged 24 at delivery) would have attended school in the 1980s. These children still obtained an IQ of 89 on the WISC-IV Perceptual Reasoning (and even lower on other indexes). At this point their environment wouldn't have been that bad, no? Yet their IQ is no better than in earlier studies.
It may be that the IQs of Inuits are in fact depressed in Inuit communities, but the evidence is mixed. Would be interesting with better data on Inuits and whites living in similar communities, e.g. Inuits and Danes in Greenland or in Denmark.
The speed of cognitive evolution — and human evolution in general — was much slower back then. Think of a logarithmic curve. More genetic change has occurred in our species over the past 10,000 years than over the previous 100,000. This was the finding of Hawks et al. (2007) in their study of the human genome. When hunting and gathering gave way to farming, the rate of genetic evolution increased more than a hundred-fold.
Hawks, J., Wang, E. T., Cochran, G. M., Harpending, H. C., & Moyzis, R. K. (2007). Recent acceleration of human adaptive evolution. Proceedings of the National Academy of Sciences, 104(52), 20753-20758. https://doi.org/10.1073/pnas.0707650104
Thanks for this summary of the state of current thinking. Forgive me for commenting having only read to the beginning of Roman Christianity at this time.
For me it highlights the importance of selection from within existing variation (as would be expected from the very short time period for genetic evolution to occur).
I am very interested in learning what light animal (fruit fly?) studies cast on your hypotheses on the interplay between genetic change (introduction of new genetics either through mutation or inward migration of different genetics) and phenotypic changes arising from selective pressures.
This is the "Baldwin effect." Daniel Dennett saw it as a frequent pathway of genetic evolution:
"Thanks to the Baldwin effect, species can be said to pretest the efficacy of particular different designs by phenotypic (individual) exploration of the space of nearby possibilities. If a particularly winning setting is thereby discovered, this discovery will create a new selection pressure: organisms that are closer in the adaptive landscape to that discovery will have a clear advantage over those more distant." https://doi.org/10.7551/mitpress/2871.001.0001
In other words, a species can create a new adaptive landscape and adapt to it initially through phenotypic plasticity. The new phenotype then becomes a template for genetic adaptation. This process has been notably demonstrated for the evolution of house finches. https://doi.org/10.1098/rstb.2008.0285
Yes. Inherent genetic potential (sh)could have considerable impact on cultural outcomes. Without going back to sources about horse domestication and the emergence of the Yamnaya people:
I recall the earliest known domestication of horses was some thousand (?) years earlier and far to the east. These people developed some technologies to process horse milk to reduce lactose levels (ie yoghurt) but ultimately disappeared.
Whereas the progenitor population of the Yamnaya may have had mutation(s) providing lactose tolerance which enabled the population to expand geographically with the Yamnaya ultimately developing a culture (with a suite of [cultural] technologies including the wheel) that facilitated exponential expansion in population and geographic range that both displaced other peoples and enabled them to successfully colonise new habitats.
Hi Peter,
I have been a big fan of your work for many years now. Not many academics I like are posting on substack: I guess they are too busy publishing academic papers. So I was so excited when I found your substack! And you're communicating your work in a manner that is understandable to the general population, which is not the case with academic articles. Not only that, but this is like a small book and it was for free! I shall return to this article many times and use it as a base for much further reading. Thank you so much for taking the time to do this for us!
Phil
Thanks for your kind words!
You are a wonderful writer, Peter. When it comes to exposition no one is better.
That said, I notice that the word "conquest" occurs nowhere in this article. Military conquest (as opposed to warfare in general) is a distinct human institutional innovation that no doubt could not get started until the agricultural revolution was widely established, making it (a) possible for armed groups to seize food storages and (b) impossible for weaker people to run away and live off the land (as was possible in hunter/gatherer societies). In other words, it became possible for some groups to physically subdue others and put them to work. We see a new kind of society based on class rule and agricultural servitude.
Which is another way of saying that we shouldn't assume that the first food surpluses just somehow magically appeared via some kind of voluntary process, as is often implied when scholars write about the rise of civilizations. Based on the absence of any evidence to the contrary, they were more likely compelled as a result of military conquest. And yet we almost never read about the central role that this kind of violence has played in the historical process: how it would inevitably lead, first, to the organization of political states; secondly, to the appearance of capital cities; and then, eventually, to the establishment of geographical empires.
Indeed, once conquest became widely known as a political possibility, it is not too much to say that history becomes little more than a story of warring states in a relentless competition for power—a competition that has continued right on up into modern times.
And yet when you go to any modern encyclopedia you will find no entry under the headings "conquest" or "military conquest" as a distinct human institution. Why is that?
I have a theory: https://shorturl.at/RhncR
Thank you for your kind words. I see conquest, and imperialism in general, as playing a negative role in cognitive evolution. An empire will eventually liquidate its own founding people and the mental qualities that made their success possible (see https://evoandproud.blogspot.com/2020/12/frank-salter-and-national-question.html).
In the case of the Roman Empire, imperialism led to a steady influx of foreign slaves, initially prisoners of war and then slaves purchased from outside the Empire. Since lower class niches were now permanently filled by slaves, this put an end to the "Clarkian cycle", i.e., the upper classes were no longer replacing the lower classes through higher fertility and lower mortality,
As for your point on food surpluses and the rise of the State, we see dominant families controlling the supply of stored food even in pre-State societies. So this was a cause, and not a consequence, of State power, which then made military conquest possible.
You might find this interesting: https://www.vectorsofmind.com/p/the-snake-cult-of-consciousness
Great piece as usual. One thing puzzles me though. If cold climate had an effect on cognitive ability, why did the early hunter-gatherers in Europe have an IQ equivalent of 55 (3 SD lower than today), which is lower than that of Sub-Saharan Africans?
1. Please provide a reference for this estimate of 55 IQ points. I have never encountered it in the literature.
2. We don't have adequate samples of DNA from early European hunter-gatherers. Any estimate of Edu PGS from that period is unreliable.
3. European hunter-gatherers were not a uniform group. There were probably regional differences.
4. There is no accepted methodology for translating Edu PGS scores into IQ. The relationship between the two seems to be non-linear, especially at lower IQ values. In other words, educational attainment also reflects non-cognitive factors at lower levels of academic success (ability to sit still in a classroom, rule following, susceptibility to boredom and distraction, etc.)
In figure 3 of your post, Western Hunter-Gatherers (orange) have an "Intelligence" PGS 3 SD below that of modern Europeans, which would be equivalent to an IQ of 55.
Again, you're assuming a linear relationship between Edu PGS and IQ. I wouldn't make that assumption.
So I guess then Western Hunter-Gatherers may not actually have had an IQ 3 SD lower than today, but probably lower by a small amount, say 1.5 SD or so?
BTW, I pointed out a problem with the rise in Mongolian PGS in a comment on your eastern Eurasia article if you don't mind checking. TL;DR you attributed the rise in Mongolian PGS to increase in northern Han admixture, but the ancient Mongolian PGS was already the same as that of northern Han, and then increased beyond that of northern Han.
Maybe. This is also a problem with IQ results at very low levels (less than 85). At such levels, you're probably dealing with a population that has low thresholds for distraction and boredom. This is particularly true for hunter-gatherers who need to forage over a large area and must therefore be continually on the lookout for potential food sources.
Farmer have higher thresholds for distraction and boredom because they need to stay put in one location and take care of their food source.
I will comment at the other comment section.
Off-topic, but any comments on this new report in the Daily Mail that "Britons had black skin" until 3000 years ago? It makes this claim: "almost all Europeans had dark skin until around 3,000 years ago." Is this accurate? Sounds way too recent to me. I would have assumed light skin emerged much earlier, like 8,000-10,000 years ago.
https://www.dailymail.co.uk/news/article-14454011/Britons-black-study-stonehenge.html
"Almost all Europeans had dark skin until around 3,000 years ago."
What?? Europeans are portrayed in Egyptian tombs and on Minoan frescos that are dated to over 4,000 years ago. They’re shown as being pale-skinned.
Yes, ancient DNA indicates that dark skin prevailed until c. 7,000 years ago in western and southern Europe, perhaps as late as 5,000 to 4,000 years ago in parts of Britain (Brace et al., 2019; Lazaridis et al., 2014; Olalde et al., 2014). These relict populations may be recalled in mythology. An ancient Norse poem, the Rigsthula, describes how the god Rig created a class of black-haired, swarthy, and flat-nosed thralls (Jonassen, 1951; Karras, 1988).
But ancient DNA also shows that Europeans had pale skin much earlier in Scandinavia and the eastern Baltic (Günther, et al. 2018; Mittnik et al., 2018). Using inferential methods, three research teams have estimated the time when pale skin first appeared among Europeans: 19,200 to 7,600 years ago (Canfield et al., 2014); 19,000 to 11,000 years ago (Beleza et al., 2013); and 12,000 to 3,000 years ago (Norton & Hammer, 2007).
Western and southern Europeans became lighter-skinned at a later date, either through evolution in situ (especially through selection for lighter-skinned women) or through northeast Europeans expanding westward and southward, such as during the Yamnaya expansion that began some 5,000-6,000 years ago.
For more information: https://www.anthro1.net/p/a-people-of-many-colors
Great article, as usual.
I have a question: you and Clark say that cottage industrialist had many sons FOR raising their workforce.
Why you think the process was teleological instead of casual, i.e. rich cottage industrialists had more sons because they could feed them?
Yes, the reasons for having more children were not only economic but also cultural. It was considered good to have children — in and of itself.
But culture is influenced by economic priorities. Today, many people have the means to have large families, yet most don't (Elon Musk being a notable exception). Our culture no longer values large families, and even views them as a hindrance to the pursuit of other goals.
This was not the case in the late medieval and post-medieval period. For one thing, culture was much more constrained by the family economy. The family was the main unit of economic production, and it was economically impossible to move toward a child-free culture. That kind of cultural evolution was not yet possible.
An outstanding article. I'm wondering what your view is of the cognitive effects of the 14th century Black Death might have been. Phillippe Rushton once hypothesized that since the the poor were more likely to die from the plague which killed up to 50% of the population of NW Europe, this likely had a positive effect on average European cognitive abilities. I believe he even suggested that European IQ suddenly leapt ahead of China's as a result of this mass culling of the lower strata of the population. It's an interesting idea although I have no way to evaluate the claims. Aside from intelligence, it may have also had significant effects on social structure due to the sudden depopulation. Besides the increased value of labour, could this has led to a greater openness to non-kin and a greater willingness to cooperate?
I'm skeptical about claims that the Black Death selected for smarter people. First, natural selection has a greater impact through minor mortality over many generations than through major mortality over one generation. This is especially true in the case of recessive alleles.
Second, at the time of the Black Death, selection for cognitive ability was strongest among members of the middle class (merchants, yeoman farmers, independent artisans, etc.). This group tended to have more social interactions with strangers and thus were more exposed to the risk of infection. We do know that Jews suffered less mortality, apparently because they regularly washed their hands after social interactions. But was this true for Christian merchants?
It looks like Chinese IQ overtook European IQ long before the Black Death. The following graph is a bit rough, but it gives the general idea.
https://x.com/DavidePiffer/status/1886032099779502459
Thanks for the Piffer link. The historical dimension of intelligence is interesting. Of course, it is harder to measure. I can't find the Rushton reference just now, but I think he acknowledged Chinese IQ superiority well before the Black Death. He thought the Europeans only enjoyed a temporary advantage due to selection pressures after 1348 which eventually were lost. Most likely he would have projected Chinese dominance going forward due to its higher IQ compared with Europeans.
Your point is well-taken about the disease vulnerability about high social interaction professions.
I suspect that parts of Europe surpassed China in the 19th century (in terms of mean cognitive ability), but we need more data to answer that question.
If hunter gatherers entering farmer communities was partially the cause of their cognitive evolution, why is the WHG IQ PGS three standard deviations below the EEF one? Simplest explanation for the data -> hunter gatherer lifestyles cause little to no selection for intelligence, independent of climate. Cold-weather farming causes intense selection for intelligence. Thats why the curve was flat for ~50,000 years until farmers began to move into colder climes.
First, I wasn't referring to Western Hunter-Gatherers. I was referring to the hunter-gatherers of northern and eastern Europe, essentially the Great European Plain. This was where hunting distances were at a maximum and where hunters had to store, process, and manipulate large volumes of spatiotemporal data.
Second, we don't really have enough DNA from European hunter-gatherers to estimate their Edu or IQ PGS with confidence.
If you are referring to the EHGs with heavy ancient north eurasian admixture, there is little to no evidence of EHG admixture into farmer communities iirc. HG admixture in general is not high in Europe, and was especially not high in the earliest regions that displayed high PGSs for intelligence – Italy and Greece. You yourself acknowledge that the N European primacy in IQ is of a relatively recent vintage, long after the genetic composition of europeans had settled.
Amerindians, who were ancestrally exactly the sort of big game hunters you describe, are not noted for their intellectual prowess.
Then we'll have to agree to disagree. There is a fundamental difference of opinion between the two of us.
I lean toward estimates that place hunter-gatherer admixture in the range of 60 to 75% of the current European gene pool, the proportion being higher in the North and lower in the South. Such estimates are difficult to make for the reasons given in my essay: low estimates are based on the assumption that ALL genetic change across the hunter-gatherer/farmer time boundary is due to population replacement of hunter-gatherers by farmers. Some of the genetic change is, but some is not. The latter resembles farmer ancestry, but is actually due to:
- alleles that originate among farmers and then increase through natural selection, regardless of the degree of hunter-gatherer admixture
- convergent evolution (similar selection pressures produce similar results, even if the farmers are simply hunter-gatherers who have adopted farming).
- founder effects among those hunter-gatherers who join farming communities. These effects are random but a certain number will be "false positives" that wrongly indicate farmer ancestry.
I have only an undergraduate level understanding of the mathematics that goes into modeling drift vs. selection vs. admixture in ancient populations, so I will defer to you if you believe current models are not robust enough to differentiate the above factors and give us accurate models of HG ancestry in Europeans.
The low range of your estimate of 60% isn't actually too far off; northern europeans are basically 50% Globular Amphora and 50% Yamnaya; Globular Amphora farmers were themselves ~25% WHG in ancestry, while Yamnaya were half-EHG and half-CHG.
Perhaps the farmers were higher in verbal intelligence and the hunter gatherers in spatial intelligence? Might this explain Nordic AUTism and Med ARTism :-) ?
Alleles associated with autism have increased over time in Europeans (Piffer & Kirkegaard, 2024). So the higher incidence of autism in Western Europe may be due to the late medieval/post-medieval increase in cognitive ability, which seems to have begun earlier in England and then spread to Western Europe and then the rest of Europe.
One paragraph that really piqued my interest was 'Alleles associated with educational attainment." Especially these two sentences. 'It is constructed from the human genome, specifically from alleles associated with differences in educational attainment (EA) — a good proxy for cognitive ability. Such alleles have been identified at 1,271 loci in over one million people.' I was very surprised to see the relatively large number of loci where alleles associated with cognitive ability are found. What is the current number of alleles that are associated with cognitive ability? I realize that the calculated EA polygenic score is unreliable in predicting individual IQs, but it seems to be a good beginning toward that goal. This seems to me to be a great subject for a future article.
To date, 3,952 loci have been identified, specifically by Okbay et al., 2022. A locus can have more than one allele, so the number of alleles is much greater.
I should have mentioned Okbay et al. (2022), since I refer to that study elsewhere in my essay. I've corrected the paragraph to read as follows:
"In an initial study, such alleles were identified at 1,271 loci in over one million people. Using them, we can calculate an "EA polygenic score" that explains 11-13% of the differences in educational attainment among individuals (Lee et al., 2018). In a more recent study, EA-associated alleles were identified at 3,952 loci in about three million people. The resulting EA polygenic score explains 12-16% of the differences in educational attainment among individuals (Okbay et al., 2022)."
"To date, 3,952 loci have been identified, specifically by Okbay et al., 2022. A locus can have more than one allele, so the number of alleles is much greater."
I knew that cognitive ability was polygenic since intelligence is a spectrum trait. Thanks for the update. At least progress is being made.
The usual and surely most important effect of culturally-initiated natural selection is civilization-enhancing. Organismal biologists would likely term this process "niche construction." However, it is also conceivable that some cultural niche construction (sensu lato) works against civilization. Thuggees of the past and gangbangers of today likely discourage the kind of coherent civil life that would prevent them from plying their trades, i.e., a niche deconstruction effect that harms the larger society while the criminals enhance the social chaos in which they can thrive. The extent to which other groups that practice anti-social behaviors are deconstructive of the general society's niche while they construct their own niche is an open question.
This is why subcultures have historically been viewed with suspicion, if not worse. A society works best when there is a consensus on the rules to be followed, even though a different set of rules would work just as well.
A subculture, simply by existing, tends to delegitimize the majority culture. Either it feels that the majority's rules lack moral authority or the majority comes to doubt its moral authority through exposure to different rules. This problem is now arising all over the world through globalization. People are doubting the legitimacy of their own culture because they see a different set of rules (or an absence of rules) when they travel or turn on the TV.
Thanks for putting this together. I’m not sure that I’m fully sold on some of the details, such as the decline in cognitive ability during the Roman Empire, but you may be right. Two questions, if you don’t mind. First, what do you think about Lyman Stone’s recent attack on Greg Clark’s argument?
https://substack.com/home/post/p-157553408
Second, do you think that the decline in genetic potential will continue or be reversed? I’ve read that, with the collapse in fertility including among the low income, the old gradients are beginning to re-emerge.
https://ifstudies.org/blog/more-babies-for-the-rich-the-relationship-between-status-and-children-is-changing
Lyman Stone has convinced himself that the graphs in Akbari et al. (2024) refute the comparison that Piffer & Kirkegaard (2024) made between medieval and contemporary genomes. Since Akbari et al. (2024) use a larger dataset (a little over 3 times larger), he sees this as an open-and-shut case.
There are several problems with his reasoning:
- A three-fold increase in sample size should not dramatically change the results, unless the initial dataset was biased in some way.
- Akbari et al. (2024) did not compare medieval genomes with contemporary genomes. Lyman Stone has simply made that comparison by looking at the lines on the graphs in Akbari et al. (2024). This is not a valid method. A line through a scatter plot will necessarily reduce the ups and downs in the actual data, on the assumption that these ups and downs are mostly noise. As a result, a genuine rise and fall will be reduced or even eliminated.
- For instance, we have good evidence from several studies that Edu PGS declined throughout the 20th century, but this decline is absent from the graphs of Akbari et al. (2024). It is simply treated as noise. As a result, the previous rise — including the late medieval/post-medieval surge in cognitive ability — is greatly reduced.
- Another example is the cognitive decline during the Imperial Era of Rome. It is visible on the graphs, but the magnitude has been reduced, with the result that the subsequent rise is likewise reduced.
I'm glad that Lyman Stone is interested in this question, but he repeatedly responds with anger and insults whenever I point out the flaws in his reasoning. Perhaps I'm wrong, but that still isn't an excuse for acting like a goon in a seedy bar.
The only solution, as I see it, is to go back to the raw data in Akbari et al. (2024) and redo the same comparison that Piffer and Kirkegaard (2024) did.
For the former: this looks like an academic tiff. Clark is more of a conceptual thinker than Stone who hews more closely to the verities of more abstract methods. For the latter, my understanding is in the USA it is only the very wealthiest who are having more than 2 kids. These are a tiny proportion of the population whose reproduction is way overwhelmed by relatively high birthrates among the poor.
The big difference is between the non-religious and the religious, especially those who try to insulate themselves from the surrounding culture, e.g., Amish, Mormons, Hassidim, etc.
NZ is a little different (low levels of formal religious practice) as TFR remained quite high in nonreligious communities.
NZ Maori in (sub)urban areas have a (somewhat) higher birthrate than (sub)urban NZ Caucasians who share essentially the same culture. Since the 1960's when Maori moved into the towns and cities there has been a great deal of intermarriage so that now describing yourself as Maori (for many) is a political act and having more kids is a political demographic move.
Up to about 2010 NZ TFR had stayed around 2.1 for 30 years. As a rule the poor have more kids when they receive government assistance and in rural areas. Also migrants from the Pacific Islands (mostly Polynesians but also Indians from Fiji and fewer Melanesians) tended to have relatively high TFR, though this has dropped in line with the drop in TFR of the general NZ population over the past 15 years.
Are there no communities like the Amish in New Zealand? Here in Canada, I'm starting to see them in areas where they didn't exist before.
We have an insular high TFR religious community of a few hundred in our (remote) South Island west coast, though it was started by a charismatic Pastor in the recent past ('60s or 1970s). It has been under systematic 'progressive' attack (by CPS etc) for about the past 15 years. Otherwise we have closed Brethren who tend to live among other Kiwis in small towns. Their TFR is about the same as the people they live among. The Polynesian communities are comparable.
Some comments regarding the correlation between brain size and latitude. While brain size clearly correlates with latitude, there appears to be strong evidence against the notion that this correlation is due to cognitive ability.
In the Smith & Beals (1990) data, Native Americans in Canada have almost the same cranial capacity as Europeans (~1356cc). The CC in equatorial South America, on the other hand, is much smaller; the Goajiro, Carib, Quechua, Botocudo have CCs of 1263-1350, average=1306. But once you go south from there, to southern South America (near Antarctica), the climate gets cold again and the CC once again increases significantly; the Araucanians, Yahgan and Ona, which are indigenous to southern South America, have CCs of 1363-1391, mean=1380. This is higher than the European CC of 1365 (the average of nine west/central European groups). In other words, CCs of Native Americans are largest in the cold northernmost and southernmost parts of the Americas, surpassing Europeans, and smallest in the warm equator region. In fact, the difference between Native Americans in equatorial and southern South America is 74cc, almost as large as the European-African difference of 83cc. Yet, I am unaware of any significant IQ differences between different groups of Native Americans. Apparently, holding IQ and race constant, brain size still differs as much as ever with distance from the equator.
This is also true within Northeast Asia; the Japanese and Chinese (1318cc-1418cc) have much smaller CCs than the more northern Mongols and Yakuts (1478cc-1489cc), despite the lower IQs of the latter.
Thus, couldn't the climate-brain size correlation be entirely due to something else, like lower surface area-to-volume ratio due to heat conservation?
References:
Smith, C. L., & Beals, K. L. (1990). Cultural Correlates with Cranial Capacity. American Anthropologist, 92(1), 193–200. doi:10.1525/aa.1990.92.1.02a00150
Inuit have an IQ level the same as or close to that of Europeans. See my discussion on this at:
https://www.amren.com/features/2020/03/why-are-human-groups-so-different/
There is also anecdotal evidence that cognitive ability is higher among them than among indigenous groups farther south.
John Hawks examined geographic variation in brain size, most of which seems to have arisen during the Holocene. He concluded:
"The evolution of smaller brains in many recent human populations must have resulted from
selection upon brain size itself or on other features more highly correlated with brain size than are gross body dimensions."
Hawks, J. (2011). Selection for smaller brains in Holocene human evolution. arXiv:1102.5604 [q-bio.PE] https://arxiv.org/abs/1102.5604
I agree that brain size correlates only in part with cognitive ability. The increase in brain size in Arctic environments probably reflects the need to store, process, and manipulate large volumes of spatiotemporal data for hunting over long distances.
Inuits do not score better than whites on IQ tests. In the article you linked, you state: "Inuit children in Arctic Quebec perform as well as white children from southern Quebec and even exceed U.S. norms." It appears you were referring to Wright et al. (1996). First of all, this is only one single study of Inuit IQ out of 18 others, and second of all, the problem with that study is that it did not correct for the Flynn effect and the American norms included a significant number of non-whites. Lynn looked at this study in his 2015 book, quote: "The authors claim that the Inuit children scored higher than Americans, but this is because American norms were depressed by the inclusion of ethnic minorities, [and] they made no allowance for the secular increase of scores." Correcting for the Flynn effect and comparing to American white/British norms, the actual IQ of this sample is 92. Lynn summarizes 18 IQ studies of Arctic Peoples in his book, all giving results in the range 74-96 with a median of 91. There's even a recent study from 2015 of 282 Inuits aged 8-14 in Canada tested with the (American) WISC-IV. They scored highest on Perceptual Reasoning (94), but adjusting for the Flynn effect and for the fact that whites score 3 points higher than the overall WISC-IV norms (due to inclusion of minorities) brings the IQ further down to 89. So I think it is safe to say that Inuits definitely don't score the same as Europeans, and even their EDU PGS seem to be lower according to Piffer's recent Eastern Eurasia study.
In any case, as you noted, Inuits do have higher IQ than Native Americans. In Lynn's book, the median of 23 studies give Native Americans an IQ of 86, lower than the Inuit IQ of 91. However, almost all studies of the Inuit are with non-verbal tests, whereas a lot of the Native American results are from verbal tests, which they are handicapped on since many do not speak English as first language. So the difference between Inuits and Native Americans is probably simply an artifact of inclusion of more verbal tests among the Native Americans. Strictly taking the nonverbal medians of the Native American and Inuit results respectively, they both end up with similar IQs around 90.
In fact, I'd guess most racial groups actually have around the same IQ of 90; when raised in Western societies, Southeast Asians, Middle Easterners & North Africans, Pacific Islanders, Inuits and Native Americans all obtain similar IQs of about 90. E.g. the 2nd generation Filipinos in Hawaii have an IQ around 90, and these immigrants were rather unselected unlike those who have later migrated to the mainland US. 2nd generation Turks & Moroccans in the Netherlands also have an IQ of 90. Native Hawaiians and Maoris also have an IQ of around 90. Same with Native Americans. So I guess the Inuits don't necessarily have a higher IQ than most other racial groups, despite occupying the harshest climate. On the other hand, maybe one could argue that their IQ is high considering that they never developed agriculture etc., unlike most other racial groups.
References
Jacobson, J. L., Muckle, G., Ayotte, P., Dewailly, É., & Jacobson, S. W. (2015). Relation of prenatal methylmercury exposure from environmental sources to childhood IQ. Environmental Health Perspectives, 123, 827-833.
Adjusting for the Flynn effect would actually increase Inuit IQ relative to European IQ, since the Inuit are closer to the beginning of the Flynn effect.
I agree that their IQ seems to be qualitatively different. The Inuit have an incredible ability to navigate over extensive territory and remember the spatial coordinates long after.
The mean age was 5.5 and the mean raw score was 14.06. Using British 2007 CPM norms gives them an IQ of 87. They were tested in ~1991 and the norms are from 2007, so a Flynn effect of 0.3 points per year raises the IQ upwards by 4.8 points to ~91 (the FE did continue in Britain among young children at least until 2007, though there's been no FE for those aged over 14). The same study also did show them scoring lower than Canadian norms, just as they scored lower than British norms.
Well, I'm not sure if the Flynn effect can expect to increase Inuit IQ relative to European IQ if that's what you mean. The Flynn effect appears to have been continuing at the same rate among Inuits as Europeans for the last 80 years; the Inuit-European gap hasn't changed from the first study in the 1930s to the most recent study from 2015.
It is true though that their visual memory is quite good and even better than Europeans. The same is true of Northeast Asians and Native Americans.
Let me quote from the two studies on Inuit in Canada:
"There were no significant differences between Inuit and White Ss in spatial, verbal-educational, or inductive reasoning abilities." https://psycnet.apa.org/doi/10.1037/h0081930
"Inuit children's CPM scores were consistently higher than age-appropriate U.S. norms and were comparable with data for White children in southern Quebec. In addition, the scores of children with two Inuit parents did not differ significantly from those of children with mixed Inuit/White heritage."
https://doi.org/10.1177/0022022196276006
For Inuit in the Canadian Arctic, public education began to be established in the 1950s and was not fully available until the early 1960s. Given that the above studies were conducted in the mid-1970s and mid-1990s, these children were at most 20 to 40 years into the Flynn effect — in practice, even less because their parents were either illiterate or partially illiterate.
Adjusting for the Flynn effect should therefore increase Inuit IQ relative to European IQ.
On another note, I agree that the term "age-appropriate U.S. norms" doesn't mean a lot, but there was also a comparison with "White children in southern Quebec" as well as children of mixed ancestry.
The first study (Taylor & Skanes, 1976): This study showed Inuits on the Labrador coast scoring lower than one sample of whites and higher than two other samples of whites. However, it was a small sample of only 22 Inuits and the three white samples only numbered 14-22 each. A subsequent study by the same authors (Taylor & Skanes, 1977, https://doi.org/10.1037/h0081636) tested more Inuits, numbering 63 (presumably they expanded the initial sample by adding more subjects?). It also included three larger samples of whites consisting of "148 middle-class urban children, 173 low socio-economic urban children, and 175 White children living in coastal communities of Labrador", the last of which lived in similar communities as the Inuit. I've calculated the IQs (and corrected for the FE) as follows:
Inuit: 87
Coastal white: 78
Low SES white: 93
Middle-class white: 100
So the whites in the same community as Inuits (coastal whites) did in fact score lower than the Inuits. These may be a substandard sample of whites and thus not representative, but on the other hand this alone probably can't explain their very low IQ of 78. This could indeed suggest that Inuits obtain similar IQs to whites when raised in similar environments. However, another study, albeit smaller, tested 87 Inuits and 33 whites "attending school in integrated classes" in northern Canada, and Inuits scored 0.6d lower than whites on the Raven's, equivalent to an IQ of 91, exactly the same difference as when they aren't living in the same communities (MacArthur, 1969: https://doi.org/10.1037/h0082686)
The second study (Wright et al., 1996): The Canadian norms for white Quebec cited in this study don't go below age 6, whereas the three age groups of Inuits tested were aged 5, 5.5, and 6. Only those aged 6 could be compared to Canadian norms. They numbered no more than 14 but they scored 16.3, slightly lower than 17 for the Canadian norms. Britain on the other hand does have norms for those aged 5 and 5.5, and Inuits did in fact score lower than those. The average of the three age groups is equivalent to an IQ 91 on British norms. The authors did note that mixed-race Inuits didn't differ from full-blood Inuits, possibly suggesting that European admixture doesn't affect Inuit IQ, although the mixed-race ones only numbered around 25.
The Jacobson et al. (2015) study I cited earlier tested 282 11-year-olds from 14 Inuit villages around 2010. They would've been born around 1999, and their mothers (who were aged 24 at delivery) would have attended school in the 1980s. These children still obtained an IQ of 89 on the WISC-IV Perceptual Reasoning (and even lower on other indexes). At this point their environment wouldn't have been that bad, no? Yet their IQ is no better than in earlier studies.
It may be that the IQs of Inuits are in fact depressed in Inuit communities, but the evidence is mixed. Would be interesting with better data on Inuits and whites living in similar communities, e.g. Inuits and Danes in Greenland or in Denmark.
One thing that wasn't addressed was what prompted the increase in human cognitive ability over other hominids before 30,000 years ago.
The speed of cognitive evolution — and human evolution in general — was much slower back then. Think of a logarithmic curve. More genetic change has occurred in our species over the past 10,000 years than over the previous 100,000. This was the finding of Hawks et al. (2007) in their study of the human genome. When hunting and gathering gave way to farming, the rate of genetic evolution increased more than a hundred-fold.
Hawks, J., Wang, E. T., Cochran, G. M., Harpending, H. C., & Moyzis, R. K. (2007). Recent acceleration of human adaptive evolution. Proceedings of the National Academy of Sciences, 104(52), 20753-20758. https://doi.org/10.1073/pnas.0707650104
Thanks for that information.
Thanks for this summary of the state of current thinking. Forgive me for commenting having only read to the beginning of Roman Christianity at this time.
For me it highlights the importance of selection from within existing variation (as would be expected from the very short time period for genetic evolution to occur).
I am very interested in learning what light animal (fruit fly?) studies cast on your hypotheses on the interplay between genetic change (introduction of new genetics either through mutation or inward migration of different genetics) and phenotypic changes arising from selective pressures.
This is the "Baldwin effect." Daniel Dennett saw it as a frequent pathway of genetic evolution:
"Thanks to the Baldwin effect, species can be said to pretest the efficacy of particular different designs by phenotypic (individual) exploration of the space of nearby possibilities. If a particularly winning setting is thereby discovered, this discovery will create a new selection pressure: organisms that are closer in the adaptive landscape to that discovery will have a clear advantage over those more distant." https://doi.org/10.7551/mitpress/2871.001.0001
In other words, a species can create a new adaptive landscape and adapt to it initially through phenotypic plasticity. The new phenotype then becomes a template for genetic adaptation. This process has been notably demonstrated for the evolution of house finches. https://doi.org/10.1098/rstb.2008.0285
Yes. Inherent genetic potential (sh)could have considerable impact on cultural outcomes. Without going back to sources about horse domestication and the emergence of the Yamnaya people:
I recall the earliest known domestication of horses was some thousand (?) years earlier and far to the east. These people developed some technologies to process horse milk to reduce lactose levels (ie yoghurt) but ultimately disappeared.
Whereas the progenitor population of the Yamnaya may have had mutation(s) providing lactose tolerance which enabled the population to expand geographically with the Yamnaya ultimately developing a culture (with a suite of [cultural] technologies including the wheel) that facilitated exponential expansion in population and geographic range that both displaced other peoples and enabled them to successfully colonise new habitats.