Theft of fire by Prometheus, Christian Griepenkerl (1878)

This past week, Nature published a paper on the evolution of Europeans over the past ten thousand years — a span of time that has transformed them in many aspects of mind and body:

… in the past ten millennia, we find that many hundreds of alleles have been a affected by strong directional selection. We also document one-standard-deviation changes on the scale of modern variation in combinations of alleles that today predict complex traits. This includes decreases in predicted body fat and schizophrenia, and increases in measures of cognitive performance. (Akbari et al., 2026)

After initial release as a preprint in September 2024, the paper remained in peer review for more than a year and a half — much longer than the usual four to six months, and a sign of its controversial nature.

Its findings certainly challenge the dominant view that genetic evolution has changed Homo sapiens only in minor ways, with the major changes being in the realm of culture. Consequently, most peer-reviewers assume that human populations differ genetically only in simple traits, and not in complex ones like cognitive ability.

This assumption has been challenged by a minority of geneticists and anthropologists, beginning in the 1980s with L.L. Cavalli-Sforza, Robert Boyd, Peter Richerson, and Pierre van den Berghe. Over the last two decades, their place has been taken by John Hawks, Greg Cochran, and Henry Harpending.

Their argument can be summarized as follows:

• Humans adapt to their cultural environment just as they adapt to their natural environment

• In both cases, the adaptations can be genetic

• Because humans adapt to a faster-changing cultural environment, and not just a slow-changing natural environment, they undergo more genetic change, not less. They create their environment, and it recreates them … through natural selection.

This can be seen in the timeline of changes to the human genome: new adaptive alleles were gained at a rate more than a hundred times higher when hunting and gathering gave way to farming, and this faster rate continued well into recorded history (Cochran & Harpending, 2009; Hawks et al., 2007).

During the time of the earliest farmers, humans had already adapted to the full range of natural environments from the equator to the arctic. They were now adapting to the innovations that farming made possible: new foods, sedentism, larger communities, and more interaction with non-kin. There then followed a cascade of further changes: the rise of the state, the rule of law, the development of trade, the growth of literacy and numeracy, and much more.

This cascade of changes created new demands on mind and body that varied from one culture to another. Demands on the mind especially increased. It now became necessary to process a greater volume of information from a greater number of sources in a greater variety of ways. Those who could meet such demands had an advantage over those who couldn’t.

Ancient DNA, a tool for charting cognitive evolution

DNA has been retrieved from human remains at many archaeological sites, particularly over the last two decades. Meanwhile, researchers have discovered thousands of alleles that affect IQ or educational attainment (Okbay et al., 2022). By identifying which of them are present in a population, we can estimate its average cognitive ability. This measure — the “cognitive polygenic score” — shows a 90% correlation with the mean IQ of the population (Piffer, 2019).

If we have enough DNA from each period of history and prehistory, we can chart how cognitive ability has evolved over time. This sort of evolution has been charted mostly for Europeans, since more DNA has been collected from them. Specifically, there have been four studies:

Woodley et al. (2017)

The authors used a dataset of 99 genomes dating from 4,560 to 1,210 years ago, as well as 503 modern genomes. The genomes were not only from Europe but also from Central Asia. The authors calculated three types of polygenic score, using alleles linked to educational attainment.

The results show an upward trend over the past four to five thousand years. Given the small size of the dataset, it was impossible to measure changes in the speed or direction of cognitive evolution during that time period.

Kuijpers et al. (2022)

The authors used a dataset of 827 genomes dating from over 11,000 years ago to under 5,000 years ago, as well as 250 modern genomes. They calculated three types of polygenic score, using alleles linked to intelligence, fluid intelligence, or educational attainment.

The results show:

• no change in cognitive ability during the long period of hunting and gathering

• a sustained upward trend from the advent of farming to the dawn of history

• stagnation during Roman times

• a renewed upward trend during the last few centuries.

Higher levels of cognitive ability were favored by “the strong increase in social complexity resulting from the Neolithic revolution and the process of urbanization and occupational specialization.”

Piffer & Kirkegaard (2024)

The authors used a dataset of 2,625 genomes dating from the past 35,000 years. They calculated three types of polygenic score, using alleles linked to intelligence or educational attainment.

The results show:

• no change in cognitive ability between 35,000 and 15,000 years ago

• an upward trend beginning near the end of the last ice age and lasting to the present.

The stagnation during Roman times, observed in the previous study, is probably obscured here by the large number of genomes from non-Romanized regions of Europe. When the authors examined data from central Italy alone, they found that cognitive ability stopped rising near the end of the Roman Republic and then fell throughout the Imperial Era (Piffer et al., 2023; see also Frost, 2023).

Akbari et al. (2026)

This is the latest and most prestigious study. It was carried out by the Reich Lab at Harvard Medical School, using a dataset of 15,836 genomes dating from the past 10,000 years. The authors calculated three types of polygenic score, using alleles linked to intelligence, household income, or years of schooling.

The results show:

• a surge in cognitive ability between 9,000 and 7,000 years ago among the earliest farmers

• a slower upward trend, with a possible decline between 2,000 and 1,000 years ago.

The initial surge may have been driven by strong selection for cognitive ability among the earliest farmers. Or the cause may have been assimilation of hunter-gatherers, particularly those from the cognitively demanding environments of northern Europe (Frost, 2019). The cause is difficult to pinpoint because we have only small numbers of European hunter-gatherer genomes from that time period, with even fewer from the north.

A replicated finding presented as a new one

At first sight, the four studies seem consistent with normal science. The first one produced a new finding, and the later ones replicated it with more data and better methodology.

However, the latest study does not acknowledge the three earlier ones. Look for yourself. All three are absent from the 220 references. The latest study is presented as an entirely novel work.

In a private reply to Davide Piffer, Ali Akbari explained that the Piffer and Kirkegaard paper was not cited because it “lacked proper calibration, did not adequately deal with inflation or multiple testing, and therefore did not count as relevant prior evidence for directional selection” (Piffer, 2026). If these concerns were as crucial as Akbari makes them out to be, the prior studies would not have been so strikingly similar in their results. Clearly, these studies had some value. Is Akbari denying that they motivated the Reich Lab to go ahead and do its own study?

Actually, there is a good reason for not acknowledging the prior studies. If they had been acknowledged, Nature might have rejected the paper and forced the Reich Lab to resubmit to a lesser journal. Keep in mind that Davide Piffer, Emil Kirkegaard, and Michael A. Woodley of Menie are personae non gratae in much of academia due to their work on IQ differences. The mere mention of their names might have antagonized the reviewers and tipped the balance toward rejection.

As things turned out, the paper barely got accepted. This may also explain why it was kept out of the media spotlight … until now. David Reich wanted any controversy to be after publication, and not during peer review.

Of course, by taking credit for this scientific discovery, he and his research team are embracing the very thing they were distancing themselves from. If mean IQ can differ from one time period to another within the same population, it probably also differs between populations. Why not? Time and space are two sides of the same coin.

Think that over. Differences in mean IQ across space are due to differing trajectories of mean IQ over time.

It’s not sporting and it sucks, but …

Let’s put personal vanity aside and see the big picture. The Reich Lab paper has the potential to settle the “race and IQ” issue — first, by opening up debate and, second, by showing that IQ differences are frequent, expectable, and even ubiquitous.

Mean IQ can differ between populations that no one would consider to be “races.” It can differ even between two time periods of the same population. Such ubiquity arises because natural selection acts on cognitive ability in a multitude of ways and for a multitude of reasons.

All of this seems obvious to me and is likely obvious to Ali Akbari and David Reich. But it is far from obvious to most people, including many proponents of cancel culture. For this, we should be thankful.

References

Akbari, A., Perry, A., Barton, A.R. et al. (2026). Ancient DNA reveals pervasive directional selection across West Eurasia. Nature (2026). https://doi.org/10.1038/s41586-026-10358-1

Cochran, G. & Harpending, H. (2009). The 10,000 Year Explosion: How Civilization Accelerated Human Evolution. Basic Books: New York. https://www.amazon.com/000-Year-Explosion-Civilization-Accelerated/dp/0465020429

Frost, P. (2019). The Original Industrial Revolution. Did Cold Winters Select for Cognitive Ability? Psych, 1(1), 166-181. https://doi.org/10.3390/psych1010012

Frost, P. (2023). Was the Roman Empire eugenic? Peter Frost’s Newsletter, December 11.

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 (USA), 104, 20753-20758. https://doi.org/10.1073/pnas.0707650104

Kuijpers, Y., Domínguez-Andrés, J., Bakker, O.B., Gupta, M.K., Grasshoff, M., Xu, C.J., Joosten, L.A.B., Bertranpetit, J., Netea, M.G., & Li, Y. (2022). Evolutionary Trajectories of Complex Traits in European Populations of Modern Humans. Frontiers in Genetics, 13, 833190. https://doi.org/10.3389/fgene.2022.833190

Okbay, A., Wu, Y., Wang, N., Jayashankar, H., Bennett, M., Nehzati, S.M., Sidorenko, J., Kweon, H., Goldman, G., Gjorgjieva, T., Jiang, Y., Hicks, B., Tian, C., Hinds, D.A., Ahlskog, R., Magnusson, P.K.E., Oskarsson, S., Hayward, C., Campbell, A., Porteous, D.J., … Young, A. I. (2022). Polygenic prediction of educational attainment within and between families from genome-wide association analyses in 3 million individuals. Nature Genetics, 54, 437–449. https://doi.org/10.1038/s41588-022-01016-z

Piffer, D. (2019). Evidence for Recent Polygenic Selection on Educational Attainment and Intelligence Inferred from Gwas Hits: A Replication of Previous Findings Using Recent Data. Psych, 1(1), 55-75. https://doi.org/10.3390/psych1010005

Piffer, D. (2026). What Akbari’s Reply Gets Wrong About Science. PifferPilfer, April 16.

Piffer D, Dutton, E., & Kirkegaard, E.O.W. (2023). Intelligence Trends in Ancient Rome: The Rise and Fall of Roman Polygenic Scores. OpenPsych. Published online July 21, 2023. https://doi.org/10.26775/OP.2023.07.21

Piffer, D., & Kirkegaard, E. O. (2024). Evolutionary trends of polygenic scores in European populations from the Paleolithic to modern times. Twin Research and Human Genetics, 27(1), 30-49. https://doi.org/10.1017/thg.2024.8

Woodley, M. A., Younuskunju, S., Balan, B., & Piffer, D. (2017). Holocene selection for variants associated with general cognitive ability: Comparing ancient and modern genomes. Twin Research and Human Genetics, 20(4), 271-280. https://doi.org/10.1017/thg.2017.37

At the deathbed, 1895. Edvard Munch (1863-1944), Kode Museum

Whites made up 96% of euthanized Canadians in 2024. Yes, euthanasia mostly involves seniors, and older Canadians are whiter. But the 65+ age bracket was only 86% white at the last census in 2021 and is less so today (Statistics Canada, 2025).

I found the relevant paragraph in the latest report on Medical Assistance in Dying (MAID):

A total of 15,927 of the 16,499 people who received MAID in 2024 responded to this question, the vast majority of whom (95.6%) identified as Caucasian (White). For context on how this compares to the overall population of Canada, approximately 70% of people in Canada identified as Caucasian in the most recent Census. (Health Canada, 2025, p. 32)

Why are Euro-Canadians “over-euthanized”? Let me answer by examining the approval process, the medical justifications, the place of death, and the numbers by province.

Who, whom?

Who is making whiteness a risk factor? Let’s begin with those whom the system euthanizes. Euro-Canadians may be over-represented in MAID deaths because they often live on their own after retirement with neither a spouse nor children to help. They may thus be less motivated to go on living, regardless of their medical condition.

What about those who approve requests for euthanasia? Two medical professionals must approve a MAID request, and no one else reviews their approval. Without review, their decisions are less constrained by personal biases that may or may not target Euro-Canadians. A physician may, for instance, be more willing to sit down with a requester of the same culture and religion to discuss alternatives to MAID. Otherwise, the request will be rubber-stamped — not out of animosity but because the physician feels less able to talk the requester out of it.

Talking the requester out of it also depends on the time available. The physician must choose between spending more time with the current patient or moving on to the next one. Primary care physicians see about a hundred patients a week and are always under pressure to keep their caseload manageable. Keep in mind that a wrongly euthanized person cannot lodge a complaint — the complaint must come from a friend or relative. This is where the individualism of Euro-Canadians works against them. If a physician has a heavy caseload, who is easier to “unload”? Someone with ties to many friends and relatives or someone with none?

When natural death is near versus when it isn’t

At what point in old age does whiteness become a risk factor? We can answer by examining the two tracks that a MAID request must follow. Track 1 is for recipients whose death is “reasonably foreseeable.” Track 2 is for those whose death is not “reasonably foreseeable.”

In 2024, whites were 94.8% of Track 1 deaths and 97.4% of Track 2 deaths (Health Canada, 2025, p. 32, fn. 39). Track 2 deaths also had a lower median age, 75.9, even though younger seniors are proportionately less white.

Euro-Canadians are thus over-euthanized for reasons that run counter to the natural processes of aging and illness. Whiteness matters less in a decision to hasten natural death and more in a decision to hasten death independently of medical justification.

MAID at home versus MAID in an institution

In what sort of environment does whiteness become a risk factor? A private home or an institution? We can answer with indirect evidence.

The latest MAID report provides a breakdown by ethnicity and province (Health Canada, 2025, Table C.5, p. 80). We previously saw a national total of 15,226 deaths among “Caucasians” (Health Canada, 2025, p. 32). The provincial and territorial subtotals add up to 14,215 deaths. The second total is 6.6% lower.

In death statistics, it isn’t unusual for a national total to exceed the total of all the regions. The discrepancy arises because the form filler is unsure what to write for the deceased’s address and leaves the field blank. Yes, the province of death can be inferred from other fields, and even from the form itself, but such inference requires manual intervention when the forms are processed.1

The address is almost always filled out for deaths at home (private residence or retirement home) but is often left blank for deaths at institutions (hospice, hospital, palliative care facility, residential care facility, correctional facility, or shelter). In an institution, it is harder to get information about the deceased, particularly the home address. A study of death certificates in Vermont concluded: “Certificates for deaths in hospitals were more likely to have major errors than certificates for deaths in a private residence … Certificates indicating deaths in a nonhospital facility were also more likely to have major errors than certificates indicating deaths in a private residence” (McGivern et al., 2017; see also Hahn et al., 2002).

This discrepancy reveals another. Whites are 95.6% of the national total but only 86.2% of the combined provincial/territorial total. The gap of 9.4 points is almost equal to the 9.6 points that separate the white share of MAID deaths from the white share of Canada’s senior population. Euro-Canadians are over-euthanized almost entirely in those cases where the address field is left blank — most likely in institutional cases. This is consistent with the higher percentage of white deaths in Track 2, as already noted. Track 2 also has a higher percentage of institutional deaths, with private residences accounting for only 81.2% of Track 2 deaths versus 88.4% of Track 1 deaths (Health Canada, 2025, p. 55).

For now, let’s use the provincial totals while remembering that they exclude most of the white over-representation in MAID deaths. This underreporting should, if anything, reduce the differences between provinces in white over-representation, making them seem smaller than they really are.

Whiter provinces versus less white ones

In which provinces does whiteness become a risk factor? Those where Euro-Canadians are proportionately fewer?

Canada’s population was about 97% of European descent in 1961, but this percentage has fallen with the rise of global immigration since the 1960s, particularly in British Columbia and Ontario — where most non-European immigrants have settled. Quebec and the Atlantic provinces have received less immigration and are now the whitest regions of Canada.

To answer the above question, we can compare the white percentage of MAID deaths in each province with the white percentage of the senior population:

The white percentages of the 65+ bracket were calculated from 2021 census data, using the formula: White percentage = 100% − (Visible minority % + Indigenous identity %).

Sources: Health Canada, 2025; Statistics Canada, 2026a; Statistics Canada, 2026b

In Ontario and British Columbia, Euro-Canadians are euthanized at a much higher rate than their share of the senior population. In Alberta, Quebec, and the Atlantic provinces, they are euthanized at the same rate or perhaps a lower one. (Note: white over-representation should be adjusted upwards due to the approximately one thousand white MAID deaths missing from the provincial totals).

The Atlantic provinces are poorer, less attractive to immigrants, and whiter. Their inhabitants are also more religious and more traditional. They prefer to stay in one place because they value support from relatives, longtime friends, the local church and the community. This support may explain why fewer of them choose euthanasia.

But what about Alberta? Its demographics are like Ontario’s: 66.5% white versus 62.9% white in 2021. It is also one of the provinces with the fewest locally born residents — only 52%. Yet it is also the province where whites are least likely to be euthanized compared to other groups. It isn’t because they are more religious. Self-identified Christians are less common in Alberta than in Canada as a whole, and other measures rank the province as average in religiosity (Cornelissen, 2021).

And what about Quebec? It is the province where whites are second-least likely to be euthanized compared to other groups. It is also the least religious province by any measure, with the lowest proportion of people attending religious activities at least once a month (Cornelissen, 2021).2

Alberta and Quebec may simply be less “woke.” They have not gone as far as other provinces in normalizing anti-white thinking, discourse, and behavior.

Takeaways

The more freedom the physician has to approve MAID requests, the whiter are the resulting deaths. This happens:

• when approval isn’t reviewed (all requests)

• when no medical justification is needed (Track 2)

• when the requester lives in an institution

• when the requester and the physician are not bound by the same culture and religion

• when the ideological environment permits anti-white bias

It would be politically easier to say that Euro-Canadians are over-euthanized because they want to be. They are less traditional than other groups, more solitary, and thus more open to euthanasia when old age creeps up on them. Yet this explanation doesn’t fit the data.

First, people most often prefer to be euthanized at home and as a means to hasten natural death. We see the opposite, however, with white over-representation in MAID deaths.

Second, MAID deaths are disproportionately white in institutional settings — so much so, that this one factor may explain most of the white over-representation. The “home turf” of a private residence or a retirement home seems to offer seniors more freedom to make up their mind — and not have it made for them.

Third, in whiter provinces, Euro-Canadian seniors are euthanized at the same rate as other seniors or even at a lower one. This might reflect a more conservative attitude toward euthanasia by traditional Euro-Canadians. But how does this explanation apply to Alberta and Quebec?

Alberta is the province where whites are least over-represented in MAID deaths and may even be underrepresented. Yet it is not much whiter than Ontario, and almost half its residents are born elsewhere. Nor is it very traditional, at least if we use religion as a metric.

Quebec is the province where whites are second-least over-represented in MAID deaths, and yet institutional resistance to euthanasia is almost absent there. In addition to having more unique MAID practitioners than all other provinces combined, it has the lowest percentage of hospitals that report no MAID deaths on their premises. Nor is there much resistance from organized religion.

However, Alberta and Quebec are similar in one respect: their estrangement from Canadian political culture. For different reasons, neither province feels wholly part of Canada, and neither feels bound by the country’s current political culture, including the belief that anti-white bias is normal and justified.

Is wokeness a factor?

We’re thus left with woke ideology: the more prevalent it is, the more euthanasia becomes disproportionately white. Wokeness has spread less easily into French Canada — due to the language barrier, the existence of separate institutions, and a general mistrust of Canadian political culture. A similar mistrust exists in Alberta.

Meanwhile, wokeness has spread into the rest of Canada much more easily — due to the shared language with the U.S., as well as shared institutions and a naïve acceptance of American mass culture.

Yet this is not the whole story. According to a recent report from the Macdonald-Laurier Institute, Canadians aren’t more woke than Americans; in fact, public opinion on cancel culture, critical race/history, and transgender issues is almost the same in Canada, the UK, and the US. But the same report does reveal two differences:

• Although younger people are considerably more woke in Canada than older people, the generation gap is smaller than in the UK or the US;

• Canadians have three times more trust in journalists than do people in the UK and over 50% more than do Americans (Kaufmann, 2024).

Both findings indicate a higher level of group conformity. In general, Canadians are less confrontational and more deferential, especially toward representatives of authority, such as government officials, university professors, church leaders, and journalists (Lipset, 1986). Once these authority figures have been ideologically captured, everyone else falls into line (Kaufmann, 2024, pp. 61-63). Canadians may feel unhappy about the woke revolution, but they generally keep their unhappiness to themselves.

There is also the fear of job loss and reputation loss: “55 percent of Canadians say they feel less free than they did 5 years ago to express their views on immigration; 61 percent say that the political climate prevents them from expressing their views as it might offend others; and 78 percent say political correctness has gone too far” (Kaufmann, 2024, pp. 38-39).

Finally, Canada’s elite is smaller, more centralized, and more interconnected (Brodie, 2001; Clement, 1975; Savoie, 1999). Because Americans have a larger, less centralized, and less interconnected elite, on top of being less deferential and more confrontational, they have had less trouble pushing back and building anti-woke institutions.

French Canadians and Albertans seem to be just as deferential and non-confrontational as other Canadians, but they have less trust in the federal government (Léger, 2025). They are thus more likely to challenge wokeness, which has become identified with federal policies. French Canadians also diverge from other Canadians on certain issues. While they lean further left on economic and foreign policy, in addition to being more republican and anticlerical, they are less woke than English Canadians on gender and racial diversity. In particular, they are more opposed to measures for racial or gender diversity at university (by 14 to 16 percentage points) and to transgender rights and flying the pride flag on public buildings (by 10 to 20 percentage points). In sum, French Canadians are generally further to the left, but their leftism is much more pre-woke (Kaufmann, 2024, pp. 36, 49).

To what extent does wokeness, in relation to other factors, explain white over-representation in MAID deaths? The question is difficult to answer because the different factors interact.

Take the fact that whites are not over-euthanized in Quebec. Is the main reason the lower prevalence of wokeness? Or is it the language barrier, the existence of separate institutions, or the lower percentage of foreign-educated physicians? There is no easy answer because all of these factors are entangled with each other. The French language has hindered the inflow of woke ideology while leading to the creation of separate institutions and a political culture that favors the training of local physicians, rather than the licensing of foreign ones.

On a final note, whatever is causing white over-representation in MAID deaths — woke ideology, contemporary Anglo culture, white individualism, lack of rapport between physician and requester — we may be seeing only the tip of the iceberg. Anti-white bias is probably just as present whenever a Canadian physician makes a life-or-death decision.

References

Brodie, I. (2001). Interest group litigation and the embedded state: Canada's court challenges program. Canadian Journal of Political Science/Revue canadienne de science politique, 34(2), 357-376. https://doi.org/10.1017/S0008423901777931

Clement, W. (1975). The Canadian Corporate Elite. McGill-Queen’s University Press. https://www.jstor.org/stable/j.ctt9qf2rw

Cornelissen, L. (2021). Religiosity in Canada and its evolution from 1985 to 2019. Insights on Canadian Society, https://www150.statcan.gc.ca/n1/pub/75-006-x/2021001/article/00010-eng.htm

Hahn, R. A., Wetterhall, S. F., Gay, G. A., Harshbarger, D. S., Burnett, C. A., Parrish, R. G., & Orend, R. J. (2002). The recording of demographic information on death certificates: a national survey of funeral directors. Public Health Reports, 117(1), 37. https://www.jstor.org/stable/4598716

Health Canada. (2025). Sixth Annual Report on Medical Assistance in Dying in Canada. https://www.canada.ca/en/health-canada/services/publications/health-system-services/annual-report-medical-assistance-dying-2024.html

Kaufmann, E. (2024). The Politics of the Culture Wars in Contemporary Canada. Macdonald-Laurier Institute, February. https://macdonaldlaurier.ca/wp-content/uploads/2024/02/20240129_Culture-wars-Kaufmann_PAPER-B-v2-FINAL.pdf

Léger (2025). Trust in Government and Views on Provincial Sovereignty, May 12. https://leger360.com/wp-content/uploads/2025/05/Special-report-May-20th.pdf

Lipset, S. M. (1986). Historical traditions and national characteristics: A comparative analysis of Canada and the United States. Canadian Journal of Sociology/Cahiers canadiens de sociologie, 113-155. https://doi.org/10.2307/3340795

McGivern, L., Shulman, L., Carney, J. K., Shapiro, S., & Bundock, E. (2017). Death certification errors and the effect on mortality statistics. Public Health Reports, 132(6), 669-675. https://doi.org/10.1177/0033354917736514

Savoie, D. J. (1999). Governing from the Centre: The Concentration of Power in Canadian Politics. University of Toronto Press. https://www.jstor.org/stable/10.3138/9781442675445

Statistics Canada. (2025). Sources of income of racialized individuals 65 years and over in Canada, 2020. Dorcas Hindir (author). Ethnicity, Language and Immigration Thematic Series, 89-657-X2025003. https://www150.statcan.gc.ca/n1/pub/89-657-x/89-657-x2025003-eng.htm

Statistics Canada (2026a). Indigenous identity population by gender and age: Canada, provinces and territories, census metropolitan areas and census agglomeration. Table: 98-10-0292-01. https://www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=9810029301

Statistics Canada (2026b). Visible minority by gender and age: Canada, provinces and territories. Table: 98-10-0351-01. https://www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=9810035101

Philosopher in meditation, Rembrandt (1606-1669)

At the height of the Middle Ages, most Western Europeans were still trapped in irrational ways of thinking — revealed truth, finalism, predestination, animism, and anthropomorphism. The stars, the moon, the sea, and the weather were all seen as conscious entities that could help or harm, depending on their mood. “Good luck” and “bad luck” were similarly viewed as intentional processes. There was no fine line between animate and inanimate, subjective and objective, or cause and effect (Oesterdiekhoff, 2012).

Some individuals did rise above this mindset, but they had few opportunities to meet others like themselves. Even as counsellors to monarchs, they were alone in the world, their full worth neither recognized nor appreciated, and their best ideas confined to unspoken thoughts. Intellectual progress remained slow without a critical mass of people who could understand new ideas, criticize them, improve on them, and put them into practice.

This situation began to change during the Late Middle Ages. A ferment was stirring among scholars, as shown by a renewed interest in ancient writings and the rise of scholasticism — the use of analytical argument to reconcile Christianity with pagan philosophy. “Men of reason” had more opportunities to come together. At first, they met in each other’s homes or at monastic schools; later, as they became more numerous, they gathered in more secular venues, like salons, coffeehouses, and learned societies. Such men would usher in the post-medieval era: a time of accelerating progress not only in the sciences but also in literature, music, and the arts (de Courson et al., 2023).

What caused this great advance to happen? For Gregory Clark, who has studied English demographic history, the cause was the growth of the middle class. As the market economy expanded from late medieval times onward, it created niches for young middle-class men and women, who could now marry earlier and have more children. But their higher rate of natural increase often overshot the number of niches created, forcing some children down the social ladder. Because the lower class was not reproducing itself, due to a higher death rate and lower marriage rate, the entire English population became increasingly of middle-class descent.

This demographic shift was paralleled by a mental and behavioral one: “Thrift, prudence, negotiation, and hard work were becoming values for communities that previously had been spendthrift, impulsive, violent, and leisure loving” (Clark, 2007, p. 166; see also Clark, 2009; Clark, 2023).

Were the mental and behavioral characteristics of the middle class due to good upbringing or good genes? Likely both. Middle-class parents were passing on both a certain morality and a certain genotype, including alleles associated with higher cognitive ability. This is what we see in the data.

A new research tool

Our understanding of recent human evolution has undergone a sea change with the retrieval of DNA from human remains. By examining the alleles of certain genes at different points of history and prehistory, we can observe how human populations have evolved—not only physically but also mentally and behaviorally.

We can infer the evolution of cognitive ability from alleles associated with IQ or educational attainment. Although this polygenic measure is unreliable for any one person, it can be reasonably accurate for a group (Piffer, 2025). In fact, it shows a high correlation with mean population IQ, in the range of 0.82 to 0.98 (Piffer, 2019).

To chart the cognitive evolution of Western Europeans, geneticists Davide Piffer and Emil Kirkegaard compared 467 genomes from two periods: Late Antiquity to the Middle Ages; and the present day. Their comparison showed a large rise in cognitive ability over time — between one third and one half of a standard deviation (Frost, 2024a; Piffer & Kirkegaard, 2024).

The actual rise may have been even larger because the comparison was between medieval and present-day genomes. According to Gregory Clark, cognitive ability was already rising during the Late Middle Ages, at least in England and probably elsewhere in Western Europe. At the other end of the timeline, cognitive evolution seems to have stalled during the Victorian Era and then reversed in the 20th century (Frost, 2022; Kong et al., 2017; Woodley et al., 2013).

To identify the actual timeline, Davide Piffer and Gregory Connor did a second study with 269 genomes from one region of England (Cambridge and surrounding area) between the 11th and 19th centuries (Piffer & Connor, 2025a). This study brings the trend of cognitive evolution into sharper focus: little change at first, followed by a steady rise from the 1300s onward. Particularly impressive is the increase in the “smart fraction”: the top 1% in 1850 was as smart as the top 0.1% in the year 1000.

Piffer and Connor then did a third study with a larger sample of 600 genomes and a broader geographic and temporal scope — Belgium, Denmark, England, Germany, and the Netherlands between the 8th and 19th centuries (Piffer & Connor, 2025b).

They again found the same timeline:

• No clear change until about 1350, followed by a steady rise until 1850. The total increase was a little over three quarters of a standard deviation.

• The inflection point seems to be the Black Death (1346-1353). This major pandemic may have kickstarted cognitive evolution in Western Europe by weakening feudalism and freeing up human capital for the emerging market economy: “labour became scarcer, wages rose, and land and capital became relatively cheaper. In that new environment, individuals who could exploit opportunities in trade, crafts, and skilled work had strong advantages” (Piffer, 2025b).

• On a per capita basis, the highly intelligent became over ten times more numerous in Western Europe between 700 and 1850. In other words, the top 1% in 1850 were smarter than the top 0.1% in 700 (Piffer & Connor, 2025b; Piffer, 2025b).

Cognitive evolution in Western Europe from medieval to modern times. Was it kickstarted by the Black Death? (Piffer, 2025b)

Discussion

These findings shed new light on Western European history, particularly the arbitrary line drawn between the Middle Ages and later times — typically at the year 1500. In reality, the entry into modernity was gradual. If a line must be drawn between pre-modernity and modernity, it should be set at around 1350, when the Black Death ravaged feudal Europe, wiping out between 30 and 60% of the population and facilitating the rise of a new economic system and dominant class.

Its aftermath saw a sustained expansion of the market economy. Young middle-class men and women could marry earlier and have more children, who could also provide their family business with cheap labor. Such families formed a growing proportion of the population. With each passing generation, there were more and more people who could exploit market opportunities through their skills in literacy, numeracy, and planning. The new economic context was favoring their success and, hence, the success of their moral values and genetic traits.

In sum, a changing economy was coevolving with a changing genome, with each pushing the other forward. And the spillover effects on intellectual life went far beyond bookkeeping and merchandising. As mean cognitive ability increased, so did the numbers of the highly intelligent. They were becoming a class of their own.

The last point is crucial. Progress is driven not only by exceptional individuals but also by exceptional communities that can appreciate the worth of new ideas and put them to use. Otherwise, new ideas are left to rot on the vine. For example, the printing press was not really invented by Johannes Gutenberg in 1440 — this was when it became commercially viable. It could have been invented earlier, and traders had already seen movable-type presses in East Asia, but there was no need to mass-produce books, at least not in Europe. Now there was a need.

The big picture

The great cognitive advance after 1350 is part of a bigger picture. If we combine the recent study of genomes from Western Europe with an earlier one of genomes from central Italy, we can tentatively propose the following timeline from the Roman Period onward:

27 BCE – 300 (Imperial Era) - Cognitive ability declined on average, primarily because the elite was failing to reproduce itself (Frost, 2024b; Piffer et al., 2023). Possible causes:

1. Decline of marriage, as described by Augustus in a speech to the equestrian class: “[W]hen the knights were very urgent, during the games, in seeking the repeal of the law regarding the unmarried and the childless, he assembled in one part of the Forum the unmarried men of their number, and in another those who were married, including those who also had children. Then, perceiving that the latter were much fewer in number than the former, he was filled with grief” (Cassius Dio. Roman History, 56).

2. Hypogamy between elite men and women of low status, often slaves or ex-slaves. Elite women thus became less important to elite fertility. If a slave owner had no children by a wife, he could leave his estate to the children he had presumably fathered with slave women or to emancipated boys and girls for whom he had developed affection (Rawson, 1986, 2014, pp. 173-179).

3. Increase in the slave population. Previously, elite offspring could, if need be, find niches further down the social ladder. In this way, the lower class (which had negative natural increase) was continually replenished by the demographic surplus of higher classes. Mean cognitive ability thus rose progressively from generation to generation. During the Imperial Era, however, there was less downward mobility: lower social niches were deemed unfit for higher-class offspring and filled normally by slaves or freedmen. (Harris, 1999; Saller, 1994).

300 – 800? (Late Antiquity) - The decline of mean cognitive ability reversed with the rise of Christianity. Possible causes:

1. Aggressive promotion of monogamy by the Church, which forced elite men to focus on procreation with a lawful wife, usually of similar status, rather than on sex with prostitutes or slave women. Socioeconomic success thus translated into reproductive success, and hence into selection for cognitive ability.

2. Shift from orthopraxy to orthodoxy, i.e., from correct conduct (paganism) to correct belief (Christianity). The Church favored those who were better at learning its rules, teachings, and doctrines.

800? – 1350 (Middle Ages to the Black Death) - Cognitive evolution now entered a period of either stasis or slow increase that cannot be measured with current data. Possible causes:

1. Rise of feudalism, from the 9th century onward. Feudal manors provided security in exchange for personal freedom. With fewer life choices, and less scope for innovation, serfs had fewer cognitive demands to deal with (Schooler, 1976).

2. Rise of monasticism, especially the Cluniac movement (founded in 910). Monks came disproportionately from aristocratic families, who provided endowments and appreciated the education provided by monastic living (Janin & Carlson, 2023, pp. 14, 39). Their vows of chastity kept them from passing on their mental aptitudes.

1350 – 1850 (late medieval and post-medieval periods) - Cognitive evolution accelerated after the Black Death and continued at a fast pace until the Victorian Era. Possible causes:

1. Growth of the middle class, due to the expanding market economy.

2. Gradual demographic replacement of the lower class, through the downward mobility of surplus middle-class individuals and the failure of the lower class to reproduce itself due to a higher death rate and lower marriage rate.

1850 – present (Modern Era) - Mean cognitive ability plateaued and eventually declined (Frost, 2022). Possible causes:

1. Shift from cottage industry to factory capitalism, where a business no longer relied primarily on family members to do the work, and where the workforce could now be expanded or contracted at will through hiring or firing. Because businesses were no longer “ma and pa” shops, the owners, generally men, had no economic incentive to marry early and have children. Reproductive success was thus severed from socioeconomic success

2. Creation of new needs to maintain a middle-class lifestyle, including the maintenance of a wife and family. Children became more costly and offered fewer economic benefits.

3. Decline of religion, especially the relaxation of its restrictions on marriage, divorce and sexual behavior.

The pre-medieval timeline, like the later one already discussed, sheds new light on the historical process. The Imperial Era now seems less impressive. Despite its grandeur, it coincided with an unsustainable running down of human capital, as often happens in empires. This decline then reversed during Late Antiquity, which now seems less tragic. Yes, the Western Empire collapsed, as did much of Roman culture, but this period also brought the rise of Christianity and a reorganization of society on a more sustainable basis.

Future research

Hopefully, we will see more studies of the sort reviewed here, particularly for late medieval and post-medieval times. By charting the cognitive advance by region and by century, we may learn more about the historical processes leading to the Renaissance, the Reformation, the Scientific Revolution, the Enlightenment, and the Industrial Revolution, all of which are arbitrary segments of a single continuum extending across space and time.

Here are a few avenues for future research:

Did mean cognitive ability begin to rise earlier in some regions than in others? While the Black Death clearly assisted the rise of the market economy, and the consequent expansion of the middle class, both trends were already under way in North Sea ports from the 7th century onward and in Italy after the year 1000. Did these regions experience an earlier rise in cognitive ability? In Italy, for instance, scholasticism was already reaching its high point in the 1200s and early 1300s, well before the Black Death.

Were there “hot spots” of unusually strong cognitive evolution? Such hot spots may have arisen wherever cottage industry predominated between the 15th and 19th centuries, particularly in Westphalia, Saxony, the Zurich uplands, Piedmont, Lombardy, Alsace, Flanders, Brittany, Lancashire, Yorkshire, and Ulster. In such areas, semi-rural cottagers enjoyed a population boom due to low mortality and high fertility. On the one hand, they lived away from urban areas, where disease and malnutrition kept population growth in check. On the other, they employed family members in their workshops and had no other means to expand their workforce.

These cottagers produced their wares on contract for urban merchants. They thus operated in an elastic, competitive market, with some doing better than others. Those who did better, had more children — generally by marrying earlier and helping their offspring do likewise (Seccombe, 1992, pp. 205, 217). This translation of economic success into reproductive success must have created strong selection for cognitive ability.

Was the rise in mean cognitive ability part of a broader mental and behavioral shift? Clark (2007, p. 166) believes that the English population underwent a broad shift toward “middle class” qualities of mind and behavior — not only higher cognitive ability but also lower time preference, greater impulse control, and less willingness to settle disputes through violence.

Oesterdiekhoff (2023) has argued for a similar shift across Western Europe, describing it in terms of Jean Piaget’s stages of cognitive development. More and more people could go beyond “preoperational thinking,” which manifests itself in egocentrism, anthropomorphism, finalism, and animism. They could now achieve “operational thinking” — the ability to understand probability, cause and effect, and another person’s perspective.

Should the genomic data be controlled for socioeconomic status? Studies of cognitive evolution may suffer from sampling bias because the physical remains of elite individuals are more likely to survive the passage of time. Elite DNA would thus be overrepresented in older samples, and cognitive ability would seem to decrease over time. Of course, we generally see the opposite: an increase in cognitive ability over time. The actual increase may therefore be even greater than current findings suggest.

In any case, we should certainly attempt to control for the SES of ancient DNA, which can be determined from parish records or inferred through isotope analysis of bones and teeth. High SES is associated with an ample, high-meat diet, and low SES with coarse grains and malnutrition.

References

Cassius Dio (1924) Roman History. Loeb Classical Library. Boston: Harvard University Press. https://penelope.uchicago.edu/Thayer/E/Roman/Texts/Cassius_Dio/56*.html

Clark, G. (2007). A Farewell to Alms. A Brief Economic History of the World. Princeton University Press: Princeton. https://press.princeton.edu/books/paperback/9780691141282/a-farewell-to-alms

Clark, G. (2009). The domestication of man: the social implications of Darwin. ArtefaCToS, 2, 64-80. https://www.researchgate.net/publication/277275046_The_Domestication_of_Man_The_Social_Implications_of_Darwin

Clark, G. (2023). The inheritance of social status: England, 1600 to 2022. Proceedings of the National Academy of Sciences, U.S.A., 120(27), e2300926120 https://doi.org/10.1073/pnas.2300926120

de Courson, B., Thouzeau, V., & Baumard, N. (2023). Quantifying the scientific revolution. Evolutionary Human Sciences, 5, E19. https://doi.org/10.1017/ehs.2023.6

Frost, P. (2022). The Great Decline. Peter Frost’s Newsletter, December 20.

Frost, P. (2024a). Cognitive evolution in Europe: Two new studies. Peter Frost’s Newsletter, March 14. https://www.anthro1.net/p/cognitive-evolution-in-europe-two

Frost, P. (2024b). How Christianity rebooted cognitive evolution, Aporia Magazine, October 10.

Harris, W. (1999). Demography, Geography and the Sources of Roman Slaves. Journal of Roman Studies 89: 62-75. https://doi.org/10.2307/300734

Janin, H., & Carlson, U. (2023). Medieval Monks and Monasteries. McFarland.

Kong, A., Frigge, M.L., Thorleifsson, G., Stefansson, H., Young, A.I., Zink, F., Jonsdottir, G.A., Okbay, A., Sulem, P., Masson, G., Gudbjartsson, D.F., Helgason, A., Bjornsdottir, G., Thorsteinsdottir, U., & Stefansson, K. (2017). Selection against variants in the genome associated with educational attainment. Proceedings of the National Academy of Sciences, 114(5), E727-E732. https://doi.org/10.1073/pnas.1612113114

Oesterdiekhoff, G.W. (2012). Was pre-modern man a child? The quintessence of the psychometric and developmental approaches. Intelligence 40, 470–478. https://doi.org/10.1016/j.intell.2012.05.005

Piffer, D. (2019). Evidence for Recent Polygenic Selection on Educational Attainment and Intelligence Inferred from Gwas Hits: A Replication of Previous Findings Using Recent Data. Psych, 1(1), 55-75. https://doi.org/10.3390/psych1010005

Piffer, D. (2025). Why individual prediction doesn’t scale to group differences in polygenic scores: debunking a common fallacy. PifferPilfer, November 21.

Piffer, D., & Connor, G. (2025a). Genomic Evidence for Clark’s Theory of the British Industrial Revolution, preprint, ResearchGate, June. https://www.researchgate.net/publication/392808200_Genomic_Evidence_for_Clark’s_Theory_of_the_British_Industrial_Revolution

Piffer, D., & Connor, G. (2025b). Genomic Evidence for Clark’s Theory of the Industrial Revolution, preprint, ResearchGate, November. https://www.researchgate.net/publication/392808200_Genomic_evidence_for_Clark’s_theory_of_the_Industrial_Revolution

Piffer D, Dutton, E., & Kirkegaard, E.O.W. (2023). Intelligence Trends in Ancient Rome: The Rise and Fall of Roman Polygenic Scores. OpenPsych, July 21, 2023. https://doi.org/10.26775/OP.2023.07.21

Piffer, D., & Kirkegaard, E.O.W. (2024). Evolutionary Trends of Polygenic Scores in European Populations from the Paleolithic to Modern Times. Twin Research and Human Genetics, 27(1), 30-49. https://doi.org/10.1017/thg.2024.8

Rawson, B. (1986). Children in the Roman familia. In: B. Rawson (ed.) The Family in Ancient Rome: New Perspectives, (pp. 170-200). Ithaca (New York): Cornell University Press.

Saller, R. P. (1994). Patriarchy, property and death in the Roman family (Vol. 25). Cambridge University Press.

Schooler, C. (1976). Serfdom’s legacy: An ethnic continuum. American Journal of Sociology, 81(6), 1265-1286. https://doi.org/10.1086/226223

Seccombe, W. (1992). A Millennium of Family Change. Feudalism to Capitalism in Northwestern Europe. London: Verso. https://archive.org/details/millenniumoffami0000secc/page/n3/mode/2up

Woodley, M.A., J. Nijenhuis, & Murphy, R. (2013). Were the Victorians cleverer than us? The decline in general intelligence estimated from a meta-analysis of the slowing of simple reaction time. Intelligence, 41, 843-850. https://doi.org/10.1016/j.intell.2013.04.006

Head of a chimpanzee, Edward Lear, 1835 - Humans and chimps are supposed to be 98-99% genetically identical. Why, then, are they so different?

You’ve probably heard that humans and chimpanzees are genetically 98 to 99% the same and human populations 99.9% the same. Both findings are often mentioned. In a speech to high school graduates, Hillary Clinton “cited genetic research that shows humans are 99.9 percent the same.”

"The differences in how we look — in our skin color, our eye color, our height — stem from just one-tenth of 1 percent of our genes. And the differences among us — our cultures, our religious beliefs, the music we like — it is all so small a distinction in our sea of common humanity," she said. (Hawks, 2007)

Of course, one tenth of one percent is still a lot. As anthropologist John Hawks pointed out, “one-tenth of 1 percent of 3 billion is a heck of a large number — 3 million nucleotide differences between two random genomes.”

He added:

We differ by one-tenth of 1 percent of nucleotides, this is enough to make coding differences in a large fraction of our genes. (Hawks, 2007).

In other words, the 0.1% difference doesn’t refer to genes. It refers to nucleotides. A single gene is a long chain of nucleotides, often a very long one, and a single nucleotide mutation can alter how the entire chain works. Conceivably, each and every gene could differ by 0.1% among human populations, and each and every one could work differently in different people.

Genes also differ in how their chains of nucleotides are arranged. The same chain may be repeated consecutively or copied and inserted somewhere else. Such rearrangements are more important than nucleotide changes in altering how a gene works: “Structural variations, such as copy-number variation and deletions, inversions, insertions and duplications, account for much more human genetic variation than single nucleotide diversity” (Wikipedia, 2025).

How, then, can we measure the genetic distance between two groups in terms of real functional differences? One way is to look at what genes make, namely proteins. Whereas humans and chimpanzees are 98-99% the same in their nucleotides, they are only 20% the same in their proteins (Glazko et al., 2005).

But the biggest functional difference is not so much in the proteins themselves as in the way they are used to build living tissues, notably in the brain (Yoo et al., 2025). Living tissue is created by a tiny minority of genes that regulate how other genes are expressed, often thousands of others. Although these “regulator genes” are much fewer in number than other genes, they are far greater in their effects, particularly on growth and development.

The source of a legend

A half-century ago, two researchers — Mary-Claire King and Allan Wilson — discovered that humans and chimpanzees differ the most in “regulator genes.” They wrote up their finding in a landmark paper, and it is here that we first see the estimates of 98-99% and 99.9% sameness. The two authors argued that the higher primates had evolved mainly through changes at a tiny fraction of all genes:

The molecular similarity between chimpanzees and humans is extraordinary because they differ far more than sibling species in anatomy and way of life. … Because of these major differences in anatomy and way of life, biologists place the two species not just in separate genera but in separate families.

… small differences in the time of activation or in the level of activity of a single gene could in principle influence considerably the systems controlling embryonic development. The organismal differences between chimpanzees and humans would then result chiefly from genetic changes in a few regulatory systems, while amino acid substitutions in general would rarely be a key factor in major adaptive shifts. (King & Wilson, 1975, pp. 113-114)

The two researchers were thinking not only about the human-chimpanzee difference but also about the differences among human populations:

This [human-chimpanzee] distance is 25 to 60 times greater than the genetic distance between human races. In fact, the genetic distance between Caucasian, Black African, and Japanese populations is less than or equal to that between morphologically and behaviorally identical populations of other species. (King & Wilson, 1975, p. 113)

The above paragraph is within a longer discussion about the paradoxical smallness of the human-chimpanzee genetic distance. In fact, the paradox is spelled out in the next paragraph:

However, with respect to genetic distances between species, the human-chimpanzee D value is extraordinarily small, corresponding to the genetic distance between sibling species of Drosophila or mammals. Nonsibling species within a genus … generally differ more from each other, by electrophoretic criteria, than humans and chimpanzees. The genetic distances among species from different genera are considerably larger than the human-chimpanzee genetic distance. (King & Wilson, 1975, p. 113)

Genetic distance between humans and chimpanzees, compared to genetic distances in other taxa. (King & Wilson, 1975, p. 113)

So, again, how can we measure the genetic distance between two groups in terms of real functional differences? In the case of human populations, there is no easy answer because few species have had a similar evolutionary history. We evolved rapidly at the very time we were splitting up and adapting to different environments — not only natural environments from the equator to the arctic but also an ever-wider range of cultural environments. In fact, this acceleration of human adaptation to new environments largely explains the concurrent acceleration of human evolution (Akbari et al., 2024; Cochran & Harpending, 2009; Frost, 2023a; Hawks et al., 2007; Kuijpers, et al., 2022; Libedinsky et al., 2025; Piffer & Kirkegaard, 2024; Rinaldi, 2017).

Because natural selection has shaped our species in highly divergent ways, it has contributed much more to diversity between populations than to diversity within them. The second kind of diversity is due mostly to stochastic processes of little adaptive consequence, since everyone is adapting to the same environment and the same selection pressures. Within a population, natural selection creates diversity only in a few cases, essentially through frequency-dependent selection and disruptive selection.

We come again to the same paradox. If we calculate the ratio of between-population diversity to total diversity, i.e., the Fst, we get a low ratio even though human populations differ much more anatomically than do most sibling species in the animal kingdom.

As Charles Darwin noted, a naturalist would consider some human groups to be “as good species as many to which he had been in the habit of affixing specific names.” This is true because genetic diversity is much more consequential between human populations than within them. Fst is therefore an apples-to-oranges comparison (Darwin, 1936 [1888], pp. 530-531; Frost, 2023b).

Discussion

I don’t blame Hillary Clinton for drawing the wrong conclusion from the 99.9% estimate, but I’m less forgiving toward those who have silently gone along with it while knowing better. John Hawks knew better and did not silently go along when he criticized Hillary’s speech in one of his posts. That post remained on his website until he deleted it in 2021 — after he apparently got the memo that Hillary was right.

“Nice research lab you have there. Pity if anything happened to it.”

When academics choose the path of silence, and withhold their objections, they help create a fake consensus that ultimately brings academia into disrepute.

References

Akbari, A., Barton, A.R., Gazal, S., Li, Z., Kariminejad, M., Perry, A., Zeng, Y., Mittnik, A., Patterson, N., Mah, M., Zhou, X., Price, A.L., Lander, E.S., Pinhasi, R., Rohland, N., Mallick, S., & Reich, D. (2024). Pervasive findings of directional selection realize the promise of ancient DNA to elucidate human adaptation. bioRxiv. https://doi.org/10.1101/2024.09.14.613021

Anon. (2007). Finding said to show "race isn't real" scrapped http://www.world-science.net/othernews/070904_human-variation.htm

Cochran, G. & Harpending, H. (2009). The 10,000 Year Explosion: How Civilization Accelerated Human Evolution. Basic Books: New York. https://www.amazon.ca/000-Year-Explosion-Civilization-Accelerated/dp/0465002218

Darwin, C. (1936 [1888]). The Descent of Man and Selection in relation to Sex. reprint of 2nd edition, The Modern Library, New York: Random House.

Frost, P. (2023a). Human evolution didn’t slow down. It accelerated! Peter Frost’s Newsletter, July 12.

Frost, P. (2023b). Do human races exist? Peter Frost’s Newsletter. August 15.

Glazko, G., Veeramachaneni, V., Nei, M., & Makałowski, W. (2005). Eighty percent of proteins are different between humans and chimpanzees. Gene, 346, 215-219. https://doi.org/10.1016/j.gene.2004.11.003

Hawks, J. (2007). Disagreeing with Hillary Clinton on human genetic differences. John Hawks Weblog https://web.archive.org/web/20210624221131/http://johnhawks.net/weblog/topics/race/differences/clinton_2007_proportion_differences_speech.html

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

King, M-C. & Wilson, A.C. (1975). Evolution at two levels in humans and chimpanzees: Their macromolecules are so alike that regulatory mutations may account for their biological differences. Science, 188, 107-116. https://doi.org/10.1126/science.1090005

Kuijpers, Y., Domínguez-Andrés, J., Bakker, O.B., Gupta, M.K., Grasshoff, M., Xu, C.J., Joosten, L.A.B., Bertranpetit, J., Netea, M.G., & Li, Y. (2022). Evolutionary Trajectories of Complex Traits in European Populations of Modern Humans. Frontiers in Genetics, 13, 833190. https://doi.org/10.3389/fgene.2022.833190

Libedinsky, I., Wei, Y., de Leeuw, C., Rilling, J. K., Posthuma, D., & van den Heuvel, M. P. (2025). The emergence of genetic variants linked to brain and cognitive traits in human evolution. Cerebral Cortex, 35(8), bhaf127. https://doi.org/10.1093/cercor/bhaf127

Piffer, D., & Kirkegaard, E.O.W. (2024). Evolutionary Trends of Polygenic Scores in European Populations from the Paleolithic to Modern Times. Twin Research and Human Genetics, 27(1), 30-49. https://doi.org/10.1017/thg.2024.8

Redon, R., Ishikawa, S., Fitch, K. R., Feuk, L., Perry, G. H., Andrews, T. D., ... & Hurles, M. E. (2006). Global variation in copy number in the human genome. Nature, 444(7118), 444-454. https://doi.org/10.1038/nature05329

Rinaldi, A. (2017). We’re on a road to nowhere. Culture and adaptation to the environment are driving human evolution, but the destination of this journey is unpredictable. EMBO reports 18: 2094-2100. https://doi.org/10.15252/embr.201745399

Wikipedia (2025). Human genetic variation. https://en.wikipedia.org/wiki/Human_genetic_variation

Yoo, D., Rhie, A., Hebbar, P., Antonacci, F., Logsdon, G. A., Solar, S. J., ... & Eichler, E. E. (2025). Complete sequencing of ape genomes. Nature, 1-18. https://doi.org/10.1038/s41586-025-08816-3

Impressions of a cityscape, Louis-George Legrand (1801-1883)

Genetic evolution accelerated some 10,000 years ago in our species. Back then, humans already occupied the full range of natural environments from the tropics to the arctic. They were now adapting to new cultural environments. First, there was farming and life in villages, towns, and cities. Then came other changes: specialization of labor, formation of states, development of reading and writing, codification of social norms into law, and so on.

We used to think that cultural evolution simply replaced genetic evolution. In reality, the two have pushed each other forward, even into the time of recorded history. Humans create culture, and culture recreates humans — by selecting those who better fit in and cope with its demands on body and mind.

The mind, in particular, has been shaped by this coevolution with culture. To varying degrees, this has meant:

• processing an increasing volume of written texts and numerical data

• keeping track of interactions with more people, most of whom are not close kin or even friends

• imagining how to make tools for an ever-wider range of tasks

• creating mental models with a longer timeline. Simply put, we no longer live mainly in the present. To act in the present, we must move back and forth between an imagined past and multiple imagined futures, thereby increasing the two-way interaction between thought and action.

This new view of recent human evolution is backed by several recent studies. At first, researchers had to use indirect methods to estimate when and how fast the human genome changed during prehistory and history. By measuring the degree of decay in linkage disequilibrium at a gene, they could estimate the length of time since a mutation or a selective sweep.

This method has been used in two studies of recent human evolution.

Hawks et al. (2007) - humans over the past 40,000 years

Almost 4 million SNPs were analyzed to estimate the rate of change to the human genome over the past 40,000 years.

Findings:

• Genetic evolution accelerated more than a hundredfold some 10,000 years ago, when hunting and gathering gave way to farming and other cultural changes (sedentary living, growth of towns and cities, rise of social complexity, etc.).

• This period of rapid evolution lasted well into the time of recorded history, peaking 8,000 years ago in Africa and 5,250 years ago in Europe.

• In actual fact, these peaks were even more recent. “[T]he peak ages of new selected variants in our data do not reflect the highest intensity of selection, but merely our ability to detect selection. Because of the recent acceleration, many more new adaptive mutations should exist than have yet been ascertained, occurring at a faster and faster rate during historic times.”

• This rapid evolution involved physiological adaptations to new diets or new diseases and, above all, cognitive adaptations to new ways of doing things. “[T]he rapid cultural evolution during the Late Pleistocene created vastly more opportunities for further genetic change, not fewer, as new avenues emerged for communication, social interactions, and creativity” (Hawks et al., 2007).

Libedinsky (2025) - mostly ancestors of Europeans over the past 5 million years

Over 13 million SNPs were analyzed to estimate the rate of change to the human genome over the past 5 million years.

Findings:

• Human evolution was rapid during two periods: 1) 2.4 million to 280,000 years ago with a peak around 1.1 million years ago; and 2) 280,000 to 1,700 years ago with a peak around 55,000 years ago.

• The second period saw rapid evolution in three domains: 1) vision; 2) mental function; and 3) nutrient absorption, digestion, and storage. There was much less evolutionary change in metabolism, skeletal development, and the immune system.

• Changes to the neocortex were more recent than changes elsewhere in the brain (Libedinsky, 2025).

Like Hawks, Libedinsky identified a period of rapid evolution that went well past the spread of humans out of Africa, perhaps even into historic times. But his analysis shows it beginning earlier and peaking earlier, perhaps because of differences between his dataset and Hawks’. The latter dataset, though smaller, was divided equally among several human groups: two of European descent; three of East Asian descent; one of mixed European/Amerindian descent; one of South Asian descent; and four of African descent. Because Libedinsky mostly used data from people of European descent, he largely excluded the evolutionary changes that non-Europeans have experienced over the past 50,000 years since the Out-of-Africa event (Hawks, 2023).

A new research tool

With the retrieval of DNA from human remains, the past few years have seen a revolution in studies of recent human evolution. We can now directly see how human populations have evolved over time, not only physically but also mentally and behaviorally.

Kuijpers et al. (2022) - Europeans over the past ~35,000 years

This was an analysis of genomes from 827 Europeans who lived in the Upper Paleolithic (>11,000 years ago), the Mesolithic (11,000-5,500 years ago), the Neolithic (8,500-3,900 years ago), and the post-Neolithic (5,000 to “more recent ages”). Analysis also included genomes from 250 individuals who now live in present-day Europe.

Evolution of cognitive ability was measured by alleles associated with intelligence, fluid intelligence, and educational attainment (EA).

Findings:

• No change in mean cognitive ability during the long period of hunting and gathering

• With the emergence of farming, a steady rise throughout the Neolithic and into historic times

• Stagnation during the classical era of civilization

• A steady rise over the last few centuries (Kuijpers et al., 2022; see also Frost, 2024a).

Piffer et al. (2023) - Central Italians over the past 10,000 years

This was an analysis of genomes from 127 individuals who lived in central Italy during the pre-Iron Age (10,000-900 BCE), the Iron Age and the Republican Era (900-27 BCE), the Imperial Era (27 BCE-300 CE), Late Antiquity (300-700), and medieval/early modern times (700-1800). Analysis also included genomes from 33 individuals who now live in central Italy. Cognitive ability was measured by alleles associated with educational attainment (EA4)

Findings:

• A steady rise in mean cognitive ability through prehistory and into the time of the Roman Republic

• A sharp fall during the Imperial Era.

• A steady rise from Late Antiquity to early modern times (Frost, 2024b)

• A fall beginning somewhere after 1800 and continuing to the present (Piffer et al., 2023).

Rise, fall, and renewed rise of mean cognitive ability in central Italy (Piffer et al, 2023, Figure 1)

Piffer & Kirkegaard (2024) - Europeans over the past ~35,000 years

This was a replication of Kuijpers et al. (2022), using genomes from more individuals and using more alleles to calculate polygenic scores. The genomes came from 2,625 individuals who lived in Europe from 32,600 to 246 years ago. Cognitive ability was measured by alleles associated with occupational status, EA, and IQ. The alleles associated with EA were identified by other research teams from two large samples, one with over a million participants (EA3) and the other with three million (EA4).

Findings:

• No change in mean cognitive ability from 35,000 to 15,000 years ago.

• A rise beginning near the end of the last ice age and continuing to the present (Piffer & Kirkegaard, 2024; see also Frost, 2024a)

In this study, the upward trend is more continuous and thus less steep, perhaps because of samples from non-Romanized regions that suffered less from the cognitive decline of the Imperial Era.

As with the last study, we see a close alignment of all three measures of cognitive ability. The evolutionary change has clearly been in this trait, and not in some non-cognitive aspect of educational attainment.

Akbari et al. (2024) - Europeans over the past 9,000 years

This was a replication of Kuijpers et al. (2022) and Piffer and Kirkegaard (2024), using genomes from even more individuals: 8,433 Europeans from the past 14,000 years and 6,510 from present-day Europe. Cognitive ability was measured by alleles associated with IQ, household income, and EA (based on alleles identified from a sample of about a million Europeans and East Asians, see Chen et al., 2024).

Findings:

• A steep rise in mean cognitive ability from 9,000 to 7,000 years ago

• A moderate rise from 7,000 to 2,000 years ago

• A fall from 2,000 to 1,000 years ago

• A rise over the last 1,000 years.

The research team also found strong directional selection acting on many hundreds of alleles over the past 10,000 years. There was selection for lighter skin, for lower risk of schizophrenia and bipolar disease, for slower decline in health with age, and for higher cognitive ability.

Strangely, the paper makes no mention of previous work by John Hawks, Ilan Libedinsky, Yunus Kuijpers, Davide Piffer, and Emil Kirkegaard. The Reich Lab researchers are presented as the first proponents of the new view of recent human evolution.

Previous work has shown that classic selective sweeps driving alleles to fixation have been rare over the broad span of human evolution. Thus, we were surprised that over the last 14,000 years in West Eurasia there have been many hundreds of instances of directional selection with coefficients on the order of 0.5% or more. (Akbari et al., 2024, pp. 9-10)

All three measures show that mean cognitive ability has risen substantially in Europeans over the past 9,000 years (Akbari et al., 2024)

Piffer (2025a) - East Asians over the past 12,000 years

This was a study of genomes from 1,245 individuals who have lived in eastern Eurasia over the past 12,000 years. For earlier periods, not enough samples were available to draw firm conclusions. Cognitive ability was measured by alleles associated with EA or IQ (based on 176,400 East Asians, Chen et al., 2024).

Findings:

• No change in mean cognitive ability during the long period of hunting and gathering.

• A steady rise beginning 9,000 to 8,000 years ago, apparently with the advent of farming.

• Stagnation or decline over the past 1,500 years, possibly due to lower fertility among elite individuals. Stagnation is indicated by alleles associated with EA, and a real decline by alleles associated with IQ (Piffer, 2025a; see also Frost, 2025a).

Why do we see decline with IQ alleles and no change with EA alleles? Perhaps the divergence between the two measures reflects the growing importance of the Imperial Examination (Frost, 2011; Wen et al., 2024). This examination provided access to all levels of China’s civil service. As such, it rewarded not only high cognitive ability but also rule following, submissiveness, and resistance to boredom — in short, the ability to sit still at a desk and do tedious assignments. The latter ability is likely better measured by alleles associated with educational attainment than by those associated with IQ.

Mean cognitive ability in East Asia over the past 12,000 years, as measured by alleles associated with EA (Piffer, 2025a, Figure 3, p. 6)

Mean cognitive ability in East Asia over the past 12,000 years, as measured by alleles associated with IQ (Piffer, 2025a, Figure 6, p. 8)

Piffer & Connor (2025a) - An English region from the 11th to 19th centuries

This was a study of genomes from 269 individuals who lived in one region of England (Cambridge and surrounding area) from the 11th to 19th centuries. Cognitive ability was measured by alleles associated with EA.

Findings:

• No clear change until the 1300s, followed by a steady rise in mean cognitive ability until the 19th century.

• On a per capita basis, the highly intelligent became ten times more numerous in England between 1000 and 1850, i.e., the top 1% in 1850 were as smart as the top 0.1% in the year 1000 (Piffer & Connor, 2025a; see also Frost, 2025b).

Piffer & Connor (2025b) - Northwest Europeans from the 8th to 19th centuries

This was a replication of the above study with more genomes (n=600) and a broader spatial and temporal scope (Belgium, Denmark, England, Germany, and the Netherlands from the 8th to 19th centuries). Cognitive ability was measured by alleles associated with EA.

Findings:

• No clear change until the mid-1300s followed by a steady rise in mean cognitive ability until 1850. The total rise was a little over three quarters of a standard deviation.

• The Black Death (1346-1353) may have kickstarted cognitive evolution in northwest Europe by weakening feudalism and freeing up human capital for the emerging market economy: “labour became scarcer, wages rose, and land and capital became relatively cheaper. In that new environment, individuals who could exploit opportunities in trade, crafts, and skilled work had strong advantages” (Piffer, 2025b).

• On a per capita basis, the highly intelligent became over ten times more numerous in northwest Europe between 700 and 1850, i.e., the top 1% in 1850 were smarter than the top 0.1% in the year 700 (Piffer & Connor, 2025b; Piffer, 2025b).

Cognitive evolution in northwest Europe from medieval to modern times (Piffer, 2025b)

Has David Reich walked back his finding?

Of the studies presented above, one has especially gained public attention: the 2024 paper by Ali Akbari and other researchers at the David Reich Lab — a renowned research laboratory in the Department of Genetics at Harvard Medical School. David Reich himself has become widely known for his book Who We Are and How We Got Here.

Reich’s team found that mean cognitive ability has risen substantially over the past 9,000 years. The signals of selection are well above the threshold of significance:

We finally observe signals of selection for combinations of alleles that today predict three correlated behavioral traits: scores on intelligence tests (increasing 0.79 ± 0.14), household income (increasing 1.11 ± 0.14), and years of schooling (increasing 0.61 ± 0.13). These signals are all highly polygenic, and we have to drop 463 to 1109 loci for the signals to become nonsignificant (Akbari et al., 2024, p. 9)

This finding was judged important enough to appear in the Abstract:

We also identify selection for combinations of alleles that are today associated with lighter skin color, lower risk for schizophrenia and bipolar disease, slower health decline, and increased measures related to cognitive performance (scores on intelligence tests, household income, and years of schooling). (Akbari et al., 2024)

Yet David Reich seems to deny this finding in a podcast interview with Dwarkesh Patel:

If you look at traits that we know today affect immune disease or metabolic disease, these traits are highly overrepresented by a factor of maybe four in the collection of variants that are changing rapidly over time. Whereas if you look at traits that are affecting cognition that we know in modern people modulate behavior, they’re hardly affected at all. Selection in the last 10,000 years doesn’t seem to be focusing, on average, on cognitive and behavioral traits. It seems to be focusing on immune and cardiometabolic traits, on average, with exceptions. On average, there’s an extreme over representation of cardio metabolic traits. (Patel, 2024, 00:55:18)

Sasha Gusev (a cancer researcher and social activist) has cited this quote to argue that David Reich didn’t mean what he seemed to mean in his recent paper (Gusev, 2025). What, then, did he mean? The Reich Lab paper has a paragraph that closely matches the above quote.

… signals of selection are unusually associated with specific classes of traits. In particular, we find enrichment for SNPs contributing to blood-immune-inflammatory traits, compared to random SNPs with matched characteristics defining the baseline of 1-fold. In contrast, for mental-psychiatric-nervous and behavioral traits, we do not see enrichment. These patterns cannot be explained by differences in allele frequencies or purifying selection since we control for these factors. The intensity of selection on blood-immune-inflammatory and cardio-metabolic traits increased in the Bronze Age relative to the pre-farming period, which may reflect adaptation to new diets, higher population densities, or living closer to domesticated animals. (Akbari et al., 2024, p. 4)

In the appendix, Figure 1 has two aggregate categories “Mental-psychiatric-nervous” and “Behavioral”, each of which encompasses 13 traits. Presumably, the first category includes the 3 measures of cognitive ability, with the 10 others being Qualifications none, Schizophrenia, Reaction time, Bipolar disorder, Tobacco use disorder, Mental disorders related to tobacco, Noncognitive aspects of educational attainment, Qualifications A levels, Neuroticism and Depressive symptoms (Extended Data Figure 7; Extended Data Figure 13). Such a vast category would not provide a clear signal of selection, since many of these traits are independent of each other, and some are inversely correlated with each other. For instance, schizophrenia is inversely correlated with cognitive ability.

Perhaps Reich initially thought he had evidence for rapid cognitive evolution but then changed his mind. This, too, is unlikely. Patel interviewed him a month before the release of the Reich Lab paper.

More likely, Reich had seen an earlier version of the paper when he met Patel — a version that did not yet include the separate analyses of cognitive ability. In fact, in the interview transcript, he talks about these analyses as if they were still ongoing:

You might expect that in reaction to a change so economically, dietarily, cognitively transformative as agriculture, the genome might shift in terms of how it adapts. You might actually see that in terms of adaptation on the genome. You might expect to see a quickening of natural selection or a change. I don’t think we know the answer yet to whether that’s occurred, although they’re beginning to be hints. We could learn that from the DNA data. (Patel, 2024, 00:50:44)

What next?

In tropical regions, research will be hindered by the poor preservation of DNA in human remains. Complete or near-complete retrieval of ancient DNA is necessary for the study of cognitive evolution over time.

Nonetheless, we will soon see a study of cognitive evolution in South Asia. Its findings will probably resemble those of East Asia and the Roman world: a steady rise in mean cognitive ability, followed by decline. Cognitive evolution is usually driven by the higher fertility of elite individuals, and such people inevitably reach a level of development where the pursuit of wealth and power becomes severed from reproductive success.

Hopefully, we will see more studies of European cognitive evolution, particularly for the key period of late medieval and post-medieval times. During this period, mean cognitive ability rose steadily, perhaps beginning to rise earlier in England, Holland, and northern Italy. By charting this advance by region and by century, we may learn more about the historical processes leading to the Reformation, the Enlightenment and the Industrial Revolution.

In the case of East Asia, we should closely examine the evolutionary trend of the last 1,500 years. Did mean cognitive ability simply stop rising? Or did it actually decline? In addition, did China and Japan differ in their trajectories of cognitive evolution?

Finally, future studies ought to control genomic data for socioeconomic status (SES). Otherwise, there may be sampling bias. Since the remains of elite individuals are more likely to survive the passage of time, their DNA should be overrepresented in older samples. Cognitive ability would thus seem to decrease over time. Of course, we generally see the opposite: an increase in cognitive ability over time. So there may have been even more cognitive evolution than what current studies suggest.

In any case, we can and should control for the SES of those deceased individuals who provide DNA. Often, it is already known. If not, it can be inferred through isotope analysis of the bones and teeth. High SES is associated with an ample, high-meat diet, and low SES with coarse grains and malnutrition.

References

Akbari, A., Barton, A.R., Gazal, S., Li, Z., Kariminejad, M., Perry, A., Zeng, Y., Mittnik, A., Patterson, N., Mah, M., Zhou, X., Price, A.L., Lander, E.S., Pinhasi, R., Rohland, N., Mallick, S., & Reich, D. (2024). Pervasive findings of directional selection realize the promise of ancient DNA to elucidate human adaptation. bioRxiv. September 15 https://doi.org/10.1101/2024.09.14.613021

Chen, T. T., Kim, J., Lam, M., Chuang, Y. F., Chiu, Y. L., Lin, S. C., ... & Won, H. H. (2024). Shared genetic architectures of educational attainment in East Asian and European populations. Nature Human Behaviour, 8(3), 562-575. https://doi.org/10.1038/s41562-023-01781-9

Frost, P. (2011). East Asian intelligence, Evo and Proud, February 18. https://evoandproud.blogspot.com/2011/02/east-asian-intelligence.html

Frost, P. (2024a). Cognitive evolution in Europe: Two new studies. Peter Frost’s Newsletter, March 14. https://www.anthro1.net/p/cognitive-evolution-in-europe-two

Frost, P. (2024b). How Christianity rebooted cognitive evolution, Aporia Magazine, October 10.

Frost, P. (2025a). Cognitive evolution in eastern Eurasia, Aporia Magazine, February 8.

Frost, P. (2025b). The Great Cognitive Advance, Aporia Magazine, July 31.

Gusev, S. (2025). https://x.com/SashaGusevPosts/status/1932649103898366243

Hawks, J. (2023). Did two pulses of evolution supercharge human cognition? John Hawks Weblog, May 15. https://johnhawks.net/weblog/did-two-pulses-of-evolution-supercharge-human-cognition/

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 (USA), 104, 20753-20758. https://doi.org/10.1073/pnas.0707650104

Kuijpers, Y., Domínguez-Andrés, J., Bakker, O.B., Gupta, M.K., Grasshoff, M., Xu, C.J., Joosten, L.A.B., Bertranpetit, J., Netea, M.G., & Li, Y. (2022). Evolutionary Trajectories of Complex Traits in European Populations of Modern Humans. Frontiers in Genetics, 13, 833190. https://doi.org/10.3389/fgene.2022.833190

Libedinsky, I., Wei, Y., de Leeuw, C., Rilling, J. K., Posthuma, D., & van den Heuvel, M. P. (2025). The emergence of genetic variants linked to brain and cognitive traits in human evolution. Cerebral Cortex, 35(8), bhaf127. https://doi.org/10.1093/cercor/bhaf127

Patel, D. (2024). David Reich – How one small tribe conquered the world 70,000 years ago. Dwarkesh Podcast, August 29.

Piffer, D. (2025a). Directional Selection and Evolution of Polygenic Traits in Eastern Eurasia: Insights from Ancient DNA. Twin Research and Human Genetics, 28(1), 1-20. https://doi.org/10.1017/thg.2024.49

Piffer, D. (2025b). The genetic evolution of the human race and its consequences for the Industrial Revolution, November 17, PifferPilfer.

Piffer, D., & Connor, G. (2025a). Genomic Evidence for Clark’s Theory of the British Industrial Revolution, preprint, ResearchGate, June. https://www.researchgate.net/publication/392808200_Genomic_Evidence_for_Clark’s_Theory_of_the_British_Industrial_Revolution

Piffer, D., & Connor, G. (2025b). Genomic Evidence for Clark’s Theory of the Industrial Revolution, preprint, ResearchGate, November. https://www.researchgate.net/publication/392808200_Genomic_evidence_for_Clark’s_theory_of_the_Industrial_Revolution

Piffer D, Dutton, E., & Kirkegaard, E.O.W. (2023). Intelligence Trends in Ancient Rome: The Rise and Fall of Roman Polygenic Scores. OpenPsych, July 21, 2023. https://doi.org/10.26775/OP.2023.07.21

Piffer, D., & Kirkegaard, E.O.W. (2024). Evolutionary Trends of Polygenic Scores in European Populations from the Paleolithic to Modern Times. Twin Research and Human Genetics, 27(1), 30-49. https://doi.org/10.1017/thg.2024.8

Wen, F., Wang, E. H., & Hout, M. (2024). Social mobility in the Tang Dynasty as the Imperial Examination rose and aristocratic family pedigree declined, 618–907 CE. Proceedings of the National Academy of Sciences, 121(4), e2305564121. https://doi.org/10.1073/pnas.2305564121

Reverie, 1872, Edward Radford (Wikicommons)

Toxoplasma gondii is a tiny protozoan. It’s also a parasite. Like a surprising number of parasites, it can infiltrate brain tissue and make its host behave in ways that help it spread to new hosts.

For example, an infected rat is attracted to the smell of cat urine and ends up getting eaten by a cat, the only host in which T. gondii can sexually reproduce. When the parasite enters a rodent host, through food or water contaminated by cat feces, it targets the amygdala—an area of the brain that stores emotional memories. Autopsies of infected rodents show a density of T. gondii cysts in the amygdala that is twice the density of other brain regions (Vyas et al., 2007). Primates can also get infected, and the behavioral changes are similar. Infected chimpanzees like the smell of leopard urine (Poirotte, 2016).

Although a feline species is the only host in which T. gondii can sexually reproduce, “any warm-blooded predator or omnivore can function as a paratenic host, accumulating different genetically distinct strains of Toxoplasma” (Flegr, 2025). These strains can thus evolve to become more adapted to a new host species, even in the absence of sexual reproduction (Ferguson, 2022).

Humans can get infected—between 13 and 43% of adults 25 to 50 years old in Europe, with lower rates in North America and higher ones in Brazil and Africa (Friesema et al., 2025). Some strains seem to have adapted specifically to human bodies. Of the three main lineages, Type II strains are the most common in our species. They are also much less lethal than Type I, perhaps because they have coevolved with humans to a greater extent (Delgado et al., 2022; Hosseini et al., 2019; Xiao & Yolken, 2015).

Infection causes humans, like other animals, to change mentally and behaviorally. Men become more jealous, women more easy-going, and both sexes slower to respond to threats, as shown by a higher risk of traffic accidents and a longer reaction time. The last finding shows the direction of causality: the longer you have been infected, the slower you react. Slower reaction time does not increase your risk of infection (Flegr et al., 2005; Flegr, 2025; Havlíček et al., 2001; Latifi et al., 2025).

T. gondii manipulates not only behavior but also physical appearance. Infected men are taller, and women rate them as being more dominant and masculine. Infected men also have lower second-to-fourth digit ratios, a sign of greater exposure to the male hormones (Flegr et al., 2005; Flegr, 2025; Hodková et al., 2007; Latifi et al., 2025). This physical manipulation is in line with a strategy of making infected males more attractive to prospective hosts (Mackey & Immerman, 2003).

Can this behavioral and physical manipulation be demonstrated under controlled conditions? Admittedly, we cannot experiment with humans as we do with lab animals. Humans also live long, and decades may pass between the initial infection and the time when the parasite makes its presence known (Cochran et al., 2000; Frost, 2020).

Nonetheless, a group of Czech researchers is convinced that some strains of T. gondii have evolved the ability to manipulate human behavior, specifically sex. Infected men not only seek more sexual partners but also engage in practices that transmit the parasite more efficiently, i.e., into the partner’s oral cavity or anal canal. The male host gets no reproductive benefit; in fact, his fertility is reduced through lower sperm counts and lower sperm motility (Hlaváčová et al., 2021; Kaňková et al., 2020).

At the extremes of sexual behavior, infected individuals, particularly men, feel more drawn to masochism, rape fantasies, bondage, same-sex experiences, and anal sex. Women with latent toxoplasmosis also report feeling more drawn to violent sexual practices and same-sex experiences. But such fantasies are less often acted out by infected individuals. Real changes to sexual behavior seem confined to a greater propensity for sexual promiscuity and oral sex, as well as anal sex among men who have sex with men (Flegr, 2017; Flegr, 2025; Flegr & Kuba, 2016; Latifi et al., 2025).

T. gondii cyst in a mouse brain (Wikicommons, J.P. Dubey)

We know that T. gondii is present in semen as cysts containing thousands of spores (Tong et al., 2023). We also know that the cysts are spread from one body to another through fellatio, anal sex, or vaginal sex. The transmission is male to female or male to male, but not female to male (Flegr, 2025; Flegr et al., 2014).

Sexual transmission, particularly from a male host, is indicated by several lines of evidence:

• In heterosexual couples, an infected male partner increases the female partner’s risk of infection, but an infected female partner does not increase the male partner’s risk of infection.

• In women, the risk of infection correlates with sex work, unprotected sex before pregnancy, and history of genital injuries. In men it correlates with sexual promiscuity, and in male homosexuals with unprotected anal sex.

• Seroprevalence of T. gondii is higher in fellating individuals of either sex than in non-fellating controls.

• Seroprevalence is higher in homosexuals and promiscuous individuals.

• Seroprevalence is positively correlated with the prevalence of STDs across countries, including HIV.

• Seroprevalence is higher in women than in men. The gender difference emerges in the 10-to-14 age group and peaks among 20-to-39-year-olds (Flegr et al., 2005; Flegr, 2025; Latifi et al., 2025).

We are only beginning to realize that T. gondii can be sexually transmitted. Most health authorities still believe it is transmitted solely through contaminated food and water or through contact with cats. Yet there must be some sexual transmission, particularly among young men, as shown by the sex ratio of infected individuals. An Italian survey found that recent infections were more frequent in men than in women within the 20 to 40 age bracket (Pinto et al., 2015). A German survey found the rate of seroprevalence to be 1.76 times higher for men, and higher still for younger men (Wilking et al., 2016). An American survey found a rate of 11.25% for males and 9.67% for females (Jones et al., 2017). According to a meta-analysis of workers occupationally exposed to animals, the rate was 63% for male workers and 37% for female workers (Mohammed et al., 2024).

This male predominance, particularly among young adults, seems best explained by a subpopulation of men who have sex with men, either exclusively or occasionally. Alternatively, men in this age bracket are more exposed to cats or contaminated food and water.

T. gondii seems to have gained a capacity for passive sexual transmission and then active sexual transmission through three stages of coevolution with our species:

1. Entry into a human population, apparently via contact with cats. This animal began to coexist with humans some 10,000 years ago in the Middle East (Galal et al., 2022).

2. Passive sexual transmission. The parasite was now spreading to other humans via intimate contact but still could not modify human behavior. Natural selection favored those variants that were better adapted to humans. The raw material for evolution did not come from new genetic combinations arising through meiosis during sexual reproduction (which is possible only in a cat’s body). Instead, it came from mutations arising through asexual replication and, less often, through parasexual recombination-like events (Ferguson, 2022).

3. Active sexual transmission. Natural selection was now favoring those variants that could increase sexual transmission by modifying host behavior.

Sexual transmission of T. gondii remains under-studied due to a research preference for STDs that 1) produce symptoms soon after infection, 2) are easily observable, and 3) develop over a short span of time. These characteristics are not optimal in long-lived species like our own, which can offer an invasive microbe decades of opportunities to infect other individuals. It should therefore remain discreet and do as little harm as possible as long as its host is useful.

Indeed, some diseases of old age may be due not to old age itself but to long-present parasites that have nothing to lose anymore. By that stage of life, the host has stopped having sex, at least on a regular basis, and can be harmed without any real loss. It is only then that the pathogen attracts notice—after a long period of relatively peaceful coexistence. This lengthy coexistence likely involves some degree of behavioral manipulation.

Are other parasites messing with our sex lives?

T. gondii may be one of many microorganisms that have evolved to manipulate human behavior. We are too good to pass up. Our complex neural circuitry, our social nature, and our dense populations make us ideal vectors. Even if a parasite initially has no ability to manipulate human behavior, such an ability could evolve soon enough. All of the right conditions are in place.

The following are several disorders that may be caused by behavior-manipulating microbes. Again, the disorder itself could represent the final destructive stage of a long-term infection.

HIV-Associated Neurocognitive Disorders (HAND)

Although HIV-associated neurocognitive disorders (HAND) are widely attributed to HIV, the relationship between the two remains circumstantial. In fact, HAND occurs even in individuals who have lost all detectable traces of HIV through antiretroviral therapy. Yet 21% of them go on to develop dementia (McArthur & Brew, 2010).

The causal agent seems to be a pathogen other than HIV—an AIDS cofactor, but only in the sense of orienting the host toward behaviors that increase the risk of HIV infection. The pathogen may be hiding in plain sight among the opportunistic infections currently blamed on the host’s compromised immune system. While such infections primarily target the lungs, the brain is the second most-common target (Masliah et al., 2000; Jellinger et al., 2000).

The existence of an HIV cofactor is consistent with the profile of AIDS victims in a study from Bologna, Italy. In that country, AIDS is transmitted mainly via intravenous drug use, yet transmission via homosexuality/bisexuality is ten times more often associated with cognitive impairment (De Ronchi et al., 2002; Wikipedia, 2025). Certainly, AIDS does impair cognition, as shown by an association between low white cell counts and HAND in the Bologna study. But some cofactor must also be impairing cognition via the homosexuality/bisexuality route.

The cofactor might be the hepatitis C virus (HCV). In a large cohort of people with HIV, the risk of HAND was higher among those with antibodies to HCV (Mastrorosa et al., 2023).

Sexually transmissible strains of Candida albicans (vaginal yeast)

This fungal microbe has been implicated in several neurodegenerative diseases, notably Alzheimer's, multiple sclerosis, and autism spectrum disorders (Alonso et al., 2014; Benito-Leon & Laurence, 2017; Hughes & Ashwood, 2018; Pisa et al., 2016). It can thus cross the blood-brain barrier and adhere to tissues in the central nervous system, particularly white matter. Because of antibody cross-reactivity, the actual species is difficult to identify, though C. albicans is the main suspect (Denaro et al., 1995; Jong et al., 2001).

Despite its classification as a single species, C. albicans encompasses many strains that differ substantially in various ways: single nucleotide polymorphisms, inversions, copy number changes, loss of heterozygosity, and whole or partial chromosomal aneuploidies. At least one of these changes is responsible for altering the balance between commensalism and pathogenesis (Hirakawa et al., 2015; see also Tian et al., 2021).

Some strains have adapted specifically to the vagina, with a few causing vulvovaginal candidiasis (VVC). Commonly known as vaginal yeast infection), VVC affects 70 to 75% of sexually active women at least once and 5 to 8% recurrently. In China, two strains account for almost 60% of all VVC cases, with neither being present at extragenital sites (Li et al. 2008).

Sexual transmission is indicated by several lines of evidence:

• Once a vaginal infection develops, it can spread to the male partner’s glans penis via vaginal sex or to his oral cavity via cunnilingus (Li et al., 2008; Schmid et al., 1995).

• The same strains seem to infect both the vagina and the glans penis (Ge et al., 2012; see also Tian et al., 2021).

• In both men and women, genital yeast infections are associated with a higher number of sexual partners (Warszawski et al., 1996).

• There seems to have been selection for sexual transmissibility, particularly via oral sex. Vaginal strains adhere better than other strains to saliva-coated surfaces (Schmid et al. 1995).

Transmission is female to male and not male to female. Infected people do not have higher rates of vaginal sex but do have higher rates of oral sex, notably cunnilingus (Hellberg et al., 1995; Reed et al., 2003). C. albicans in the male partner’s oral cavity does not predict recurrence of VVC in the female partner (Reed et al., 2003), and treatment of the male partner with antifungals does not prevent recurrence of VVC in the female partner (Bisschop et al., 1986; Buch et al. 1982).

Infection goes through three stages: 1) colonization of the vagina as a commensal with low virulence and no VVC, potentially for a long period of relative latency; 2) colonization of brain sites that influence sexual behavior; and 3) activation of the VVC stage, during which the pathogen can spread to a male partner. The final stage is when the parasite may alter the behavior of its female host, perhaps by weakening her sexual inhibitions and encouraging vaginal contact with her partner’s body, particularly his mouth.

Multiple sclerosis

This disease can cause lesions throughout the nervous system, often the limbic system of the brain and particularly the amygdala (Batista et al., 2016; Hillyer et al., 2023; Meyer-Arndt et al., 2022). As we have seen with T. gondii, this brain region is a primary target for manipulation of behavior.

Multiple sclerosis (MS) seems to be associated with a fungal parasite that infiltrates the brain and nervous system, perhaps a form of Candida. In people with MS, this association is indicated by:

• Antibodies against various Candida species

• High levels of immune defense proteins that bind to mannoproteins, which are ubiquitous in fungal cell walls but rare in bacterial and mammalian cell walls

• High levels of chitotriosidase, which the immune system produces to destroy chitin—a component of fungal cell walls but not of bacterial and mammalian cell walls.

• Successful treatment of MS with a fungicide, dimethyl fumarate (Benito-Leon & Laurence, 2017).

MS seems to be sexually transmitted. It is rare before puberty, two to three times more common in women, more common in women taking oral contraceptives, and associated with smoking—a sociological marker of sexual activity among women. It is also associated with herpes simplex virus type 2 (Benito-Leon & Laurence, 2017; Golden & Voskuhl, 2017; Hawkes, 2002).

Unlike the unknown parasite that causes HAND, the one that causes MS begins doing harm early in adulthood, at around 30 years of age, perhaps because this is when female hosts tend to abandon their multi-partner lifestyle. Consequently, the parasite no longer has anything to gain from keeping its host healthy. The mean age of onset for MS has nonetheless risen in recent decades (Golden & Voskuhl, 2017; Romero-Pinel et al., 2022).

A role in alternative sexualities?

We will now turn to alternative sexual lifestyles. Could these also be due, in some cases, to behavioral manipulation by unknown parasites?

The caveat “in some cases” should be highlighted, since such parasites must become sexually transmissible before they can evolve the ability to manipulate sexual behavior. They should thus facilitate an existing behavior, which would exist even without them. Nonetheless, as we shall see, creation of novel behaviors is not excluded.

Exclusive male homosexuality

More than two decades ago, Greg Cochran argued for the existence of a “gay germ,” i.e., a pathogen that reorients male sexual preferences to increase the number of hosts it can infect via sexual relations (Cochran et al., 2000). He ruled out a genetic cause because of the high fitness cost of exclusive male homosexuality and its heritability of only 20%. Instead, the cause may be a pathogen that targets the limbic system. “Indeed, anecdotal reports indicate that changes in human sexual orientation have occurred following changes in the limbic area due to trauma or infection.” The pathogen may be exploiting a niche that provides more opportunities for sexual transmission. “[H]omosexual behavior could facilitate spread because of the larger numbers of partners homosexual males may have on average, relative to heterosexual males” (Cochran et al., 2000).

This theory may be supported by recent findings about the gut microbiome of men who have sex with men. Specifically, they have fewer of the commensal bacteria that help repair and preserve the intestinal lining. Perhaps an invasive pathogen destroys such bacteria in order to facilitate its entry into the body. Although the altered microbiome facilitates HIV infection, having been in fact discovered through AIDS research, HIV itself does not seem to be responsible (Armstrong et al., 2018; Lin et al., 2024; Tuddenham et al., 2020).

Cuckold fetish

Other pathogens may cause certain paraphilias, such as the cuckold fetish—where a man no longer feels sexual jealousy and even derives pleasure from the prospect of being cuckolded. This fetish is not mentioned in Greco-Roman literature, which nonetheless attests to a wide range of alternative sexualities. The earliest references come from 17th century England, where it was said to be common among merchants (Frost, 2023).

The date and milieu point to an external source—most likely, slaves imported through trade from West Africa. The region’s high polygyny rate would have been conducive to the evolution of STDs that can spread from one polygynous household to another. Keep in mind that the barriers to transmission were already low. On the one hand, the head of household was generally an older man who could not satisfy all his wives; on the other, the latter were often solicited by young single men, as inevitably happens in a polygynous society.

This point is made by anthropologist Pierre van den Berghe:

The temporary celibacy of young men in polygynous societies is rarely absolute, however. While it often postpones the establishment of a stable pair-bond and the procreation of children, it often does not preclude dalliance with unmarried girls, adultery with younger wives of older men, or the rape or seduction of women conquered in warfare. Thus, what sometimes looks like temporary celibacy is, in fact, temporary promiscuity. (van den Berghe, 1979, pp. 50-51)

Polygynous households were thus vulnerable to any STD that could overcome the already low barriers to entry. Once this behavioral niche was colonized, selection favored those strains that could further lower the barriers by inhibiting or even reversing male jealousy.

Conclusion

In theory, all STD pathogens should be under selection for the ability to manipulate host behavior. Their very existence depends on how the host behaves, and even a slight behavioral change could significantly improve their ability to infect other individuals. Sexual transmission is an ideal target for manipulation because it provides the pathogen with potentially numerous opportunities for spreading to new hosts.

Such manipulation can take different forms. Some pathogens may make the male or female host less discriminating, as is the case with T. gondii. Others may reorient the host’s sexual orientation. Still others may cause sexual fetishes and other paraphilias. In fact, all STDs may be evolving to some degree along this trajectory. It has been suggested that certain well-known ones (HIV, HPV, Chlamydia trachomatis, Neisseria gonorrhoeae) help an infected individual continue to infect others by suppressing production of pro-inflammatory cytokines and thereby inhibiting the loss of interest in sex that comes with sickness behavior (Adamo, 2014).

It is difficult to determine the full extent of this phenomenon. One avenue of research would be to examine the human brain for signs of strategies to counter behavioral manipulation, as two evolutionary biologists have wondered:

How much of our neural complexity is a necessary defense against manipulative invaders? How much of the enormous redundancy is to provide system level functionality if part of the system is attacked? How much of the complex process of wiring a brain during development is to prevent pathogen re-wiring? (Read & Braithwaite, 2012, p. 195)

These defensive strategies achieve their aim in three ways: 1) restrict access to the brain, primarily by defending the blood-brain barrier; 2) increase the costs of manipulation, either by increasing the amount of neuroactive substance needed for a behavioral response or by making this substance toxic at the microscopic level; and 3) increase robustness, by creating redundant or modular neural systems as “backups” to ensure the integrity of behavioral responses (Del Giudice, 2019).

The first strategy may be attested by the presence of decoy surface molecules. Such decoys mimic the surface molecules used by invasive pathogens to enter a cell while not allowing them to enter. An example in humans is CEACAM3, which mimics the normal surface molecule CEACAM1. It is fast-evolving in our species, apparently because it has been in an evolutionary arms race with certain pathogens: the latter evolve to avoid this surface molecule, which in turn evolves to bind the pathogen even more (Del Giudice, 2019). Are such decoys produced by our neural tissues?

Evolutionary psychologist Marco Del Giudice is skeptical about the possibility that behavior-manipulating parasites have adapted to human bodies, notably because of the size and complexity of our brains. We thus have so much neural redundancy that a parasite would have to pull too many switches in too many places to produce a substantial effect. We are also much bigger than any invasive microbe and can metabolically invest much more in counter-measures.

Other things, however, do not work in our favor. First, we have a much longer generation time than any invasive microbe. The latter can evolve faster than we can. Second, while our large, complex brains are more difficult to control completely, they offer more opportunities for partial control. Third, as already noted, if the parasite is sufficiently discreet, at least until its human host is too old to have sex, the cost of its presence may be too low for the host to invest in rigorous countermeasures.

But one other thing does work in our favor: humans can fight parasites through cultural rules and prohibitions. Rules against sexual promiscuity and non-vaginal sex may reflect a vague awareness that the sex act can transmit not only semen but also unwanted free riders (Mackey & Immerman, 2001; Mackey & Immerman, 2003).

References

Adamo, S. A. (2014). Parasitic aphrodisiacs: manipulation of the hosts’ behavioral defenses by sexually transmitted parasites. Integrative and Comparative Biology, 54(2), 159-165. https://doi.org/10.1093/icb/icu036

Alonso, R., Pisa, D., Rábano, A., & Carrasco, L. (2014). Alzheimer’s disease and disseminated mycoses. European Journal of Clinical Microbiology & Infectious Diseases, 33(7), 1125-1132. https://doi.org/10.1007/s10096-013-2045-z

Armstrong, A. J., Shaffer, M., Nusbacher, N. M., Griesmer, C., Fiorillo, S., Schneider, J. M., ... & Lozupone, C. A. (2018). An exploration of Prevotella-rich microbiomes in HIV and men who have sex with men. Microbiome, 6(1), 198. https://doi.org/10.1186/s40168-018-0580-7

Batista, S., d’Almeida, O. C., Afonso, A., Freitas, S., Macário, C., Sousa, L., ... & Cunha, L. (2017). Impairment of social cognition in multiple sclerosis: Amygdala atrophy is the main predictor. Multiple Sclerosis Journal, 23(10), 1358-1366. https://doi.org/10.1177/1352458516680750

Benito-Leon, J., & Laurence, M. (2017). The Role of Fungi in the Etiology of Multiple Sclerosis, Frontiers in Neurology 16 October https://doi.org/10.3389/fneur.2017.00535

Bisschop, M. P. J. M., Merkus, J. M. W. M., Scheygrond, H., & Van Cutsem, J. (1986). Co‐treatment of the male partner in vaginal candidosis: a double‐blind randomized control study. BJOG: An International Journal of Obstetrics & Gynaecology, 93(1), 79-81. https://doi.org/10.1111/j.1471-0528.1986.tb07818.x

Buch, A., & Christensen, E. S. (1982). Treatment of vaginal candidosis with natamycin and effect of treating the partner at the same time. Acta Obstetricia et Gynecologica Scandinavica, 61(5), 393-396. https://doi.org/10.3109/00016348209156578

Cochran, G.M., Ewald, P.W., & Cochran, K.D. (2000). Infectious Causation of Disease: An Evolutionary Perspective. Perspectives in Biology and Medicine, 43(3), 406-448. https://doi.org/10.1353/pbm.2000.0016

De Ronchi, D., Faranca, I., Berardi, D., Scudellari, P., Borderi, M., Manfredi, R., & Fratiglioni, L. (2002). Risk Factors for Cognitive Impairment in HIV-1-Infected Persons with Different Risk Behaviors. Archives of Neurology, 59(5), 812-818. https://doi.org/10.1001/archneur.59.5.812

Delgado, I. L., Zúquete, S., Santos, D., Basto, A. P., Leitão, A., & Nolasco, S. (2022). The apicomplexan parasite Toxoplasma gondii. Encyclopedia, 2(1), 189-211. https://doi.org/10.3390/encyclopedia2010012

Del Giudice, M. (2019). Invisible designers: brain evolution through the lens of parasite manipulation. The Quarterly Review of Biology, 94(3), 249-282. https://doi.org/10.1086/705038

Denaro, F.J., López-Ribot, J.L., & Chaffin, W.L. (1995). Adhesion of Candida albicans to Brain Tissue of Macaca mulata in an Ex Vivo Assay. Infection and Immunity , 63(9), 3438-3441. https://doi.org/10.1128/iai.63.9.3438-3441.1995

Ferguson, D.J.P. (2022). Toxoplasma gondii: Detailed Description of the Coccidian (Asexual and Sexual) Development and Oocyst Sporulation. In: de Souza, W. (eds) Lifecycles of Pathogenic Protists in Humans. Microbiology Monographs, vol 35. Springer, Cham. https://doi.org/10.1007/978-3-030-80682-8_9

Flegr J. (2013). Influence of latent Toxoplasma infection on human personality, physiology and morphology: pros and cons of the Toxoplasma-human model in studying the manipulation hypothesis. Journal of Experimental Biology, 216, 127–133. https://doi.org/10.1242/jeb.073635

Flegr, J. (2017). Does Toxoplasma infection increase sexual masochism and submissiveness? Yes and no. Communicative & integrative biology, 10(5-6), e1303590. https://doi.org/10.1080/19420889.2017.1303590

Flegr, J. (2025). Thirty years of studying latent toxoplasmosis: behavioural, physiological, and health insights. Folia Parasitologica, 72, 1-16. http://dx.doi.org/10.14411/fp.2025.005

Flegr, J., Hrŭsková, M., Hodný, Z., Novotná, M., & Hanušová, J. (2005). Body height, body mass index, waist-hip ratio, fluctuating asymmetry and second to fourth digit ratio in subjects with latent toxoplasmosis. Parasitology, 130(6), 621-628. https://doi.org/10.1017/S0031182005007316

Flegr, J., Klapilová, K., & Kaňková, Š. (2014). Toxoplasmosis can be a sexually transmitted infection with serious clinical consequences. Not all routes of infection are created equal. Medical Hypotheses, 83(3), 286-289. https://doi.org/10.1016/j.mehy.2014.05.019

Flegr, J., & Kuba, R. (2016). The relation of Toxoplasma infection and sexual attraction to fear, danger, pain, and submissiveness. Evolutionary Psychology, 14(3), 1474704916659746. https://doi.org/10.1177/1474704916659746

Friesema, I. H., Waap, H., Swart, A., Györke, A., Le Roux, D., Evangelista, F. M., ... & Opsteegh, M. (2025). Systematic review and modelling of Toxoplasma gondii seroprevalence in humans, Europe, 2000 to 2021. Eurosurveillance, 30(34), 2500069. https://doi.org/10.2807/1560-7917.es.2025.30.34.2500069

Frost, P. (2023). Cuckoldry: Sexual Fantasies. In: Shackelford, T.K. (ed.) Encyclopedia of Sexual Psychology and Behavior. Springer. https://doi.org/10.1007/978-3-031-08956-5_757-1

Frost, P. (2020). Are Fungal Pathogens Manipulating Human Behavior? Perspectives in Biology and Medicine, 63(4), 591-601. https://doi.org/10.1353/pbm.2020.0059

Galal, L., Ariey, F., Gouilh, M. A., Dardé, M. L., Hamidović, A., Letourneur, F., ... & Mercier, A. (2022). A unique Toxoplasma gondii haplotype accompanied the global expansion of cats. Nature Communications, 13(1), 5778. https://doi.org/10.1038/s41467-022-33556-7

Ge, S. H., Xie, J., Xu, J., Li, J., Li, D. M., Zong, L. L., ... & Bai, F. Y. (2012). Prevalence of specific and phylogenetically closely related genotypes in the population of Candida albicans associated with genital candidiasis in China. Fungal Genetics and Biology, 49(1), 86-93. https://doi.org/10.1016/j.fgb.2011.10.006

Golden, L.C., & Voskuhl, R. (2017). The Importance of Studying Sex Differences in Disease: The example of Multiple Sclerosis. Journal of Neuroscience Research, 95(1-2), 633-643. https://doi.org/10.1002/jnr.23955

Havlíček, J., Gasová, Z. G., Smith, A. P., Zvára, K. & Flegr, J. (2001). Decrease of psychomotor performance in subjects with latent ‘asymptomatic’ toxoplasmosis. Parasitology, 122, 515-520. https://doi.org/10.1017/S0031182001007624

Hawkes, C. H. (2002). Is multiple sclerosis a sexually transmitted infection?. Journal of Neurology, Neurosurgery & Psychiatry, 73(4), 439-443. https://doi.org/10.1136/jnnp.73.4.439

Hellberg, D., Zdolsek, B., Nilsson, S., & Mårdh, P. A. (1995). Sexual behavior of women with repeated episodes of vulvovaginal candidiasis. European Journal of Epidemiology, 11, 575-579. https://doi.org/10.1007/BF01719311

Hillyer, A., Sharma, M., Kuurstra, A., Rosehart, H., Menon, R., & Morrow, S. A. (2023). Association between limbic system lesions and anxiety in persons with multiple sclerosis. Multiple Sclerosis and Related Disorders, 79, 105021. https://doi.org/10.1016/j.msard.2023.105021

Hirakawa, M. P., Martinez, D. A., Sakthikumar, S., Anderson, M. Z., Berlin, A., Gujja, S., ... & Cuomo, C. A. (2015). Genetic and phenotypic intra-species variation in Candida albicans. Genome research, 25(3), 413-425. http://www.genome.org/cgi/doi/10.1101/gr.174623.114

Hlaváčová J., Flegr J., Řežábek K., Calda P., Kaňková Š. (2021). Association between latent toxoplasmosis and fertility parameters of men. Andrology, 9, 854–862. https://doi.org/10.1111/andr.12969

Hodková H., Kolbeková P., Skallová A., Lindová J., & Flegr J. (2007). Higher perceived dominance in Toxoplasma infected men – a new evidence for role of increased level of testosterone in toxoplasmosis-associated changes in human behaviour. Neuroendocrinology Letters, 28, 110–114.

Hosseini, S.A., Amouei, A., Sharif, M., et al. (2019). Human toxoplasmosis: a systematic review for genetic diversity of Toxoplasma gondii in clinical samples. Epidemiology and Infection, 147:e36. https://doi.org/10.1017/S0950268818002947

Hughes, H.K., & Ashwood, P. (2018). “Anti-Candida albicans IgG Antibodies in Children with Autism Spectrum Disorders.” Frontiers in Psychiatry, 9, 627. https://doi.org/10.3389/fpsyt.2018.00627

Jellinger, K.A., Setinek, U., Drlicek, M., Böhm, G., Steurer, A., & Lintner, F. (2000). Neuropathology and General Autopsy Findings in AIDS during the Last 15 Years. Acta Neuropathologica, 100(2), 213-220. https://doi.org/10.1007/s004010000245

Jones, J. L., Kruszon-Moran, D., Elder, S., Rivera, H. N., Press, C., Montoya, J. G., & McQuillan, G. M. (2017). Toxoplasma gondii infection in the United States, 2011–2014. The American Journal of Tropical Medicine and Hygiene, 98(2), 551. https://doi.org/10.4269/ajtmh.17-0677

Jong, A.Y., Stins, M.F., Huang, S-H., Chen, S.H.M., & Kim, K.S. (2001). Traversal of Candida albicans across Human Blood-Brain Barrier In Vitro. Infection and Immunity, 69(7), 4536-4544. https://doi.org/10.1128/iai.69.7.4536-4544.2001

Kaňková Š., Hlaváčová J., Flegr J. (2020). Oral sex: a new, and possibly the most dangerous, route of toxoplasmosis transmission. Medical Hypotheses, 141, 109725. https://doi.org/10.1016/j.mehy.2020.109725

Kurtzke, J. F., Hyllested, K., Heltberg, A., & Olsen, A. (1993). Multiple sclerosis in the Faroe Islands. 5. The occurrence of the fourth epidemic as validation of transmission. Acta neurologica scandinavica, 88(3), 161-173.

Latifi, A., Flegr, J., & Kaňková, Š. (2025). Re-assessing host manipulation in Toxoplasma: the underexplored role of sexual transmission–evidence, mechanisms, implications. Folia Parasitologica, 72, 015. http://dx.doi.org/10.14411/fp.2025.015

Li, J., Fan, S. R., Liu, X. P., Li, D. M., Nie, Z. H., Li, F., ... & Bai, F. Y. (2008). Biased genotype distributions of Candida albicans strains associated with vulvovaginal candidosis and candidal balanoposthitis in China. Clinical Infectious Diseases, 47(9), 1119-1125. https://doi.org/10.1086/592249

Lin, H., Chen, Y., Abror-Lacks, G. et al. (2024). Sexual behavior is linked to
changes in gut microbiome and systemic inflammation that lead to HIV-1
infection in men who have sex with men. Communications Biology, 7, 1145.
https://doi.org/10.1038/s42003-024-06816-z

Mackey, W. C., & Immerman, R. S. (2001). Restriction of sexual activity as a partial function of disease avoidance: A cultural response to sexually transmitted diseases. Cross-Cultural Research, 35(4), 400-423. https://doi.org/10.1177/106939710103500403

Mackey, W. C., & Immerman, R. S. (2003). New ideas/viewpoints: A proposed feedback loop of sexually transmitted diseases and sexual behavior: The red queen’s dilemma. Social Biology, 50(3-4), 281-299. https://doi.org/10.1080/19485565.2003.9989077

Masliah, E., DeTeresa, R.M., Mallory, M.E., & Hansen, L.A. (2000). Changes in Pathological Findings at Autopsy in AIDS Cases for the Last 15 Years. AIDS, 14 (1), 69-74. https://doi.org/10.1097/00002030-200001070-00008

Mastrorosa, I., Pinnetti, C., Brita, A. C., Mondi, A., Lorenzini, P., Del Duca, G., ... & Antinori, A. (2023). Declining prevalence of human immunodeficiency virus (HIV)–associated neurocognitive disorders in recent years and associated factors in a large cohort of antiretroviral therapy–treated individuals with HIV. Clinical Infectious Diseases, 76(3), e629-e637. https://doi.org/10.1093/cid/ciac658

McArthur, J.C., & Brew, B.J. (2010). HIV-associated Neurocognitive Disorders: Is There a Hidden Epidemic? AIDS, 24(9), 1367-1370. https://doi.org/10.1097/qad.0b013e3283391d56

Meyer-Arndt, L., Kuchling, J., Brasanac, J., Hermann, A., Asseyer, S., Bellmann-Strobl, J., ... & Weygandt, M. (2022). Prefrontal-amygdala emotion regulation and depression in multiple sclerosis. Brain Communications, 4(3), fcac152. https://doi.org/10.1093/braincomms/fcac152

Mohammed, A., Ahmed, M., & Ibrahim, N. (2024). The global seroprevalence of Toxoplasma gondii infection in workers occupationally exposed to animals (1972–2023): a systematic review and meta-analysis. The Veterinary Quarterly, 44(1), 1. https://doi.org/10.1080/01652176.2024.2396577

Pinto, B., Mattei, R., Moscato, G. A., Cristofano, M., Giraldi, M., Scarpato, R., ... & Bruschi, F. (2017). Toxoplasma infection in individuals in central Italy: does a gender-linked risk exist? European Journal of Clinical Microbiology & Infectious Diseases, 36(4), 739-746. https://doi.org/10.1007/s10096-016-2857-8

Pisa, D., Alonso, R., Rábano, A., & Carrasco, L. (2016). Corpora amylacea of brain tissue from neurodegenerative diseases are stained with specific antifungal antibodies. Frontiers in Neuroscience, 10, 86. https://doi.org/10.3389/fnins.2016.00086

Poirotte C., Kappeler P.M., Ngoubangoye B., Bourgeois S., Moussodji M., & Charpentier M.J.E. (2016). Morbid attraction to leopard urine in Toxoplasma-infected chimpanzees. Current Biology, 26, R98–R99. https://doi.org/10.1016/j.cub.2015.12.020

Read A., & Braithwaite V. (2012). Afterword. In: Host Manipulation by Parasites, (eds) D. P. Hughes, J. Brodeur, & F. Thomas (pp. 195-197), Oxford(United Kingdom): Oxford University Press.

Reed, B. D., Zazove, P., Pierson, C. L., Gorenflo, D. W., & Horrocks, J. (2003). Candida transmission and sexual behaviors as risks for a repeat episode of Candida vulvovaginitis. Journal of Women's Health, 12(10), 979-989. https://doi.org/10.1089/154099903322643901

Romero-Pinel, L., Bau, L., Matas, E., León, I., Muñoz-Vendrell, A., Arroyo, P., ... & Martínez-Yélamos, S. (2022). The age at onset of relapsing-remitting multiple sclerosis has increased over the last five decades. Multiple Sclerosis and Related Disorders, 68, 104103. https://doi.org/10.1016/j.msard.2022.104103

Schmid, J., Hunter, P. R., White, G. C., Nand, A. K., & Cannon, R. D. (1995). Physiological traits associated with success of Candida albicans strains as commensal colonizers and pathogens. Journal of Clinical Microbiology, 33(11), 2920-2926. https://doi.org/10.1128/jcm.33.11.2920-2926.1995

Schmid, J., Rotman, M., Reed, B., Pierson, C. L., & Soll, D. R. (1993). Genetic similarity of Candida albicans strains from vaginitis patients and their partners. Journal of clinical microbiology, 31(1), 39-46. https://doi.org/10.1128/jcm.31.1.39-46.1993

Tian, J. Y., Yang, Y. G., Chen, S., Teng, Y., & Li, X. Z. (2021). Genetic diversity and molecular epidemiology of Candida albicans from vulvovaginal candidiasis patients. Infection, Genetics and Evolution, 92, 104893. https://doi.org/10.1016/j.meegid.2021.104893

Tong, W. H., Hlaváčová, J., Abdulai-Saiku, S., Kaňková, Š., Flegr, J., & Vyas, A. (2023). Presence of Toxoplasma gondii tissue cysts in human semen: Toxoplasmosis as a potential sexually transmissible infection. Journal of Infection, 86(1), 60-65. https://doi.org/10.1016/j.jinf.2022.10.034

Tuddenham, S., Koay, W. L., & Sears, C. (2020). HIV, sexual orientation, and gut microbiome interactions. Digestive diseases and sciences, 65(3), 800-817. https://doi.org/10.1007/s10620-020-06110-y

van den Berghe, P. L. (1979). Human Family Systems. An Evolutionary View. New York: Elsevier. https://archive.org/details/humanfamilysyste0000vand

Vyas, A., Kim, S. K., Giacomini, N., Boothroyd, J. C., & Sapolsky, R. M. (2007). Behavioral changes induced by Toxoplasma infection of rodents are highly specific to aversion of cat odors. Proceedings of the National Academy of Sciences, 104(15), 6442-6447. https://doi.org/10.1073/pnas.0608310104

Warszawski, J., Meyer, L., & Bajos, N. (1996). Is genital mycosis associated with HIV risk behaviors among heterosexuals? ACSF Investigators. Analyse des comportements sexuels en France. American Journal of Public Health, 86(8_Pt_1), 1108-1111. https://doi.org/10.2105/AJPH.86.8_Pt_1.1108

Wikipedia. (2025). “HIV/AIDS Public Health Campaigns in Italy” https://en.wikipedia.org/wiki/HIV/AIDS_Public_Health_Campaigns_in_Italy

Wilking, H., Thamm, M., Stark, K. et al. (2016). Prevalence, incidence estimations and risk factors of Toxoplasma gondii infection in Germany: a representative, cross-sectional, serological study. Scientific Reports, 6, 22551. https://doi.org/10.1038/srep22551

Xiao, J., & Yolken, R. H. (2015). Strain hypothesis of Toxoplasma gondii infection on the outcome of human diseases. Acta Physiologica, 213(4), 828. https://doi.org/10.1111/apha.12458

You’ve got your mother’s big blue eyes! Irving Berlin, 1913 (Wikicommons)

Archaeogeneticist Davide Piffer has reviewed the ancient DNA evidence on the origin of blue eyes among Europeans. The earliest evidence comes from an individual dated to around 35,000 years ago in the Crimea. This was during the last ice age, when Europeans were nomadic hunter-gatherers (Piffer, 2025).

But a lone individual may be just that. Davide concludes that blue eyes were confined to specific hunter-gatherer groups during glacial times. They then became much more common after the last ice age ended some 12,000 years ago.

Davide goes on to argue that blue eyes then became much less common when European hunter-gatherers were replaced by brown-eyed farmers from the Middle East. Afterwards, blue eyes quickly returned to their formerly high prevalence.

Here is my take on his argument:

Scarcity of aDNA from glacial times (>12,000 years ago)

Blue eyes are poorly attested from the ice age because so little DNA has survived from that time. Ice-age hunter-gatherers had a low population density, and many of their sites were destroyed by glaciation and glacial meltwaters. Post-glacial sites provide more DNA and more information on eye color. If we examine individuals who lived shortly after the ice age in Germany, Scandinavia, and the East Baltic, we find that half or a little over half were blue-eyed (Günther et al., 2018; Mittnik et al., 2018; Posth et al., 2023).

In sum, blue eyes were already common in northern Europe when the ice age ended and were probably common earlier.

Overestimation of population replacement by Anatolian farmers

Davide Piffer, like many archaeogeneticists, believes that the indigenous hunter-gatherers of Europe were considerably replaced by brown-eyed farmers from Anatolia. This belief leads him to argue that strong selection caused blue eyes to bounce back to their former prevalence.

Such a scenario requires rapid evolutionary change. Farming began to spread into northern Europe scarcely 6,000 years ago and into Finland and the East Baltic less than 4,000 years ago — and on a limited scale. The last date is already within the time of recorded history, and only a millennium before the Greek poet Homer mentioned blue eyes in his works.

There is no doubt that farmers did spread out of Anatolia and into Europe. But there is reason to doubt the claim that northern Europeans are 30 to 50% of Anatolian origin. This estimate assumes that any genetic change across the time boundary between hunter-gatherers and farmers is due to population replacement of the former by the latter. One such change is the loss of haplogroup U, which was common among Europeans before farming and is now found only among the Sami of Finland and the Mansi of Siberia, both of whom were hunter-gatherer-fishers until recently (Derbeneva et al., 2002). As expected, genomes from central and western Europe show a sharp break in the population frequency of haplogroup U at the time boundary between hunter-gatherers and farmers (Bramanti et al., 2009).

Yet things are not always what they seem. In Denmark, haplogroup U persisted at high frequencies long after the transition to farming — as late as the Early Iron Age (Melchior et al., 2010). In Latvia and Ukraine it persisted into the Neolithic (Jones et al., 2017). Is its disappearance better explained by natural selection? Haplogroup U shifts the body’s energy balance to heat production — a useful adaptation for hunter-gatherers when sleeping in makeshift shelters or pursuing game in all kinds of weather (Balloux et al., 2009). It would have been less useful for farmers, who slept in a warmer environment and could more easily plan their outdoor activities.

In addition to genetic changes due to differences in natural selection, there were also random changes in allele frequencies at the time boundary between hunter-gatherers and farmers. These are founder effects: only some hunter-gatherers adopted farming, and their allele frequencies would have differed somewhat from those who didn’t — again, for purely random reasons. Some of these stochastic changes create false similarities to the Anatolian genetic profile, thus inflating the estimated Anatolian ancestry of early farmers.

Because of these false positives, and because of convergent natural selection, we overestimate the Anatolian contribution to the North European gene pool, while underestimating the indigenous contribution.

Why blue eyes?

We know the answers to the “when” and the “where.” Blue eyes became common during the last ice age within a region encompassing Germany, Scandinavia, the East Baltic, and probably areas farther east.

At that time, Scandinavia and the Alps were under ice. Northern Europe was habitable only on the plains stretching from northern Germany eastwards. Before 12,000 years ago, these plains were steppe-tundra with wandering herds of reindeer and nomadic bands of hunter-gatherers. Actually, they were just hunters. There were few opportunities for gathering fruits, nuts, tubers, or edible greens. Food was almost entirely “meat on the hoof” (Hoffecker, 2002, pp. 8, 178, 193-194, 237).

But why would such an environment favor blue eyes? Davide offers four possible reasons:

• Lower UV exposure requiring less melanin protection

• Sexual selection for distinctive traits

• Genetic drift in smaller northern populations

• Need for lighter skin to maintain vitamin D synthesis where sunlight is weaker

Because their skin was white?

The first and last reasons imply a strong linkage between eye color and skin color. If selection causes skin to turn white, it should also cause eyes to turn blue. Actually, there is no such linkage. Pale skin often co-exists with dark eyes and dark hair, as noted by anthropologist Alice Brues:

If we were to take all the human beings in the world who have dark brown eyes and black or dark brown hair, we would not only have the vast majority of the human species, but would have a group which shows virtually the complete range of human skin color, from black to almost completely depigmented. (Brues, 1975)

A genome-wide association study of 192,986 Europeans has similarly concluded that eye color is not significantly associated with skin or hair color:

Although DNA variants within the MC1R gene are strongly associated with light skin and red hair color, no detectable association with eye color was found in our large GWAS, in line with previous albeit smaller-sized GWASs of more limited statistical power. Similarly, other DNA variants strongly associated with skin and hair color within genes, such as SILV, ASIP, and POMC, showed no statistically significant effect on eye color in this study, nor in previous studies. Moreover, we also identified 34 genetic loci that were significantly associated with eye color, but for which there is no report of significant association with hair and/or skin color. (Simcoe et al., 2021)

We might find a significant association by examining millions of Europeans, but it would be very weak and could hardly explain such a dramatic increase in the prevalence of blue eyes.

Because of genetic drift?

This explanation would be plausible if the change concerned the frequency of only one allele. To date, researchers have identified 124 alleles that together account for 53% of the diversity in European eye color (Simcoe et al., 2021). This is a massive increase in the number of alleles from the single allele that initially prevailed in ancestral Europeans and still does in most humans.

The word “diversity” is significant here. This is not simply the replacement of one color by another. It is a shift from one color to many. Eyes took on a broad range of hues, not only brown but also blue, gray, hazel, and green.

Only some sort of selection could have caused so many new colors to become so prevalent in so short a time — at most, the time since the arrival of modern humans in northern Europe some 35,000 years ago. Genetic drift seems even less likely if we consider that the diversification of eye color was paralleled by a diversification of hair color within the same region and timeframe — over 200 alleles for various permutations of black, brown, flaxen, blonde, and red (Morgan et al., 2018). In fact, the number may be even higher: a recent forensic model used 2,997 and 4,954 SNPs respectively to predict the existence of blonde and brown hair (Cabrejas-Olalla et al., 2025).

What are the chances of two color polymorphisms, one for the eyes and one for the hair, arising at different genes within the same limits of time and space? It also seems more than a matter of chance that two color polymorphisms would arise on or near the face — which is crucial for visual recognition.

Because of sexual selection?

The new colors are also brighter and purer than the original brown and black. A brighter color reflects more light. A purer color reflects light within a narrower slice of the visible spectrum. Bright, pure colors are typically created by selection of one sort or another to attract attention, be it from a pollinating bee or a potential mate.

All of this is consistent with sexual selection, which occurs when too many of one sex have to compete for too few of the other. What could have caused such an imbalance on the European steppe-tundra of the last ice age?

There were two relevant factors:

• Low polygyny rate. Food was largely meat from herbivores, mostly reindeer. Because men were the hunters, they alone had to provision their wives and children with sufficient food. Only a very capable hunter could provide for two wives and their offspring. If single women were too numerous, only a few could pair up with a polygynous male.

• High male death rate. Game animals had to be pursued over long distances: “hunter-gatherers in northern continental environments … must forage across large areas in order to secure highly dispersed and mobile prey” (Hoffecker, 2002, p. 8). Again, men were the hunters, and their risk of death was a function of hunting distance. The longer the distance, the higher their risk of death from drowning, starvation, exposure, lack of shelter, and other hunting-related accidents.

Together, these two factors led to a surplus of unmated females. Too many women had to compete for too few men. Women were thus under strong selection for colorful features that could attract male attention, including new hair and eye colors (Frost, 2006; Frost, 2022).

Today, hair and eye colors are most diverse within a region that roughly corresponds to the European steppe-tundra of the last ice age, if one allows for the cover of ice over Scandinavia and the Alps. Yet this steppe-tundra also extended into parts of North Asia. Why, then, do indigenous North Asians have only black hair and brown eyes?

The reason may be that the steppe-tundra was colder and drier in North Asia than in Europe, being farther north and further removed from the moderating influence of the Gulf Stream. The Asian steppe-tundra thus supported much smaller human populations, which frequently died out — especially during the glacial maximum. The effects of sexual selection were thus repeatedly reset to zero. Also, in a smaller population, selection has to wait longer for suitable alleles to appear through mutation.

Nonetheless, we do see frequent cases of light hair in certain indigenous groups of North Asia and even Arctic North America (Frost, 2009).

Steppe-tundra did exist in North Asia during the last ice age, but it was farther north, colder, and drier.

Women more often have the “new” hair and eye colors

If sexual selection of women created these colors, wouldn’t women be more prone to having them? This is indeed the case with natural hair color. Red hair is highly overrepresented among women, followed by blond hair and light brown hair, whereas black hair is three to five times underrepresented (Hysi et al., 2018; Shekar et al., 2008).

For eye color, green and hazel are overrepresented among women while brown and blue are underrepresented (Frost, Kleisner, & Flegr, 2017). Why are blue eyes less frequent among women? Isn’t it a “new” color?

It looks like sexual selection favored not only bright and pure colors but also rare ones. There seems to be a novelty effect: the less frequent an eye or hair color, the more it attracts attention. Thus, as new colors arose through mutation, they benefitted from their novelty and increased in frequency at the expense of more common colors. The colors green and hazel appear to have arisen later as a novel form of blue eye color; hence, a greater proportion of blue eyes are expressed in women as green or hazel.

This novelty effect is described in Darwin’s discussion of sexual selection: “It would even appear that mere novelty, or slight changes for the sake of change, have sometimes acted on female birds as a charm, like changes of fashion with us” (Darwin, 1936[1888], p. 813). Like other aspects of visual merchandising, it is most important in saturated markets that offer too many interesting choices among products of equal quality (Lea-Greenwood, 1998).

This rare-color advantage has been reported by several studies.

• An American research team showed pictures of attractive women to male participants and asked them to choose the one they most wanted to marry. There were three series of pictures: the first had equal numbers of brunettes and blondes; the second had one brunette for every five blondes; and the third had one brunette for every eleven blondes. The scarcer the brunettes were in a series, the more often they were chosen (Thelen, 1983).

• Maxim cover girls are likelier to be light blonde or dark brown than the more common dark blond or light brown of real life (Anon, 2008a).

• Viennese women tend to change their hair to a less common color (Schweder, 1994).

• In Brazil and the United Kingdom, modeling agencies prefer women with less frequent eye colors (Forti & Young, 2016).

• On American TV, women are four and a half times likelier than men to have red or auburn hair and five times likelier to have blonde hair (Davis, 1990).

• On Turkish TV, women more often have red or blonde hair (Ikizler, 2007, p. 39).

• Blonde Playboy playmates have become more frequent since the mid-1960s, increasing from a third to over half the total (Anon, 2008b). This trend may be driven by the declining prevalence of blondes in the American population.

A novelty effect can also be seen in the way a preference for one paint color for home interiors would rise until it became satiated and then give way to preference for another (Stansfield & Whitfield, 2005).

However, no such rare-color advantage was found in a replication of the original American study. It should be noted that the replication differed from the original in two ways: participants were recruited online and made their choices privately on their home computers, with no control over whatever female images they had viewed previously or might still be viewing within sight of their computer screen (Janif et al., 2015).

Less frequent hair colors seem to be chosen more often for Maxim cover girls (Anon, 2008a).

Estrogen facilitates fetal development of European hair and eye colors

This sex difference in hair and eye colors may have a hormonal cause. Specifically, estrogen seems to favor the expression of non-black hair and non-brown eyes during fetal development.

An estrogenic effect is especially likely for red hair, which is the hair color that differs the most in prevalence between men and women. According to a survey of 7,000 Czech participants, male redheads are as healthy as other men, doing better on average in three categories and worse in three. Female redheads, however, do worse on average than other women in ten categories and better in only three. They are especially prone to four types of cancer: colorectal, cervical, uterine, and ovarian — three of which are estrogen-dependent. Being both female and red-haired seems to maximize the risk of developing an estrogen-dependent disease, probably because these risk factors are both associated with higher estrogen levels in a developing fetus (Frost et al., 2017).

A higher degree of fetal estrogenization may explain why several studies have found an association between blue eyes and certain feminine traits. [Note: although blue eyes are actually less frequent in women than in men, these studies used a broad definition of “blue” that included green and hazel, which are more frequent in women].

• Blue-eyed boys tend to be shy. This is the “little boy blue” effect. A study of preschoolers found more social wariness in blue-eyed boys than in brown-eyed boys. The difference was greatest at the extremes of wariness. Among the very inhibited boys, 13 out of 14 were blue-eyed. Among the very uninhibited, only 4 out of 10 were. There was no such relationship among the girls, whose eyes were blue in 5 out of 9 among the very inhibited and in 6 out of 11 among the very uninhibited (Coplan et al., 1988).

• Blue-eyed women tend to have narrower shoulders and higher hip-to-waist ratios. A Latvian study found small but significant correlations between female eye color and certain sexually dimorphic features. Shoulders were narrower and hip-to-waist ratios higher in blue-eyed women than in brown-eyed women (Kažoka & Vētra, 2011).

• Blue-eyed men tend to have more feminine faces. This was an unintended finding of two Czech studies whose participants were asked to rate male and female facial photos. Initially, the brown-eyed male faces were rated as more dominant than the blue-eyed male faces. When, as a control, the brown-eyed faces were photoshopped to make them blue-eyed, they were still rated as more dominant. On careful examination, the originally brown-eyed faces were found to be more masculine with broader and more massive chins, broader mouths, larger noses, larger eyebrows, and closer-set eyes. The originally blue-eyed faces had smaller and sharper chins, narrower mouths, smaller noses, and greater distance between the eyes.

  Blue eyes were associated with a more feminine facial shape only when the faces were male, just as blue eyes were associated with shyness only in boys. Perhaps natural selection has ensured that female fetuses will always receive enough estrogen for their physical and behavioral development. Male fetuses may be exposed to a range of estrogen levels with more variable effects (Kleisner et al., 2010; Kleisner et al., 2013).

  Were brown eyes associated with a different facial shape in this study because some of the brown-eyed men were partly Jewish or Roma and had a more Mediterranean appearance? In that case, facial shape would have been more variable in the brown-eyed men. It was not. This explanation also fails to explain the effect of gender: why were blue eyes associated with facial feminization in men but not in women?

The sex hormones seem to affect hair color not only in the womb but also at puberty. Alleles for darker hair color are associated with earlier puberty in men, and a similar but weaker association exists in women. This looks like a testosterone effect (Hollis et al., 2020). There may thus have been some counter-selection acting on men to resist changes to their physical appearance due to spill-over from sexual selection acting on women.

Eventually, sexual selection would have produced more contrasting sex differences in hair and eye color. It didn’t because it occurred over a relatively brief period, essentially between the entry of modern humans into northern Europe some 35,000 years ago and the end of the ice age some 12,000 years ago. Briefer still was the period of intense sexual selection — the glacial maximum between 20,000 and 15,000 years ago.

Averaged faces of blue-eyed men (left) and brown-eyed men (right), Czech population (Kleisner et al., 2010).

White skin

What about the strange albino-like skin of Europeans? Could it also be due to sexual selection?

The counter-argument is that skin became white at high European latitudes to maintain sufficient synthesis of vitamin D. But why, then, did it fail to whiten to the same degree among indigenous peoples at the same latitudes in Asia and North America? Even in Europe, modern humans remained brown-skinned long after they arrived in the continent. Three research teams have estimated that white skin arose somewhere in Europe no earlier than 19,000 years ago (Canfield et al., 2014; Beleza et al., 2013; Norton and Hammer, 2007). In western Europe, Europeans remained brown-skinned until almost the dawn of history, as shown by DNA dated to 8,000 years ago from Luxembourg, 7,000 years ago from Spain, and as late as 5,000-4,000 years ago for some people in England (Brace et al., 2019; Lazaridis et al., 2014; Olalde et al., 2014).

One could argue that it was the change from hunting to farming that caused skin color to turn white. With Europeans consuming less meat, their bodies required more vitamin D. There is in fact evidence that meat contains a cofactor that independently reduces the risk of rickets (Dunnigan et al., 2005). But we still have to explain why some European populations were white-skinned long before farming. Two ancient DNA studies have shown a high prevalence of white skin in Scandinavians 9,500 to 6,000 years ago and in East Baltic peoples 7,460 to 5,360 years ago (Günther et al., 2018; Mittnik et al., 2018).

So what made Europeans white? Probably the same sexual selection that made their hair and eyes diversely colored. Unlike their hair and eyes, however, their skin did not acquire a variety of colors. It just became pale. This may be because sexual selection for light skin was guided by a pre-existing sexual dimorphism; i.e., women are lighter-colored than men across all human populations (Edwards & Duntley, 1939; Edwards & Duntley, 1949; Edwards et al., 1941; Frost, 2023; Manning et al., 2004). Lighter-skinned women are thus seen as more feminine in traditional cultures and preferred as mates (van den Berghe & Frost, 1986). When sexual selection is sufficiently strong, it should drain the gene pool of alleles for dark skin.

Ultimately, women have a lighter skin for the same reason they have a smaller nose and chin, a smoother, more pliable skin, and a higher pitch of voice. These are visual, tactile, and auditory cues that originally identified the human infant to an adult observer, who would respond by being less aggressive and more willing to provide care and nurturance. By mimicking those cues, women could strengthen the pair-bond and better ensure male provisioning (Frost, 2010, pp. 134-135; Frost, 2011; Frost, 2025; Lorenz, 1971, pp. 154-164).

Is the European face primarily a woman’s face?

When modern humans arrived in northern Europe some 35,000 years ago, they looked much like other modern humans elsewhere. They then became more and more different-looking, not through selection by the natural environment but through selection by the limited supply of men. On the steppe-tundra of the last ice age, the high male death rate and the high cost of polygyny for men created a continual surplus of unmated women. The resulting increase in sexual selection favored those women who could better catch male attention.

Sexual selection probably peaked during the glacial maximum some 20,000 to 15,000 years ago. Although women were the ones being selected, the effects spilled over onto men due to the weak sex-linkage of skin, hair, and eye color. Both sexes thus acquired the white skin and bright hair and eye colors that initially arose among women.

This sexual selection ended with the end of the last ice age. For reasons unrelated to their physical appearance, these pale humans would spread across Europe during post-glacial times, pushing west and south out of their original territory and eventually replacing their darker cousins throughout the continent. Thus, on the eve of recorded history, all Europeans now possessed a unique look that would later define them, as if they were a cast of actors being hastily made up and rushed onto stage moments before curtain time.

References

Anon. (2008a). Maxim’s audience prefers brunettes; distribution is bimodal. Gene Expression, July 6. http://www.gnxp.com/blog/2008/07/maxims-audience-prefers-brunettes.php

Anon. (2008b). Bygone brunette beauty: Fashion in hair color. Gene Expression, June 29. http://www.gnxp.com/blog/2008/06/bygone-brunette-beauty-fashion-in-hair.php

Balloux F., Handley, L.J., Jombart, T., Liu, H., & Manica, A. (2009). Climate shaped the worldwide distribution of human mitochondrial DNA sequence variation. Proceedings of the Royal Society B. Biological Sciences, 276, 3447-3455. https://doi.org/10.1098/rspb.2009.0752

Beleza, S., Santos, A.M., McEvoy, B., Alves, I., Martinho, C., Cameron, E., et al. (2013). The timing of pigmentation lightening in Europeans. Molecular Biology and Evolution, 30(1), 24-35. https://doi.org/10.1093/molbev/mss207

Brace, S., Diekmann, Y., Booth, T.J., Faltyskova, Z., Rohland, N., Mallick, S., et al. (2019). Ancient genomes indicate population replacement in Early Neolithic Britain. Nature Ecology & Evolution, 3(5), 765-771. https://doi.org/10.1038/s41559-019-0871-9

Bramanti, B., Thomas, M.G., Haak, W., Unterlaender, M., Jores, P., Tambets, K., Antanaitis-Jacobs, I., Haidle, M.N., Jankauskas, R., Kind, C.J., et al. (2009). Genetic discontinuity between local hunter-gatherers and Central Europe’s first farmers. Science, 326, 137-140. https://doi.org/10.1126/science.1176869

Brues, A.M. (1975). Rethinking human pigmentation. American Journal of Physical Anthropology, 43(3), 387-391. https://doi.org/10.1002/ajpa.1330430320

Cabrejas-Olalla, A., Jørgensen, F. G., Cheng, J. Y., Kjærgaard, P. C., Schierup, M. H., Mailund, T., & Athanasiadis, G. (2025). Genetic predictions of eye and hair colour in the Danish population. Forensic Science International: Genetics, 78, 103267. https://doi.org/10.1016/j.fsigen.2025.103267

Canfield, V.A., Berg, A., Peckins, S., Wentzel, S.M., Ang, K.C., Oppenheimer, S., & Cheng, K.C. (2014). Molecular phylogeography of a human autosomal skin color locus under natural selection. G3, 3(11), 2059-2067. https://doi.org/10.1534/g3.113.007484

Coplan, R., Coleman, B., & Rubin, K. (1998). Shyness and little boy blue: Iris pigmentation, gender, and social wariness in preschoolers. Developmental Psychobiology, 32(1), 37-44. https://doi.org/10.1002/(SICI)1098-2302(199801)32:1<37::AID-DEV4>3.0.CO;2-U

Darwin, C. (1936 [1888]). The Descent of Man and Selection in relation to Sex. reprint of 2nd edition, The Modern Library, New York: Random House.

Davis, M. D. (1990). Portrayals of women in prime-time network television: Some demographic characteristics. Sex Roles, 23(5-6), 325-332. http://psycnet.apa.org/doi/10.1007/BF00290052

Derbeneva, O.A., Starikovskaya, E.B., Wallace, D.C., & Sukernik, R.I, (2002). Traces of early Eurasians in the Mansi of Northwest Siberia revealed by mitochondrial DNA analysis. American Journal of Human Genetics, 70, 1009-1014. https://doi.org/10.1086/339524

Dunnigan, M.G., Henderson, J.B., Hole, D.J., Mawer, E.B., & Berry, J.L. (2005). Meat consumption reduces the risk of nutritional rickets and osteomalacia. British Journal of Nutrition, 94, 983–991. https://doi.org/10.1079/BJN20051558 .

Edwards, E.A., & Duntley, S.Q. (1939). The pigments and color of living human skin. American Journal of Anatomy, 65(1), 1-33. https://doi.org/10.1002/aja.1000650102

Edwards, E.A., & Duntley, S.Q. (1949). Cutaneous vascular changes in women in reference to the menstrual cycle and ovariectomy. American Journal of Obstetrics & Gynecology, 57(3), 501-509. https://doi.org/10.1016/0002-9378(49)90235-5

Edwards, E.A., Hamilton, J.B., Duntley, S.Q., & Hubert, G. (1941). Cutaneous vascular and pigmentary changes in castrate and eunuchoid men. Endocrinology, 28(1), 119-128. https://doi.org/10.1210/endo-28-1-119

Forti, I.R.N., & Young, R.J. (2016). Human commercial models’ eye colour shows negative frequency-dependent selection. PLoS One, 11(12) e0168458. https://doi.org/10.1371/journal.pone.0168458

Frost, P. (2006). European hair and eye color - A case of frequency-dependent sexual selection? Evolution and Human Behavior, 27(2), 85-103. https://doi.org/10.1016/j.evolhumbehav.2005.07.002

Frost, P. (2009). Blond Inuit? Evo and Proud, January 22. https://evoandproud.blogspot.com/2009/01/blond-inuit.html

Frost, P. (2010). Femmes claires, hommes foncés. Les racines oubliées du colorisme. Quebec City: Les Presses de l’Université Laval, 202 p. https://www.pulaval.com/livres/femmes-claires-hommes-fonces-les-racines-oubliees-du-colorisme

Frost, P. (2022). European Hair, Eye, and Skin Color: Solving the Puzzle. Washington: Academica Press. ISBN 9781680538724

Frost, P. (2025). Emotional responses to the differing skin tones of men and women. Peter Frost’s Newsletter, April 3.

Frost, P., Kleisner, K., & Flegr, J. (2017). Health status by gender, hair color, and eye color: Red-haired women are the most divergent. PLoS One, 12(12), e0190238. https://doi.org/10.1371/journal.pone.0190238

Günther, T., H. Malmström, E.M. Svensson, A. Omrak, F. Sánchez-Quinto, G.M. Kilinç, et al. (2018). Population genomics of Mesolithic Scandinavia: Investigating early postglacial migration routes and high-latitude adaptation. PLoS Biol, 16(1), e2003703. https://doi.org/10.1371/journal.pbio.2003703

Hoffecker, J.F. (2002). Desolate landscapes. Ice-age settlement in Eastern Europe. New Brunswick: Rutgers University Press.

Hollis, B., Day, F. R., Busch, A. S., Thompson, D. J., Soares, A. L. G., Timmers, P. R., ... & Wilson, C. H. (2020). Genomic analysis of male puberty timing highlights shared genetic basis with hair colour and lifespan. Nature communications, 11(1), 1536. https://doi.org/10.1038/s41467-020-14451-5

Hysi, P.G., Valdes, A.M., Liu, F., Furlotte, N.A., Evans, D.M., Bataille, V., et al. (2018). Genome-wide association meta-analysis of individuals of European ancestry identifies new loci explaining a substantial fraction of hair color variation and heritability. Nature Genetics, 50(5), 652-656. https://doi.org/10.1038/s41588-018-0100-5

Ikizler, A.S. (2007). Gender role representations in Turkish television programs, Submitted as a St. Mary’s Project in Partial Fulfillment of the Graduation Requirements, St. Mary’s College of Maryland for the Degree of Bachelor of Arts in Psychology http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.514.2218

Janif, Z.J., Brooks, R.C., & Dixson, B.J. (2015). Are preferences for women’s hair color frequency-dependent? Adaptive Human Behavior and Physiology, 1(1), 54-71. https://doi.org/10.1007/s40750-014-0008-y

Jones, E.R., Zarina, G., Moiseyev, V., Lightfoot, E., Nigst, P.R., Manica, A., et al. (2017). The Neolithic transition in the Baltic was not driven by admixture with early European farmers, Current Biology, 27(4), 576-582. https://doi.org/10.1016/j.cub.2016.12.060

Kažoka, D. & Vetra, J. (2011). Variations in some anthropometrical parameters of the women with the different iris color in Latvia. Papers on Anthropology, XX, 160-170. https://doi.org/10.12697/poa.2011.20.17

Kleisner, K., Kocnar, T., Rubešová, A., & Flegr, J. (2010). Eye color predicts but does not directly influence perceived dominance in men. Personality and Individual Differences, 49(1), 59-64. https://doi.org/10.1016/j.paid.2010.03.011

Kleisner, K., Priplatova, L., Frost, P., & Flegr, J. (2013). Trustworthy-looking face meets brown eyes. PLoS One, 8(1), e53285. https://doi.org/10.1371/journal.pone.0053285

Lazaridis, I., Patterson, N., Mittnik, A., Renaud, G., Mallick, S., Kirsanow, K., et al. (2014). Ancient human genomes suggest three ancestral populations for present-day Europeans. Nature, 513(7518), 409-413. https://doi.org/10.1038/nature13673

Lea-Greenwood, G. (1998). Visual merchandising: a neglected area in UK fashion marketing? International Journal of Retail & Distribution Management, 26(8), 324-329. https://doi.org/10.1108/09590559810231797

Lorenz, K. (1971). Studies in Animal and Human Behaviour, vol. 2. London: Methuen & Co.

Manning, J.T., Bundred, P.E., & Mather, F.M. (2004). Second to fourth digit ratio, sexual selection, and skin colour. Evolution and Human Behavior, 25(1), 38-50. https://doi.org/10.1016/s1090-5138(03)00082-5

Melchior, L., Lynnerup, N., Siegismund, H.R., Kivisild, T., & Dissing, J. (2010). Genetic diversity among ancient Nordic populations. PLoS One, 5(7), e11898 https://doi.org/10.1371/journal.pone.0011898

Mittnik, A., Wang, C-C., Pfrengle, S., Daubaras, M., Zarina, G., Hallgren, F., et al. (2018). The genetic prehistory of the Baltic Sea region. Nature Communications, 9(442) https://doi.org/10.1038/s41467-018-02825-9

Morgan, M.D., Pairo-Castineira, K. Rawlik, E., Canela-Xandri, O., Rees, J., Sims, D., Tenesa, A., & Jackson, I.J. (2018). Genome-wide study of hair colour in UK Biobank explains most of the SNP heritability. Nature Communications, 9, 5271. https://doi.org/10.1038/s41467-018-07691-z

Norton, H.L., & Hammer, M.F. (2007). Sequence variation in the pigmentation candidate gene SLC24A5 and evidence for independent evolution of light skin in European and East Asian populations. Program of the 77th Annual Meeting of the American Association of Physical Anthropologists, p. 179.

Olalde, I., Allentoft, M.E., Sanchez-Quinto, F., Santpere, G., Chiang, C.W.K., DeGiorgio, M., et al. (2014). Derived immune and ancestral pigmentation alleles in a 7,000-year-old Mesolithic European. Nature, 507(7491), 225-228. https://doi.org/10.1038/nature12960

Piffer, D. (2015). The origins and spread of blue eyes in Europe: Evidence from ancient DNA. PifferPilfer, September 19.

Posth, C., Yu, H., Ghalichi, A., Rougier, H., Crevecoeur, I., Huang, Y., ... & Krause, J. (2023). Palaeogenomics of upper palaeolithic to neolithic European hunter-gatherers. Nature, 615(7950), 117-126. https://doi.org/10.1038/s41586-023-05726-0

Schweder, B.I.M. (1994). The impact of the face on long-term human relationships. Homo, 45(1): 74-93.

Shekar, S.N., Duffy, D.L., Frudakis, T., Montgomery, G.W., James, M.R., Sturm, R.A., & Martin, N.G. (2008). Spectrophotometric methods for quantifying pigmentation in human hair-Influence of MC1R genotype and environment. Photochemistry and Photobiology, 84(3), 719-726. https://doi.org/10.1111/j.1751-1097.2007.00237.x

Simcoe, M., Valdes, A., Liu, F., Furlotte, N.A., Evans, D.M., Hemani, G., et al. (2021). Genome-wide association study in almost 195,000 individuals identifies 50 previously unidentified genetic loci for eye color. Science Advances, 7(11), eabd1239 https://doi.org/10.1126/sciadv.abd1239

Stansfield, J., & Whitfield, T.W.A. (2005) Can future colour trends be predicted on the basis of past colour trends? An empirical investigation. Color Research & Application, 30(3), 235-242. https://doi.org/10.1002/col.20110

Thelen, T.H. (1983). Minority type human mate preference. Social Biology, 30(2), 162-180. https://doi.org/10.1080/19485565.1983.9988531

van den Berghe, P.L., & Frost, P. (1986). Skin color preference, sexual dimorphism and sexual selection: A case of gene-culture co-evolution? Ethnic and Racial Studies, 9(1), 87-113. https://doi.org/10.1080/01419870.1986.9993516

The Enlightenment (Wikicommons - TheErinCool). No longer voices crying in the wilderness, intellectuals became numerous enough to meet in coffeehouses, learned societies, and debating clubs.

Human populations have evolved over time, not only during prehistory but also well into recorded history. This evolution has affected a wide range of mental and behavioral traits: cognitive ability, time preference, propensity for violence, monotony avoidance, rule following, guilt proneness, and empathy, among others.

Such traits vary among human populations because different cultures have imposed different demands on mind and body. In general, the cultural environment will favor those who can better exploit its possibilities, just as the natural environment will. There has thus been a process of coevolution: we make culture, and it remakes us—by selecting those among us who survive to pass on their genes. This coevolution has proceeded along different trajectories in different populations.

One trajectory began during the Early Middle Ages on the northwestern fringe of Europe, where fishing peoples were learning how to use the North Sea for long-distance trade. From such inauspicious beginnings, they would achieve global dominance in a little over a thousand years:

In every respect, the 7th century marked a turning point: the old economic system was in its terminal phase and a new world was beginning to emerge. The accelerated decline of Mediterranean trade in the 7th century was linked by Belgian historian Henri Pirenne (1862-1935) to the Arab invasion; we have seen that the decline goes further back in time, even though Arab expansion in the 7th-8th centuries and Saracen piracy undoubtedly helped reduce trade even more …

But the great change was really the emergence of the North Sea as the main space of international trade with, around the mid-7th century, the birth of a network linking Frisia, England, Scandinavia and the Frankish world. (Chandelier, 2021, p. 192)

The failure of Rome

Why did the North Sea overtake the Mediterranean in international trade? Certainly, the latter region was disrupted by the Arab conquests of the Middle East and North Africa. But the decline in trade began much earlier. Shipwrecks on the bottom of the Mediterranean have been dated overwhelmingly to the time between the 1st century BC and the 1st century AD. Silver mining in Spain and Cyprus likewise fell sharply after the 1st century, as shown by lead contamination of Greenland’s ice sheet (Terpstra, 2019).

The decline occurred not only before the Arab conquests of the 7th and 8th centuries but also before the barbarian invasions of the 4th and 5th and the Imperial Crisis of the 3rd. It thus happened when Roman imperial expansion was at its height. Yet this timing seems inconsistent with modern economic thinking: bigger markets should create economies of scale, as well as a better match between supply and demand. So what caused things to go wrong?

Familialism, cronyism, and nepotism

First, there was the low level of social trust. People trusted only their close friends and relatives, keeping everyone else at arm’s length. As a result, economic activity was bottled up within family networks, with the exception of physical marketplaces—where buyer and seller could meet face to face. Because the market principle remained trapped within small pockets of space and time, it could not generalize to all transactions in Roman society. An economy of markets never evolved into a true market economy.

In sum, the Pax Romana opened up a large space for peaceful exchange of goods and services. Economic activity did increase but remained largely confined within the small high-trust environments of family businesses. Once that source of economic activity had been tapped out, there remained little scope for further growth.

Deterioration of physical health

Familialism, cronyism, and nepotism may explain why the Roman economy grew only to a certain point. But the ensuing decline had at least two other causes.

One was a deterioration of physical health, as indicated by the length of long bones belonging to over 10,000 adult men and women born between 500 BC and 750 AD. The data show a steady decrease from the 2nd century BC, reaching a low point in the second half of the 1st century AD, followed by a slow recovery and then a dramatic recovery from the 5th century (Jongman et al., 2019).

This trend may seem paradoxical, as it inversely correlates with the success of Imperial Rome: physical health deteriorated as the Empire expanded and then recovered as the Empire shrank. The study’s authors concluded that the Romans created not only an integrated Mediterranean economy but also “the first integrated disease regime”.

In health terms, however, the consequences were not necessarily that favourable. Roman cities had become the focal point of viruses and bacteria that all vectored in on them, to find a densely packed population. Historically, a declining biological standard of living under conditions of economic development and increasing economic integration is not unique, of course. (Jongman et al., 2019)

Physical health of adults in the Roman Empire, as indicated by skeletal data (Jongman et al., 2019, Figure 4)

Cognitive decline

The other cause was a decrease in average cognitive ability. Fewer people could master the skills of numeracy, literacy, and budgeting that are so essential to economic activity.

This decline was driven by an uncoupling of reproductive success from economic success. The wealthy were no longer using their wealth to bring children into the world. A rich man might prefer to leave his wife for a younger woman of low social status, often adopting her children. Or he might never marry.

The resulting fall in cognitive ability can be seen in DNA retrieved from human remains in central Italy. People who lived in the Imperial Era had, on average, fewer alleles associated with high cognitive ability than those who lived in the Republican Era (Frost, 2024b; Piffer et al., 2023).

Rise, fall, and renewed rise of mean cognitive ability in central Italy. During the Imperial Era, the Romans lost all of the cognitive gains they had made previously since prehistoric times (Piffer et al, 2023, Figure 1)

We can chart this cognitive decline because of scientific advances in two areas: 1) retrieval of DNA from human remains; and 2) identification of alleles associated with educational attainment (which correlates highly with cognitive ability). While we have not identified enough of these alleles to estimate the intelligence of one individual, we can now reliably estimate the average intelligence of a sample of individuals.

Civilizational decline

As average cognitive ability decreased, so did its manifestations in daily life. The Romans still excelled at directing large numbers of people, often slaves. But they were not so good at innovation. Their medieval successors took much less time to find better ways of doing things:

Unseen in Roman society were the humble wheelbarrow and stirrup, which would not appear until the Middle Ages. Apart from mechanical clocks, medieval Europe invented heavy ploughs, spectacles, windmills, iron-casting, firearms and paper. The first half of the fifteenth century would add the printing press. Well before that time, medieval texts had been published in codex form, a massive improvement on the unwieldly book scroll of Roman antiquity. Along with better carriers of written information, the Middle Ages adopted a consistent system of graphic symbols, greatly facilitating the readability of the Latin script. Improved ship design in the thirteenth century allowed Mediterranean sailing year-round, unlike in Roman times when travel largely ceased during the winter months. … [T]he maritime shippers of the Middle Ages rejected heavy, breakable ceramic vessels as transport containers. Instead, they adopted the more practical wooden barrel, which had “a better volume to weight ratio, more efficient stacking capability, and greater maneuverability.” (Terpstra, 2019)

The success of northwest Europe

How it began

In the 7th century, a new space for trade opened up around the North Sea (Melleno, 2014). These traders were unlike those around the Mediterranean, for whom markets were merely marketplaces—small islands of exchange beyond which people produced for family, kin, or lord. The North Sea traders were the first to break free from an economic model where the individual is embedded in static, long-lasting relationships based on rank, kinship, and locality.

In this, they had an inherent advantage. The North Sea and Baltic peoples are particularly prone to certain mental and behavioral tendencies that, for over a millennium, have prevailed north and west of a line running from Trieste to St. Petersburg, i.e., the “Hajnal line.” These are tendencies toward individualism, the nuclear family, late marriage, and solitary living, as well as a greater willingness to trust strangers and form bonds of impersonal prosociality (Frost, 2025).

Northwest Europeans thus offered the best behavioral conditions for the emergence of a market economy, once this possibility arose. They were best able to expand economic relations beyond the limits of the local kin group—in fact, potentially beyond the limits of any group, no matter how large (Frost, 2020).

The same behavioral conditions would facilitate not only the expansion of the market economy but also that of the State and the Church. Indeed, all three institutions similarly expanded from the 11th century onward. And all three would create a new web of relations that had little to do with kinship, except metaphorically: “brothers and sisters in Christ,” “enfants de la patrie,” etc.

There was a synergy between these parallel expansions. Growth of the market economy was assisted by a State-Church consensus on the need to execute violent males so that law-abiding people could live in peace. By the Late Middle Ages, courts were condemning to death between 0.5 and 1% of all men in each generation, with perhaps just as many dying at the scene of the crime or in prison while awaiting trial. The pool of violent men dried up until most murders occurred under conditions of jealousy, intoxication, or extreme stress. As a result, the homicide rate fell from 20–40 homicides per 100,000 in the Late Middle Ages to 0.5–1 per 100,000 in the mid-20th century (Frost, 2023; Frost & Harpending, 2015).

A person could now get ahead in life through trade and work, rather than theft and plunder. The creation of a pacified environment thus favored the success of those who possessed market-oriented skills—not only literacy, numeracy, and budgeting but also initiative and a belief that one could shape one’s own future. Such people would form the emerging middle class.

The great cognitive advance

Beginning in the 11th century, the English middle class enjoyed a higher rate of natural increase than the lower class, which failed to replace itself demographically. With each generation, the lower class was gradually replaced by downwardly mobile individuals from middle-class families.

Initially, it was the tail trying to wag the dog: the middle class was too small to have much impact on the entire gene pool. By late medieval times, however, it had become sufficiently large to have an impact. Historical economist Gregory Clark has argued that English society now became increasingly middle class, both mentally and behaviorally. "Thrift, prudence, negotiation, and hard work were becoming values for communities that previously had been spendthrift, impulsive, violent, and leisure loving” (Clark, 2007; Clark, 2009; Clark, 2023; Frost, 2022c).

Sociologist Georg Oesterdiekhoff has argued for a similar evolution across Western Europe as a whole. He views this evolution in terms of Jean Piaget’s stages of cognitive development, i.e., more and more people could go beyond preoperational thinking (egocentrism, anthropomorphism, animism) and achieve operational thinking (ability to understand probability, cause and effect, and another person’s perspective) (Oesterdiekhoff, 2023; Rindermann, 2018, pp. 49, 86-87).

The earlier way of thinking is described by psychologist Heiner Rindermann:

There were trials against animals up to the sixteenth century, e.g. rats were ordered to appear in court for having eaten major parts of the harvest. Celestial bodies were seen as conscious beings that could be communicated with and which influenced people’s lives. (Rindermann, 2018, p. 87)

DNA studies

We can now verify this model of recent Western European evolution by comparing genomes from different time periods. There have been two such studies recently.

In the first one, present-day genomes were compared with genomes from Late Antiquity to the Middle Ages (n=467). The comparison showed a substantial rise in average cognitive ability over time—between one third and one half of a standard deviation (Frost, 2024a; Piffer & Kirkegaard, 2024).

The actual rise may have been even larger, since it is imperfectly measured by a comparison between medieval and present-day genomes. At one end of the timeline, according to Gregory Clark, cognitive ability had already begun to rise during the Middle Ages. At the other, it may have peaked in the Victorian Era and then declined thereafter (Frost, 2022d). Also, the rise in cognitive ability may have begun earlier in some parts of Western Europe than in others.

To address these limitations, a second study was conducted with genomes from one region of England (Cambridge and surrounding area, n=269). The genomes are from the 11th to 19th centuries (Piffer & Connor, 2025).

Genomes by century and mean EA polygenic score by century (Piffer & Connor, 2025. Table 1)

For some centuries, notably the 16th to the 18th, we have too few genomes for a century-by-century analysis. But the overall trend is clear: a rapid increase in mean cognitive ability from the 1300s onward.

Particularly impressive is the increase in the "smart fraction": the top 1% in 1850 was as smart as the top 0.1% in the year 1000.

Estimated percentage of the population in the top 5% and top 1% as defined by 1850 norms and using the linear regression for the whole period (Piffer & Connor, 2025, Table 3). EA scores have been converted into IQ scores for ease of interpretation.

Thus, on a per capita basis, the last millennium saw a massive increase in the number and proportion of highly intelligent people. No longer voices crying in the wilderness, there were increasingly enough of them to meet in coffeehouses, learned societies, and debating clubs. This synergy would give rise to the Enlightenment and the Industrial Revolution, ultimately energizing all areas of life—not only the sciences but also literature, music, and the arts (de Courson et al., 2023).

The authors of the above studies—Davide Piffer, Gregory Connor, and Emil Kirkegaard—intend to pursue this research. In particular, they hope to answer the following questions:

• Did cognitive evolution proceed at the same pace from the 11th to 19th centuries? The current data suggest that it was sluggish at first and then took off sometime in the 1300s.

• Did the take-off occur in one part of Western Europe and then spread elsewhere? One candidate region would be England and Holland, which began their economic take-off in the 1300s (Frost, 2022a). Another would be northern Italy during the Renaissance (Rindermann, 2018, p. 133, 141-142, 259-260).

• When did cognitive evolution come to a halt? Did it peak in the late 19th century and decline thereafter with the rise of industrial capitalism and the severance of the link between economic and reproductive success? (Frost, 2022d).

• Was there a parallel evolution of other mental and behavioral traits? For example: time preference, propensity for violence, impulse control, guilt proneness, empathy, etc. Did this parallel evolution create a positive feedback loop? i.e., high frequency of “middle-class” traits → growth of the market economy → growth of the middle class → higher frequency of “middle-class” traits.

• Did cognitive evolution include alleles that are cognitively beneficial as heterozygotes but deleterious as homozygotes? One example might be alleles for autism, which became more frequent in England with rising cognitive ability (Piffer & Kirkegaard, 2024). This sort of heterozygote advantage has already been noted in Ashkenazi Jews with respect to lysosomal storage diseases: Tay-Sachs (2 alleles); Gaucher's (5 alleles); Niemann-Pick; and Mucolipidosis Type IV. All of these alleles became frequent in the same period of time, the same population, and the same metabolic pathway—an indication of natural selection, and strong selection at that (Cochran et al., 2006; Diamond, 1994; Frost, 2022b). Such alleles attest to the rapidity and recentness of the rise in cognitive ability. There are still adverse side-effects because selection has not had enough time to work the bugs out. Evidently, these alleles account for only a fraction of the total rise in cognitive ability among Ashkenazi Jews.

To answer these questions with sufficient detail and clarity, we will need more genomic data from medieval/post-medieval times. Such data could also help us better understand the recent evolution of other genetically influenced traits.

References

Chandelier, J. (2021). L’Occident médiéval. D’Alaric à Léonard. 400-1450. Mondes anciens, Paris: Belin. https://www.belin-editeur.com/loccident-medieval

Clark, G. (2007). A Farewell to Alms. A Brief Economic History of the World. Princeton University Press: Princeton. https://press.princeton.edu/books/paperback/9780691141282/a-farewell-to-alms

Clark, G. (2009). The domestication of man: the social implications of Darwin. ArtefaCToS, 2, 64-80. https://www.researchgate.net/publication/277275046_The_Domestication_of_Man_The_Social_Implications_of_Darwin

Clark, G. (2023). The inheritance of social status: England, 1600 to 2022. Proceedings of the National Academy of Sciences, U.S.A., 120(27), e2300926120 https://doi.org/10.1073/pnas.2300926120

Cochran, G., Hardy, J., & Harpending, H. (2006). Natural history of Ashkenazi intelligence. Journal of Biosocial Science, 38(5), 659-693. https://doi.org/10.1017/S0021932005027069

de Courson, B., Thouzeau, V., & Baumard, N. (2023). Quantifying the scientific revolution. Evolutionary Human Sciences, 5, E19. https://doi.org/10.1017/ehs.2023.6

Diamond, J.M. (1994). Jewish Lysosomes. Nature, 368, 291-292. https://doi.org/10.1038/368291a0

Frost, P. (2010). The Roman State and genetic pacification. Evolutionary Psychology, 8(3), 376-389. https://doi.org/10.1177%2F147470491000800306

Frost, P. (2020). The large society problem in Northwest Europe and East Asia. Advances in Anthropology, 10(3), 214-134. https://doi.org/10.4236/aa.2020.103012

Frost, P. (2022a). When did Europe pull ahead? And why? Peter Frost’s Newsletter. November 21.

Frost, P. (2022b). Ashkenazi Jews and recent cognitive evolution. Peter Frost’s Newsletter, December 5.

Frost, P. (2022c). Europeans and recent cognitive evolution. Peter Frost’s Newsletter, December 12. https://www.anthro1.net/p/europeans-and-recent-cognitive-evolution

Frost, P. (2022d). The Great Decline. Peter Frost’s Newsletter, December 20.

Frost, P. (2023). 1. Genetic pacification in Western Europe from late medieval to early modern times. Peter Frost’s Newsletter, December 19.

Frost, P. (2024a). Cognitive evolution in Europe: Two new studies. Peter Frost’s Newsletter, March 14. https://www.anthro1.net/p/cognitive-evolution-in-europe-two

Frost, P. (2024b). How Christianity rebooted cognitive evolution. Aporia Magazine, October 10.

Frost, P. (2025). Adapting to an environment of their making. Peter Frost’s Newsletter, February 25. https://www.anthro1.net/p/adapting-to-an-environment-of-their

Frost P., & Harpending, H. (2015). Western Europe, state formation, and genetic pacification. Evolutionary Psychology, 13(1), 230-243. https://doi.org/10.1177%2F147470491501300114

Jongman, W. M., Jacobs, J.P. & Goldewijk, G.M.K. (2019). Health and wealth in the Roman Empire. Economics & Human Biology, 34, 138-150. https://doi.org/10.1016/j.ehb.2019.01.005

Melleno, D. (2014). North Sea networks: trade and communication from the seventh to the tenth century. Comitatus: A Journal of Medieval and Renaissance Studies, 45, 65-89. https://doi.org/10.1353/cjm.2014.0055

Oesterdiekhoff, G.W. (2012). Was pre-modern man a child? The quintessence of the psychometric and developmental approaches. Intelligence, 40, 470–478. https://doi.org/10.1016/j.intell.2012.05.005

Piffer, D., & Connor, G. (2025). Genomic Evidence for Clark's Theory of the British Industrial Revolution, preprint, ResearchGate, June. https://www.researchgate.net/publication/392808200_Genomic_Evidence_for_Clark's_Theory_of_the_British_Industrial_Revolution

Piffer D, Dutton, E., & Kirkegaard, E.O.W. (2023). Intelligence Trends in Ancient Rome: The Rise and Fall of Roman Polygenic Scores. OpenPsych. Published online July 21, 2023. https://doi.org/10.26775/OP.2023.07.21

Piffer, D., & Kirkegaard, E.O.W. (2024). Evolutionary Trends of Polygenic Scores in European Populations from the Paleolithic to Modern Times. Twin Research and Human Genetics, 27(1), 30-49. https://doi.org/10.1017/thg.2024.8

Rindermann, H. (2018). Cognitive Capitalism. Human Capital and the Wellbeing of Nations, 1st ed.; Cambridge University Press. https://www.cambridge.org/core/books/cognitive-capitalism/7C10B724756D97F00B7AF0515B800CC5

Terpstra, T. (2020). Roman technological progress in comparative context: The Roman Empire, Medieval Europe and Imperial China. Explorations in Economic History, 75, 101300. https://doi.org/10.1016/j.eeh.2019.101300

Total fertility rate by country in 2024 (Wikicommons - Korakys)

The fertility rate began to decline in the Western world before spreading elsewhere. Perhaps surprisingly, this decline is going much farther in many non-Western populations.

Today, fertility rates are lowest in East Asia, where women typically have one child or none at all. You have likely heard that South Korea is down to a fertility rate of 0.82 per woman. But you may not know that this country is part of a much larger zone of ultra-low fertility that includes the entire north and northeast of China. The Chinese province of Heilongjiang now has a fertility rate of 0.57 (Chen, 2025).

Most of the world is not far behind. Jamaica is down to 1.3 per woman, and similar low fertility is developing across Latin America and the Caribbean (Hill, 2025; McLeod, 2025). If we look at Europe, we see that the northwest is mostly above the European average of 1.4, whereas the south and the east largely have ultra-low fertility.

The same pattern appears among population groups in the United States. In 2024, Euro Americans surpassed African Americans in terms of fertility, after having already surpassed Asian Americans and Amerindian Americans (Frost, 2024; Hamilton et al., 2025). They are now set to overtake Hispanic Americans — not because their fertility rate is rising but because it is falling more slowly.

Northwest Europeans seem to have created an anti-natalist culture that is more toxic to other groups than to themselves. This culture acts like an acid bath that dissolves traditional non-Western supports for family formation and child raising. The same fertility reduction is happening to non-Western populations everywhere as they assimilate Western culture, be it in their diasporas or in their home countries. South Korea is an extreme example, but the same holds true for East Asia in general and, increasingly, Latin America.

Better adapted to individualism?

This greater tolerance for anti-natalism by northwest European populations seems to be part of a greater tolerance for the adverse effects of individualism.

In previous posts, I reviewed the evidence that WEIRD people of northwest European descent have long been mental and behavioral outliers. For over a millennium, they have shown a consistent pattern of nuclear families, relatively weak kinship ties, and solitary living during at least part of adulthood. As a result, they have a greater tolerance for individualism, personal autonomy, and social atomization. They are also more adept at impersonal prosociality — social interactions that are less personal, less emotionally intense, and more inclusive of non-friends and non-kin. Finally, they tend to form “moral communities” based not on shared kinship but on shared ideology and rule-following (Frost, 2020; Frost, 2025a; Frost, 2025b; hbd*chick, 2014; ICA, 2020; MacDonald, 2019; Schulz et al., 2019).

These characteristics have enabled northwest Europeans to transcend the limitations of kinship and organize themselves on a much larger scale, socially and economically. They were thus the first to develop the modern state and a true market economy. Over time, these larger entities have grown at the expense of smaller ones based on kinship, including the family unit, with the result that kinship is today reduced to a vestigial role.

Although northwest Europeans can better tolerate the adverse effects of individualism, they too have seen a substantial decline in their fertility rate, which reflects this shift toward greater individualism and weaker kinship ties, particularly those within the family unit. They are thus close to the psychological limits of social atomization.

Better able to resist the fertility decline?

How are northwest Europeans maintaining a higher fertility rate? One way is by belonging to “moral communities” that support procreation through mutual assistance and shared commitment to a moral system. In the United States, such communities include Mormons, Amish, Hasidic Jews and, to some extent, evangelicals and traditional Catholics. It is above all these fertile subcultures that are buoying up the Euro American fertility rate.

In addition, because northwest Europeans don’t expect to receive child support from aunts, uncles, and other close kin, any difficulties in getting such support will not affect their decision to have children. The situation is different with other populations, particularly those of Sub-Saharan African descent. They are undergoing a greater decline in fertility because Westernization — and the social atomization it brings — increases the uncertainty of child support from close kin. There remains support from one’s partner, but it too is far from certain.

This point came up in interviews with Jamaican women about the fertility decline:

Kids? They drain your wallet with diapers, formula, and school fees. I’m not giving up my weekend travel or fancy dinners just yet. Plus you can never, ever, ever trust a man to come through for you. And it’s not au fait for moms to be the ones to leave, and allow men to be the single parents. So knowing all I know, I won’t allow myself to be trapped. (Hill, 2025)

References

Chen, S. (2025). China’s northeast provinces face demographic crisis with fertility rates plummeting below 1.0. Demstate, August 15. https://demstate.com/article/chinas-northeast-provinces-face-demographic-crisis-with-fertility-rates-plummeting-below-10

Frost, P. (2020). The large society problem in Northwest Europe and East Asia. Advances in Anthropology, 10(3), 214-134. https://doi.org/10.4236/aa.2020.103012

Frost, P. (2024). Trump: White America’s savior? Peter Frost’s Newsletter, October 7. https://www.anthro1.net/p/trump-white-americas-savior

Frost, P. (2025a). When did northwest Europeans become WEIRD? Peter Frost’s Newsletter, August 5. https://www.anthro1.net/p/when-did-northwest-europeans-become

Frost, P. (2025b). Reply to Joseph Henrich. Peter Frost’s Newsletter, August 14. https://www.anthro1.net/p/reply-to-joseph-henrich

Hamilton, B.E., Martin, J.A., & Osterman, M.J.K. (2025). Births: Provisional data for 2024. Vital Statistics Rapid Release. April;(38):1–10. https://dx.doi.org/10.15620/cdc/174587

hbd*chick (2014). Big summary post on the Hajnal Line. October 3. https://hbdchick.wordpress.com/2014/03/10/big-summary-post-on-the-hajnal-line/

Hill, S. (2025). The birth drought: Women tell why fertility rates will continue to plummet, Jamaica Observer, August 11. https://www.jamaicaobserver.com/allwoman/2025/08/11/the-birth-drought/

ICA (2020). Research Themes—Marriage Patterns. Institutions for Collective Action. https://web.archive.org/web/20190329070516/http://www.collective-action.info/_THE_MarriagePatterns_EMP

MacDonald, K. (2019). Individualism and the Western Liberal Tradition: Evolutionary Origins, History, and Prospects for the Future. Amazon. https://www.amazon.ca/Individualism-Western-Liberal-Tradition-Evolutionary/dp/1089691483

McLeod, S. (2025). Jamaica’s fertility rate drops to 1.3 — among the lowest in the world. Caribbean National Weekly, August 4. https://www.caribbeannationalweekly.com/news/jamaicas-fertility-rate-drops-to-1-3-among-the-lowest-in-the-world/

Schulz, J.F., Bahrami-Rad, D., Beauchamp, J.P., & Henrich, J. (2019). The Church, intensive kinship, and global psychological variation. Science, 366(707), 1-12. https://doi.org/10.1126/science.aau5141

An early fishing village in Europe (Ridpath’s History of the World - Wikicommons)

How did northwest Europeans become more individualistic, less clannish, and more adept at impersonal prosociality? Was it because of the Western Church’s restrictions on cousin marriage and polygyny? Or did Christianity adopt these norms as it spread into Europe?

What does aDNA tell us about northwest Europeans before they became Christian? Was their individualism due to Western Christianity and its restrictions on cousin marriage and polygyny? Or did these norms already exist as Christianity spread north and west into Europe?

Also, how do the elements of this behavioral package relate to each other? Did a certain marriage pattern lead to individualism? Or did both have a common cause?

These are the questions I tried to address in my last post, which was first published in Aporia Magazine. My reposting led to another round of criticism from Joseph Henrich on X. His criticisms and my responses are given below in the hope that debate will bring both of us closer to the truth.

First, the work referenced by my colleagues and I is about kinship intensity, only one element of which involves cousin marriage. It's quite possible to have low rate of cousin marriage but high kinship intensity. This is a fundamental misreading.

In your landmark paper in Nature, you argued that the Christian bans on polygyny and cousin marriage reduced kinship intensity and thereby promoted individualism. Kinship intensity was thus lower in Western Europe because the Western Church imposed a more radical ban on cousin marriage:

Crucially, while the Eastern Church did adopt some of the same policies as the Western Church, it never endorsed the Western Church’s extensive prohibitions on cousin marriage, adopted many policies only later, and was unenthusiastic about enforcement. Thus, we expect similar but substantially weaker effects for the Eastern Church. Other sects of Christianity adopted even less of the MFP [marriage and family policy]. Nestorian and Coptic Christians, for example, continued marrying their cousins for at least another millennium. (Schulz et al., 2019, p. 10)

You define the “kinship intensity index (KII)” as follows:

The KII is an omnibus measure for kinship intensity that captures the presence of cousin-marriage preferences, polygamy, co-residence of extended families, clan organization, and community endogamy. (Schulz et al., 2019, p. 3)

Let’s assume that these elements are not listed in order of importance. Let’s also assume that cousin marriage has a minor influence on the KII. What do we know about the other elements when the Western Church was still working out its marriage and family policy? Was the KII higher then than it is today?

In Carolingian Gaul at a time of increasing population on the estates of the Abbey of St Germain-des-Prés near Paris about 801-20 some 4,316 married or once-married adults had 4060 children. Households were nuclear and small amongst the married folk — rather under four persons on average — and, in addition, there were about 843 single adults, mostly men. About 16.3 per cent of the adult population were unmarried, so that in many respects there was already a Western European marriage pattern.

… During the ninth century the situation in Villeneuve-Saint-Georges was similar. The now or ever-married adults constituted 47.9 per cent of the entire population and the children 46.6 per cent. Households mainly consisted of parents and children. Seven children lived with their father only, and 15 with their mother. Only 22 parents lived with their son and twelve unmarried sons with their parents. Four widowers, four widows, nine bachelors and five spinsters seem to have lived alone. Of the adult population 11.5 per cent had never married. The picture is very like that of Western Europe in modern times. (Hallam, 1985, p. 56)

So even in the early ninth century, when the Western Church had not yet adopted its most radical policies on marriage and family, we already see a familiar picture of nuclear families, as well as high rates of adults living alone.

Second, proper quantitative reanalyses of historical data, such as epigraphies, suggest lots of cousin marriages in the Roman Empire. Our team of classists is using the same approaches as past historians except we rely on more extensive datasets.

According to aDNA from a wide range of regions and time periods, cousin marriage was rare during the Roman period and during Antiquity in general:

We found that only 1 out of 1785 ancient individuals have long ROH typical for the offspring of first-degree relatives (e.g., brother–sister or parent–offspring). Historically, matings of first-degree relatives are only documented in royal families of ancient Egypt, Inca, and pre-contact Hawaii, where they were sporadic occurrences7. The only other example of an offspring of first-degree relatives found using aDNA to date is the recently reported case from an elite grave in Neolithic Ireland. Our findings are in agreement that first-degree unions were generally rare in the human past (Ringbauer et al., 2021).

Cousin marriage seems to have become common only in later times, possibly during the Islamic period.

… In two specific regions with high levels of long ROH in the present-day2, the dataset contained a sufficient number of ancient individuals to allow analyzing time transects. For both transects (the Levant and present-day Northwest Pakistan), we observe a substantial shift in the levels of long ROH. In contrast to the high abundance of long ROH typical of close kin unions in the present-day individuals, long ROH was uncommon in the ancient individuals, including up to the Middle Ages. (Ringbauer et al., 2021)

Even if we assume that cousin marriage became common as early as Roman times, we cannot assume that the same change in marriage patterns affected the non-Romanized and semi-Romanized populations of northwest Europe.

Third, the ancient DNA paper referenced was done by a post-doc in the lab next door to my office. Harold's method only distinguishes 1st cousin marriages with any confidence, not more distant pairings. Most cousins are more distant; Church also banned spirit kin and affines.

I referenced four different papers, which used different methods.

Gnecchi-Ruscone et al. (2024) used two methods:

• KIN (Kinship INference), which can confidently identify first- and second-degree relationships;

• ancIBD, which can detect close genetic relationships up to the sixth degree.

They found evidence of only one possible marriage between first cousins out of around 300 individuals, and no evidence of cousin marriage at higher degrees. Admittedly, this study used aDNA from Avars and may thus be irrelevant to the issue of marriage patterns among indigenous northwest Europeans.

Wang et al. (2025) likewise used KIN and ancIBD. They examined aDNA from two sites in Austria: a non-Avar one (Mödling, or MGS) and an Avar one (Leobersdorf, or LEO). They concluded:

Given that none of the newly reported individuals carry high amounts of runs-of-homozygosity genomic regions—the indication of inbreeding—as estimated by hapROH … we infer that consanguinity was strictly avoided in both MGS and LEO across six generations. That was mainly achieved by exogamy: 17 of the 19 (90%) mothers buried in Leobersdorf with identifiable offspring have no ancestors buried on site; in the much larger community of Mödling, they are 46 out of 59 (78%). Many daughters seem to have left to be married elsewhere; between ages 7 and 17 years, the sex ratios of the deceased male to female individuals at LEO and MGS are about 1.5:1 and 1.7:1 respectively, and among adults, hardly any female individuals born by parents on site remain.

Blöcher et al. (2025) used the ancIBD method. They concluded:

The near absence of long (>12 cM) runs of homozygosity (ROH) and the lack of shared IBD segments (>8 cM) between spouses support strict incest avoidance, excluding relationships closer than the sixth degree.

Cassidy et al. (2025) used a method called ped-sim, which is accurate up to the fourth degree of relatedness (Caballero et al., 2019). They concluded:

Y chromosome diversity is high …, and patterns of ROH imply that these were relatively large outbreeding communities.

Furthermore, many cousins were 'half cousins' because of polygyny--our fathers were half brothers (different moms). Or, our grandfathers were half brothers. The aDNA method has little resolution for this.

Polygyny was common only among the Avars, a Turkic people who lived in Austria and Hungary at that time. It was rare within pre-Christian indigenous populations elsewhere in northwest Europe:

England – no mention of polygyny (Cassidy et al., 2025).

Southern Germany –

In Altheim and Büttelborn, five individuals (four men and one woman) had multiple partners, though it remains unclear whether this reflects polygamy or serial monogamy. Nonetheless, the predominance of 61 single-partner unions suggests that monogamy was the norm in Early Medieval southern Germany (Blocher et al., 2025).

Austria (Avars and indigenous Europeans) – polygyny is attested at the Avar site. At the non-Avar site, multi-reproductive unions seem to be levirate marriages.

At Mödling, female individuals also had children with two or more partners almost as often as male individuals (15 out of 31). As most partners of these multi-reproductive female individuals were related to each other (brothers, half-brothers, stepsons), these were probably levirate unions, an arrangement under which a widow marries a male relative of her deceased partner. (Wang et al., 2025).

Hungary (Avars) –

In RK only, we discovered 15 cases involving a male partner and 7 cases involving a female one. Male individuals had two partners in ten cases, three partners in four cases, and four partners in one case (RKF042); around 85% of these individuals are older men (aged 35–59). The young ages of female partners at death may indicate serial monogamy (RKC011), but the presence of older female partners in multiple partnerships suggest polygyny (RKF042 and RKF180). Multiple reproductive partners were also discovered in HNJ and KFJ (one and four cases, respectively). That means that polygyny might not have been restricted to the highest stratum of society that is known from the historical sources, but also occurred in the general population (Gnecchi-Ruscone et al., 2024).

Fourth, we have been compiling all the studies of the ancient DNA from European burials. Taken together, they provide strong support for intensive kinship, including clans, extended families, patrilineal inheritance and polygyny. So, Cherry picking...

The term “cherry picking” implies that I ignored some of the relevant aDNA studies. Actually, I did miss two when I wrote my last post, one concerning Europe in general and the other concerning Ireland. Both seem in line with the other studies, particularly with respect to low rates of polygyny and cousin marriage since the Neolithic transition:

In fact, several Neolithic burials show evidence of nuclear families, which may reflect a monogamous marriage system. A shift from polygyny to monogamy would have the effect of decreasing male variance in reproductive success, since more males would now be able to mate, and consequently would increase Nm. This could result in a signal of population growth in NRY data that would be more recent compared to that observed in mtDNA and is exactly what Dupanloup and colleagues have argued and found. Our results are in good agreement with theirs. (Rasteiro & Chikhi, 2013)

Importantly, where multiple genomes have been sequenced from Neolithic contexts in Ireland, studies have shown that most of the individuals buried together were not closely biologically. This contrasts strongly with the findings from the well-preserved burial deposits of Frälsegården or Hazleton, but matches the general picture emerging from Britain, including Orcadian passage tombs (with their admittedly small sample sets), of people buried together not being closely related, especially in the Later Neolithic. Where such inter- or intra-site relations have been identified from Ireland, they are frequently distant (e.g. fifth degree or further: e.g. second cousins or a great-great-great grandparent), rather than close genetic relationships (e.g. first to fourth degree: parents, children, siblings, grandparents/grandchildren, uncles or aunts or nieces and nephews, or first cousins). (Carlin et al., 2025)

Fifth, the study referenced in PLoS One, which supposedly undermines our approach, actually supports it. The analysis is riddled with problems. We'll be providing a proper reanalysis.

We are preparing a full-throated reply, which we'll publish in an actual journal instead of blogging. This thread is so replete with errors, misstatements and misreadings, it's difficult to know where to begin. I will hold most of my comments for that forum. Nevertheless...

I await your paper. I hope it will cover all of the aDNA studies of pre-Christian northwest Europeans.

Discussion

We all can become invested in our ideas, myself included. It’s important, however, to revise our ideas when they’re proven wrong. Perhaps that point has not been reached in this debate. Perhaps avoidance of cousin marriage is irrelevant. It may just be a general North Eurasian adaptation, as indicated by the Avar aDNA. On the other hand, a low polygyny rate seems to be much more specific to northwest Europeans, and it likewise seems to precede Christianization.

Hopefully, aDNA will be examined for other heritable WEIRD traits, such as impersonal prosociality, affective empathy, and guilt proneness. But such research will be conducted only if it seems justified. Hence the importance of debate on the origins of WEIRDness.

If WEIRDness didn’t begin with Christianity, when and how did it? HBD*chick has argued that it began with the spread of manorialism in the early Middle Ages (HBD*chick, 2012). This explanation has its share of problems:

• Manorialism didn’t take off in northwest Europe until the seventh to eighth centuries, so we still have to explain the low rates of polygyny and cousin marriage in earlier times.

• Even if we assume that WEIRDness is an outcome of purely cultural evolution — with no genetic changes — we still have to explain why the inhabitants of French manors were already fully WEIRD at the beginning of the ninth century. Is such a fundamental cultural shift possible over such a short time interval?

• Finally, feudal India had a similar system of land tenure. Why didn’t WEIRDness evolve there?

I’m inclined to see WEIRDness as being much older. To me, it looks like an adaptation to the fluid social environment that prevailed around the North Sea and the Baltic in the Late Mesolithic. These coastal peoples spent the winter inland, as small hunting bands of closely related individuals. In the summer, they congregated in large coastal settlements to subsist on fishing and sealing. They thus had to live at close quarters with non-kin, and not necessarily the same ones in each successive year. Such an environment may have selected for WEIRDness, especially weaker kinship ties and impersonal prosociality (Frost, 2020).

I may be wrong. And I always have to live with that possibility.

References

Blöcher, J., Vallini, L., Velte, M., Eckel, R., Guyon, L., Winkelbach, L., … Burger, J. (2025). Historic Genomes Uncover Demographic Shifts and Kinship Structures in Post-Roman Central Europe. bioRxiv, 2025.03.01.640862; https://doi.org/10.1101/2025.03.01.640862

Caballero, M., Seidman, D.N., Qiao, Y., Sannerud, J., Dyer, T.D., Lehman, D.M., et al. (2019) Crossover interference and sex-specific genetic maps shape identical by descent sharing in close relatives. PLoS Genet, 15(12): e1007979. https://doi.org/10.1371/journal.pgen.1007979

Carlin, N., Smyth, J., Frieman, C. J., Hofmann, D., Bickle, P., Cleary, K., ... & Pope, R. (2025). Social and genetic relations in Neolithic Ireland: re-evaluating kinship. Cambridge Archaeological Journal, 1-21. https://doi.org/10.1017/S0959774325000058

Cassidy, L. M., Russell, M., Smith, M., Delbarre, G., Cheetham, P., Manley, H., ... & Bradley, D. G. (2025). Continental influx and pervasive matrilocality in Iron Age Britain. Nature, 637(8048), 1136. https://doi.org/10.1038/s41586-024-08409-6

Frost, P. (2020). The large society problem in Northwest Europe and East Asia. Advances in Anthropology, 10(3), 214-234. https://doi.org/10.4236/aa.2020.103012

Gnecchi-Ruscone, G. A., Rácz, Z., Samu, L., Szeniczey, T., Faragó, N., Knipper, C., ... & Hofmanová, Z. (2024). Network of large pedigrees reveals social practices of Avar communities. Nature, 629(8011), 376-383. https://doi.org/10.1038/s41586-024-07312-4

Hallam, H.E. (1985). Age at First Marriage and Age at Death in the Lincolnshire Fenland, 1252-1478. Population Studies, 39, 55-69. https://doi.org/10.1080/0032472031000141276

HBD*chick. (2012). Medieval manorialism and the Hajnal line. HBD*chick, June 2. https://hbdchick.wordpress.com/2012/02/06/medieval-manoralism-and-the-hajnal-line/

Rasteiro, R., & Chikhi, L. (2013). Female and Male Perspectives on the Neolithic Transition in Europe: Clues from Ancient and Modern Genetic Data. PLoS ONE, 8(4): e60944. https://doi.org/10.1371/journal.pone.0060944

Ringbauer, H., Novembre, J., & Steinrücken, M. (2021). Parental relatedness through time revealed by runs of homozygosity in ancient DNA. Nature Communications, 12, 5425. https://doi.org/10.1038/s41467-021-25289-w

Schulz, J.F., Bahrami-Rad, D., Beauchamp, J.P., & Henrich, J. (2019). The Church, intensive kinship, and global psychological variation. Science, 366(707), 1-12. https://doi.org/10.1126/science.aau5141

Wang, K., Tobias, B., Pany-Kucera, D., Berner, M., Eggers, S., Gnecchi-Ruscone, G. A., ... & Hofmanová, Z. (2025). Ancient DNA reveals reproductive barrier despite shared Avar-period culture. Nature, 1-8. https://doi.org/10.1038/s41586-024-08418-5

The countries in green are more individualistic than the world average. The ones in red are more collectivistic (Beugelsdijk & Welzel, 2018). This difference between northwest Europeans and other humans goes back at least a millennium. Was it due to Western Christianity or does the cause go farther back?

WEIRD is an acronym for Western, Educated, Industrialized, Rich, and Democratic. It refers to people of northwest European descent and was coined to warn against using them in studies to understand human nature in general. They are not necessarily the same as people elsewhere.

Two decades ago, Joseph Henrich and other social scientists concluded that WEIRD people are outliers in many aspects of mind and behavior, including sensory ability, economic preferences, personality structure, morality, and cognition. They are divergent because they have adapted to an atypical environment of weak kinship, strong individualism, and “impersonal pro-sociality”—in other words, social interactions that are less personal, less emotionally intense, and extending much further than friends and family (Henrich, 2024; Schulz et al., 2019).

For at least a thousand years, this behavioral environment has prevailed north and west of a line running from Trieste to St. Petersburg (known as the Hajnal Line). It is characterized by certain longstanding patterns of behavior:

• Solitary living for at least part of adulthood, with many individuals remaining single their entire lives.

• Departure from the home upon reaching adulthood, either to form a new household or to circulate among unrelated households, typically as servants.

• Less loyalty to kin and greater willingness to trust strangers (Schulz et al., 2019; see also Frost, 2017; Frost, 2020; Hajnal, 1965; Hartman, 2004; hbd*chick, 2014; ICA, 2020; MacDonald, 2019; Seccombe, 1992, pp. 94-95, 150-153, 184-190).

Some authors, like Joseph Henrich, see WEIRDness as a legacy of Western Christianity, which arose in Western Europe after the fall of the Roman Empire and the rise of barbarian kingdoms. According to this theory, northwest Europeans became less clannish and more individualistic because the Western Church increasingly restricted cousin marriage:

1. Roman times: only first-cousin marriages were banned.

2. Seventh century: the ban was extended two degrees further when the Western Church adopted the anti-incest prohibitions of the Visigothic Code.

3. Early ninth century: the Western Church began to calculate degrees of kinship through the so-called “Germanic system,” thus doubling the number of forbidden marriage partners (Chandelier, 2021, p. 224; see Note).

The last restriction forced almost everyone to marry outside their local kin group, thus causing clans to disappear and making people more individualistic and less concerned about kinship ties (hbd*chick, 2014; McCann, 2010, pp. 57-58; Schulz et al., 2019).

Other authors, however, have argued that the cousin marriage ban was simply a Christianization of existing norms, specifically Germanic ones (Frost, 2020; Kirkegaard, 2025; MacDonald, 2019; Policy Tensor, 2021). As the Christian faith spread north and west, it absorbed local customs, including those relating to marriage:

During the period preceding the Teutonic invasion, speaking broadly, the church adhered to the Roman law and custom; thereafter those of the Germans, even when the marriage consisted in the formal sale and tradition [i.e., transfer] of the bride, were accepted. (Howard, 1904, p. 291).

Germanic provenance is evident in the bans themselves. The seventh-century ban was taken from the Visigothic Code, and the new kinship calculation method, adopted in the ninth century, was referred to as “Germanic” (Chandelier, 2021, p. 224; Frost, 2020; McCann, 2010, pp. 57-58; see Note).

Moreover, as shown by data from early medieval estates, northwest Europeans were already WEIRD in the ninth century, when the most extreme ban on cousin marriage came into effect. French households were small and nuclear, with 12% to 16% of adults not yet married and adults usually marrying in their mid to late twenties (Hallam, 1985, p. 56). High rates of delayed marriage seem to have long been common among northwest Europeans, as suggested by the writings of Julius Caesar and Tacitus on the Germanic tribes of Antiquity:

Those who have remained chaste for the longest time, receive the greatest commendation among their people: they think that by this the growth is promoted, by this the physical powers are increased and the sinews are strengthened. And to have had knowledge of a woman before the twentieth year they reckon among the most disgraceful acts.

Caesar, De Bello Gallico 6: 21

Late comes love to the young men, and their first manhood is not enfeebled; nor for the girls is there any hot-house forcing; they pass their youth in the same way as the boys.

Tacitus, Germania 20

Evidence from aDNA

For earlier evidence of cousin marriage avoidance, we can look at DNA from remains in ancient cemeteries, specifically the degree of genetic similarity between spouses as measured by runs of homozygosity (ROH).

To date, there have been four such studies from pre-Christian Western Europe. All four show that cousin marriages were already rare.

A study of 57 genomes from the first century BC to the first century AD in southern England

“Y chromosome diversity is high …, and patterns of ROH imply that these were relatively large outbreeding communities.” (Cassidy et al., 2025)

A study of 248 genomes from the third to eighth centuries in southern Germany

“The near absence of long (>12 cM) runs of homozygosity (ROH) and the lack of shared IBD segments (>8 cM) between spouses support strict incest avoidance, excluding relationships closer than the sixth degree.” (Blöcher et al., 2025)

A study of 722 genomes from the seventh to eighth centuries in Austria

“Given that none of the newly reported individuals carry high amounts of runs-of-homozygosity genomic regions—the indication of inbreeding—as estimated by hapROH … , we infer that consanguinity was strictly avoided in both MGS [Mödling site] and LEO [Leobersdorf site] across six generations …. That was mainly achieved by exogamy: 17 of the 19 (90%) mothers buried in Leobersdorf with identifiable offspring have no ancestors buried on site; in the much larger community of Mödling, they are 46 out of 59 (78%). Many daughters seem to have left to be married elsewhere; between ages 7 and 17 years, the sex ratios of the deceased male to female individuals at LEO and MGS are about 1.5:1 and 1.7:1 respectively, and among adults, hardly any female individuals born by parents on site remain.” (Wang et al., 2025)

A study of 424 genomes from the second to mid-ninth centuries in Hungary

“[W]e find no cases of biological consanguinity, based on the absence of long runs of homozygosity (ROH) segments in all analysed individuals …. We do not even detect ROH patterns consistent with more-distant consanguineous unions, such as at the level of second-degree cousins, despite a high occurrence of levirate and multipartner unions. Among Eurasian steppe peoples, intermarriage within the paternal line was permitted only after a certain number of generations, which could range between five and nine. Such rules would explain the absence of even distant biological consanguinity. It is intriguing that the only case we detected of reproductive partners being related was to the sixth degree (which would still be consistent with such rules) and involves the only non-exogamous female individual in RK. This further suggests the uniqueness of this single case.” (Gnecchi-Ruscone et al., 2024)

Discussion

Cousin marriage

The aDNA evidence points to a single conclusion: in pre-Christian times, cousin marriage was already rare in northwest Europe.

But it was also rare farther east. In the Austrian and Hungarian studies, the DNA came not only from indigenous Europeans (probably Slavic peoples) but also from Avars—a Turkic people who entered the Carpathian Basin in the sixth century. Avoidance of cousin marriage may therefore be part of a broader, North Eurasian adaptation.

This point is made, in part, by one of the above studies: “Among Eurasian steppe peoples, intermarriage within the paternal line was permitted only after a certain number of generations, which could range between five and nine. Such rules would explain the absence of even distant biological consanguinity” (Gnecchi-Ruscone et al., 2024).

In fact, cousin marriage may have been rare across all of Eurasia before becoming frequent in those regions where it is now common, i.e., the Middle East, North Africa, and Central/South Asia. This is the conclusion of two other studies, one of 411 genomes over the past 15,000 years and another of 1,785 genomes over the past 45,000 years (Ceballos et al., 2021; Ringbauer et al., 2021). Neither study, however, is sufficiently powered to tell us exactly when, where, or how cousin marriage became more frequent. Perhaps it began among the elites of early Middle Eastern civilizations and was later emulated by everyone else with the spread of Islam.

In addition, neither study distinguishes between avoidance of first-cousin marriage and avoidance of all cousin marriages up to the sixth degree. The latter may have reached higher levels within a smaller zone of Eurasia.

WEIRD morality

If avoidance of cousin marriage was not specific to northwest Europeans, we should seek the origins of WEIRDness in other mental and behavioral aspects, particularly those that have left traces in aDNA.

One such aspect is the moral system. Northwest Europeans are inclined toward a more universal and absolute conception of morality, which in turn is linked to certain heritable traits.

In this moral system:

• Rules are framed in universal and absolute terms, as opposed to the situational and relativistic framing of kinship.

• Help is willingly given to non-kin, as long as they belong to the same community of rule followers.

• Continual rule breaking leads to expulsion from the community. The line between insiders and outsiders is much more a line between the morally worthy and the morally worthless. Xenophobia is a moral judgment, rather than a simple rejection of the “Other.”

This system has favored—and been favored by—an increased capacity for affective empathy and guilt proneness:

• Affective empathy is extended from the mother-child relationship to all social relationships. Through this involuntary transfer of another person’s feelings to oneself, rule breaking is experienced emotionally as harm not only to others but also to oneself.

• Rule breaking is punished much more by guilt than by shame. A rule breaker feels guilty even when no one else has witnessed the rule breaking. In contrast, shame is felt only when there are witnesses. Everyone thus accumulates a burden of guilt, which can be reduced only through forgiveness, penance, confession, absolution, etc. (Benedict, 1946; Frost, 2017; Frost, 2020).

Like avoidance of cousin marriage, this moral system likely preceded the Christianization of northwest Europeans. In fact, the latter may have transformed the former, creating a branch of the Christian faith that is not only more fearful of cousin marriage but also more preoccupied with guilt, particularly the inevitability of guilt and the continual need to reduce one’s burden of guilt. This may be why the doctrine of original sin is more important in Western Christianity than in Eastern Christianity, and more important in Protestantism than in Catholicism.

To track how guilt proneness and affective empathy have evolved in northwest Europeans, we should look for indications in aDNA. For instance, changes to the mean population level of affective empathy could be tracked by examining the alleles associated with this trait, which is 52-57% heritable (Frost, 2020). These alleles have already been identified and used to calculate polygenic scores (Wendt et al., 2022). The same approach might show us how guilt proneness has varied over space and time.

Note

In the Roman Empire, marriage between close relatives was forbidden up to the third degree, that is, marriage was prohibited between brothers and sisters, parents and children, and between uncles and nieces or aunts and nephews, as well as between ascendants and descendants without limit of degree.

Beginning in the 8th century, the Church began to impose a broader restriction and forbade unions up to the seventh degree, while changing the calculation method to impose the tradition of Germanic origin whereby the degree no longer means the number of relatives between two individuals, but the number of generations before reaching a common ancestor. Thus, two first cousins are kin in the fourth degree under the Roman system, but only in the second degree under the Germanic system: their common ancestor is their grandfather or their grandmother.

The seventh-degree prohibition promoted by the Church therefore corresponds to the fourteenth degree under the old system. This considerably broadened the scope of prohibited unions. It was enough to have in common a great-great-great-great-great-great grandparent for a marriage not to be permitted (my translation) (Chandelier, 2021, p. 224).

Postscript: Discussion on X

After this article was published in Aporia Magazine on June 26, 2025, Joseph Henrich posted a reply on X, which I reproduce below. It seems to be a generic reply to critics of his work, rather than a specific reply to me. In reality, I don’t deny that a relationship of causality exists between WEIRDness and Western Christianity. Nor do I try “to clump Europeans together psychologically.”

I disagree with him only on the direction of the causality. Western Christianity didn’t create WEIRDness. Rather, the latter transformed the former by making the Western branch of Christianity more focused on the individual, more ideological, and more preoccupied with guilt as a means of behavioral control.

Joseph Henrich argues for the reverse: over time, the Western Church made northwest Europeans more and more WEIRD. As proof, he cites the correlation between WEIRDness and the time depth of Western Christianity in various regions. But this correlation is due largely to the expansion of Western European states into regions south and east of the Hajnal Line (notably southern Spain and southern Italy). As these states expanded, they imposed Western Christianity on populations that had previously practiced Islam or Eastern Christianity. So the correlation with time depth is really a correlation with the geopolitical expansion of Western Europe—an expansion that would eventually encompass the entire world.

I would also like to know about the aDNA evidence he mentions with regard to polygyny. The aDNA evidence I’ve seen indicates that Europe had a low polygyny rate even in pre-Christian times.

Joseph Henrich’s comments

Peter misses the argument here, making the entire piece misleading. Once you understand the actual argument, the aDNA data is largely supportive.

The argument is that ONE branch of Christianity (not "Christianity") diffused a set of prescriptions and prohibitions that dissolved the pre-Christian kinship structures into monogamous nuclear families. It's NOT about "cousin marriage" per se.

These norms enforced monogamous marriage (ending common polygyny, concubinage, sex slaves) and tabooed levirate marriage along many types of unions, which included genetic cousins as well as spiritual kin (e.g., goddaughter) and all in-laws. Lots of non-close-genetic relatives.

The norms also altered post-marital residence and inheritance (by testament instead of customary). This is elaborated in great detail in my book.

These changes dramatically reduced what anthropologists call "kinship intensity", which captures the broad and tight social networks formed by kinship norms. (economists renamed these "kinship tightness"). The classic corporate "tight" kin group is the clan.

The aDNA strongly supports this view by confirming the prevalence of polygynous patrilineal clans (without marrying women) in pre-Christian Europe. There's even evidence for levirate marriage, which is the first thing the Church banned in Late Antiquity.

So, even if we take the claims about cousin marriage at face value (which I don't), the aDNA still supports a decline in kinship intensity... which is the key issue. The clans were dismantled.

In his effort to clump Europeans together psychologically, he ignores the application of this theory to explain the variation among Europeans today. E.g., knowing how long a population was under the medieval Church predicts their individualism, conformity & imp. Prosociality

The Church's exposure predicts psychological variation in Europe.

It also works globally

However, if Peter's reading of the aDNA is right, why does Church exposure predict rates of European cousin in the 20th century?

Curiously, after banning cousin marriage, the medieval Church realized that Europeans wanted to marry their kin so much, they monetized by selling dispensations...made big bucks. There was demand... why? Suddenly, people want to marry cousins?

Why does knowing the rates of cousin marriage among Italian provinces tell us so much about psychological variation within Italy?

References

Benedict, R. (1946 [2005]). The Chrysanthemum and the Sword. Patterns of Japanese Culture. First Mariner Books.

Beugelsdijk, S., & Welzel, C. (2018). Dimensions and dynamics of national culture: Synthesizing Hofstede with Inglehart. Journal of Cross-Cultural Psychology, 49(10), 1469-1505. https://doi.org/10.1177/0022022118798505

Blöcher, J., Vallini, L., Velte, M., Eckel, R., Guyon, L., Winkelbach, L., … Burger, J. (2025). Historic Genomes Uncover Demographic Shifts and Kinship Structures in Post-Roman Central Europe. bioRxiv 2025.03.01.640862; https://doi.org/10.1101/2025.03.01.640862

Caesar, J. (1915). De Bello Gallico & Other Commentaries of Caius Julius Caesar (Translated by W.A. Macdevitt). London: J.M. Dent.

Cassidy, L. M., Russell, M., Smith, M., Delbarre, G., Cheetham, P., Manley, H., ... & Bradley, D. G. (2025). Continental influx and pervasive matrilocality in Iron Age Britain. Nature, 637(8048), 1136. https://doi.org/10.1038/s41586-024-08409-6

Ceballos, F. C., Gürün, K., Altinişik, N. E., Gemici, H. C., Karamurat, C., Koptekin, D., ... & Somel, M. (2021). Human inbreeding has decreased in time through the Holocene. Current Biology, 31(17), 3925-3934. https://doi.org/10.1016/j.cub.2021.06.027

Chandelier, J. (2021). L’Occident médiéval. D’Alaric à Léonard. 400-1450. Mondes anciens, Paris: Belin. https://www.belin-editeur.com/loccident-medieval

Frost, P. (2017). The Hajnal line and gene-culture coevolution in northwest Europe. Advances in Anthropology, 7, 154-174. https://doi.org/10.4236/aa.2017.73011

Frost, P. (2020). The large society problem in Northwest Europe and East Asia. Advances in Anthropology, 10(3), 214-134. https://doi.org/10.4236/aa.2020.103012

Frost, P. (2024). How Christianity rebooted cognitive evolution. Peter Frost’s Newsletter, November 13.

Gnecchi-Ruscone, G. A., Rácz, Z., Samu, L., Szeniczey, T., Faragó, N., Knipper, C., ... & Hofmanová, Z. (2024). Network of large pedigrees reveals social practices of Avar communities. Nature, 629(8011), 376-383. https://doi.org/10.1038/s41586-024-07312-4

Hajnal, J. (1965). European marriage pattern in perspective. In: D.V. Glass and D.E.C. Eversley (eds). Population in History. Essays in Historical Demography. London, Arnold.

Hallam, H.E. (1985). Age at First Marriage and Age at Death in the Lincolnshire Fenland, 1252-1478. Population Studies, 39, 55-69. https://doi.org/10.1080/0032472031000141276

Hartman, M.S. (2004). The Household and the Making of History. A Subversive View of the Western Past. Cambridge, Cambridge University Press. https://doi.org/10.1017/CBO9780511818134

hbd*chick (2014). Big summary post on the Hajnal Line. October 3. https://hbdchick.wordpress.com/2014/03/10/big-summary-post-on-the-hajnal-line/

Henrich, J. (2024). WEIRD. Open Encyclopedia of Cognitive Science. https://doi.org/10.21428/e2759450.8e9a83b0

Howard, G.E. (1904). A History of Matrimonial Institutions, Volume 1. Chicago: The University of Chicago Press.

ICA (2020). Research Themes—Marriage Patterns. Institutions for Collective Action. https://web.archive.org/web/20190329070516/http://www.collective-action.info/_THE_MarriagePatterns_EMP

Kirkegaard, E. (2025). Why did NW Europeans become WEIRD? Just Emil Kirkegaard Things, January 29.

MacDonald, K. (2019). Individualism and the Western Liberal Tradition: Evolutionary Origins, History, and Prospects for the Future. Amazon. https://www.amazon.ca/Individualism-Western-Liberal-Tradition-Evolutionary/dp/1089691483

McCann, C.A. (2010). Transgressing the Boundaries of Holiness: Sexual Deviance in the Early Medieval Penitential Handbooks of Ireland, England and France 500-1000. Theses, South Orange, NJ: Seton Hall University. https://scholarship.shu.edu/theses/76

Policy Tensor. (2021). The Church’s crusade against cousin-marriage did not create the Western nuclear family. May 7

Ringbauer, H., Novembre, J., & Steinrücken, M. (2021). Parental relatedness through time revealed by runs of homozygosity in ancient DNA. Nature Communications, 12, 5425. https://doi.org/10.1038/s41467-021-25289-w

Schulz, J.F., Bahrami-Rad, D., Beauchamp, J.P., & Henrich, J. (2019). The Church, intensive kinship, and global psychological variation. Science, 366(707), 1-12. https://doi.org/10.1126/science.aau5141

Seccombe, W. (1992). A Millennium of Family Change. Feudalism to Capitalism in Northwestern Europe. London: Verso. https://archive.org/details/millenniumoffami0000secc/page/n3/mode/2up

Tacitus (1970). Agricola, Germania, Dialogus. Loeb Classical Library (Translated by M. Hutton & W. Peterson). Cambridge, MA: Harvard University Press.

Wang, K., Tobias, B., Pany-Kucera, D., Berner, M., Eggers, S., Gnecchi-Ruscone, G. A., ... & Hofmanová, Z. (2025). Ancient DNA reveals reproductive barrier despite shared Avar-period culture. Nature, 1-8. https://doi.org/10.1038/s41586-024-08418-5

Wendt, F. R., Warrier, V., Pathak, G. A., Koenen, K. C., Stein, M. B., Krystal, J. H., ... & Polimanti, R. (2022). Polygenic scores for empathy associate with posttraumatic stress severity in response to certain traumatic events. Neurobiology of Stress, 17, 100439. https://doi.org/10.1016/j.ynstr.2022.100439

Malcolm Gladwell, author and staff writer at The New Yorker

When I first published my last post in Aporia Magazine, a reader told me that Black mother/White father children not only have worse birth outcomes than White mother/Black father children but also are eight points lower in IQ.

The source for this fact? Malcolm Gladwell, a staff writer at The New Yorker and a popular author. Eight years ago, in an article for this magazine, he argued that genetics cannot explain IQ differences between human populations, particularly the gap between Euro and African Americans. He recalled two points that Flynn had made in a debate on this subject with Charles Murray.

First, the IQ gap differs by age:

Flynn took a different approach. The black-white gap, he pointed out, differs dramatically by age. He noted that the tests we have for measuring the cognitive functioning of infants, though admittedly crude, show the races to be almost the same. By age four, the average black I.Q. is 95.4—only four and a half points behind the average white I.Q. Then the real gap emerges: from age four through twenty-four, blacks lose six-tenths of a point a year, until their scores settle at 83.4. (Gladwell, 2007)

Second, the IQ gap is affected by a strange interaction between race and gender:

Flynn then talked about what we’ve learned from studies of adoption and mixed-race children—and that evidence didn’t fit a genetic model, either. If I.Q. is innate, it shouldn’t make a difference whether it’s a mixed-race child’s mother or father who is black. But it does: children with a white mother and a black father have an eight-point I.Q. advantage over those with a black mother and a white father. (Gladwell, 2007)

Here, Flynn was following the argument he had made in an American Psychologist article on the same subject (Flynn, 1999). But that article has no comparison between Black mother/White father children and White mother/Black father children. Nor does any other article of his.

Gladwell probably heard something like the following, in which Flynn summarized Klaus Eyferth’s study of illegitimate children born to German mothers and fathered by American soldiers:

The soldiers of the American occupation force in Germany, both White and Black, fathered thousands of children with German women after World War II. Eyferth (see Flynn, 1980) selected a representative sample of 181 Black children and a matching group of 83 White children and found that their mean IQs were virtually identical. In other words, children who had a White father seemed to possess no advantage whatsoever over those who had a Black father. (Flynn, 1999, p. 14)

Evidently, the mothers were White in both groups.

A reader has brought to my attention another study, not mentioned by Flynn in his American Psychologist article, that corresponds more closely to what Gladwell remembered. Willerman et al. (1974) administered a cognitive test at 4 years of age to 101 biracial children who had White mothers with 28 who had Black mothers. The first group averaged about 9 IQ points higher than the second.

Was Gladwell inserting this finding into his memory of Flynn’s presentation? It’s a dubious finding at best.

• First, the black mother sample is small. The rule of thumb is to have a minimum of 30 to 40 individuals for statistical comparison (Varshney, 2023).

• Second, educational level was higher for both parents when the mother was white than when she was black. This, in itself, is a red flag, since the situation is usually the reverse (see my previous post). The authors tried to control for this difference by controlling for maternal education but not for both maternal and paternal education. If you’re trying to rule out a genetic explanation—and this was the intent of the study—you should control for the education of both parents, since educational attainment correlates highly with the genetic component of cognitive ability, and since both parents would presumably be passing on this genetic component. This adjustment was not done because “because educational data on 25% of the fathers were lacking.”

• As will be shown further in this post, IQ is much more malleable at younger ages than at older ages. IQ before adulthood, and even more so during early childhood, is more strongly influenced by the family environment.

But other than that …

Does Malcolm Gladwell nonetheless have a point with respect to the Eyferth study? If genetic factors are so important, why did the biracial children in that study have the same mean IQ as the White children?

Three reasons were given by Rushton and Jensen (2005, p. 261):

• The US Army screened out low-IQ enlistees through the Army General Classification Test, essentially an IQ test. The rejection rate was about 30% for African Americans and 3% for Euro Americans. The African American soldiers in Germany were thus a higher-IQ sample of the African American population.

• Between 20 and 25% of the “black” fathers were, in fact, North African, i.e., Moroccan and Algerians soldiers serving in the French army of occupation.

• The children were still young when tested. One third were between 5 and 10 years old and two thirds between 10 and 13. Since IQ is strongly influenced by family environment before puberty, a much larger sample would have been needed to show a significant difference between the two groups.

The last point leads us to a related one: the capacity for learning seems to be greater in children than in adults. This might explain one aspect of the Minnesota Transracial Adoption Study (Weinberg et al., 1992; see also Levin, 1994; Lynn, 1994). In the enriched learning environment of middle-class Minnesota families, all of the children showed impressive IQ scores by 7 years of age. By 17 years of age, however, this IQ boost had largely faded out:

Mean IQs of participants in the Minnesota Transracial Adoption Study, by race and by age

Discussion has focused on the finding that the relative positions of the four groups remained unchanged between age 7 and age 17. Mean IQ declined, however, in all four groups. Weinberg et al. (1992) attributed the decline to a renorming of the IQ test from the WAIS to the WAIS-R, noting that this change had typically caused declines of the same order in other studies. On the other hand, the Flynn effect should have caused a 3-point increase over the intervening 10 years. Also, the WAIS was only one of three IQ tests used in the study:

In interpreting this decline, one must keep in mind that different IQ tests were employed at Time 1 and Time 2. Depending on their ages, family members were tested using the Stanford-Binet Form L-M (Terman & Merrill, 1973), WISC, or WAIS in the original study, and using the WISC-R or WAIS-R in the follow-up study. (Weinberg et al.,1992, p. 130)

The authors contented themselves with the conclusion that “most of this decline” could be explained by the WAIS renorming.

We know that the capacity to learn and retain information declines with age, as noted by IQ researcher David Weschler:

Beginning with the investigation by Galton in 1883 and continuing up to and including the most recent studies of Pacaud, nearly all studies dealing with the age factor in adult performance have shown that most human abilities, in so far as they are measurable, decline progressively, after reaching a peak somewhere between ages 18 and 25 (Weschler, 1958).

But the peak came earlier in the Minnesota study. Perhaps the peak between 18 and 25 is preceded by a gradual lowering of the ceiling for learning saturation. This ceiling may be more often reached by children in the enriched learning environment of adoptive middle-class families. A similar decline has been noticed with Head Start in the U.S. — a preschool educational program that has typically shown impressive IQ gains that fade out as the children get older (Barnett et al., 2005).

The same is true for offspring in nonhuman species. An animal is more receptive to learning during infancy, when it is exploring its world. Its capacity for learning then declines as it gets older. In dogs, age-related cognitive decline seems to be independent of variation in physiological aging and lifespan among different breeds (Watowich et al., 2020). Dogs lose their ability to learn because it is less advantageous for them to learn as they transition from infancy to adulthood.

In our species, the decline is weaker and pushed forward in time because it is more advantageous for us to keep learning.

References

Barnett, W. S., & Hustedt, J. T. (2005). Head Start's lasting benefits. Infants & Young Children, 18(1), 16-24. https://psycnet.apa.org/doi/10.1097/00001163-200501000-00003

Caplan, B. (2007). Gladwell on IQ: My Chance to Answer Two Plausible But Over-rated Arguments, EconLog, December 14. https://www.econlib.org/archives/2007/12/gladwell_on_iq.html

Eyferth, K. (1961). Leistungen verscheidener Gruppen von Besatzungskindern in Hamburg-Wechsler Intelligenztest für Kinder (HAWIK). Archiv für die gesamte Psychologie, 113, 222-241.

Flynn, J. R. (1999). Searching for justice: the discovery of IQ gains over time. American Psychologist, 54(1), 5. https://psycnet.apa.org/doi/10.1037/0003-066X.54.1.5

Frost, P. (2019). IQ of biracial children and adults, Evo and Proud, March 10. https://evoandproud.blogspot.com/2019/03/iq-of-biracial-children-and-adults.html

Gladwell, M. (2007). None of the Above. The New Yorker. https://www.newyorker.com/magazine/2007/12/17/none-of-the-above

Levin, M. (1994). Comment on the Minnesota transracial adoption study. Intelligence, 19(1), 13-20. https://doi.org/10.1016/0160-2896(94)90049-3

Lynn, R. (1994). Some reinterpretations of the Minnesota transracial adoption study. Intelligence, 19(1), 21-27. https://doi.org/10.1016/0160-2896(94)90050-7

Rushton, P. & Jensen, A.R. (2005). Thirty years of research on race differences in cognitive ability. Psychology, Public Policy, and Law, 11, 235-294. https://psycnet.apa.org/doi/10.1037/1076-8971.11.2.235 https://isgp-studies.com/miscellaneous/race-and-iq/2005-APA-Rushton-Jensen-Thirty-Years-of-Research-on-Race-Differences-in-Cognitive-Ability.pdf

Varshney, D. (2023). Re: What is the minimum sample size for t-test?. Retrieved from: https://www.researchgate.net/post/What_is_the_minimum_sample_size_for_t-test/6468fa914e2ff22fde0e5fd8/citation/download.

Watowich, M. M., MacLean, E. L., Hare, B., Call, J., Kaminski, J., Miklósi, Á., & Snyder-Mackler, N. (2020). Age influences domestic dog cognitive performance independent of average breed lifespan. Animal Cognition, 23, 795-805. https://doi.org/10.1007/s10071-020-01385-0

Wechsler, D. (1958). Changes in Intelligence and Intellectual Ability with Age. In D. Wechsler, The measurement and appraisal of adult intelligence (4th ed., pp. 135–143). Williams & Wilkins Co. https://doi.org/10.1037/11167-009

Weinberg, R. A., Scarr, S., & Waldman, I. D. (1992). The Minnesota Transracial Adoption Study: A follow-up of IQ test performance at adolescence. Intelligence, 16(1), 117-135. https://doi.org/10.1016/0160-2896(92)90028-P

Wikipedia. (2025). Malcolm Gladwell. https://en.wikipedia.org/wiki/Malcolm_Gladwell

Willerman, L., Naylor, A. F., & Myrianthopoulos, N. C. (1974). Intellectual development of children from interracial matings: Performance in infancy and at 4 years. Behavior Genetics, 4(1), 83-90. https://doi.org/10.1007/BF01066706

Free woman of color with daughter, New Orleans. Late 18th century (Wikicommons)

You received the same amount of DNA from your father and your mother, but only your mother provided the womb you grew in. To some extent, your mother’s genes matter more than your father’s, since hers alone helped create the environment for your first nine months. The genetic influence is substantial: heritability is 25-40% for birthweight, fetal growth, and gestational length (Clausson et al., 2005).

Genetic mismatches may therefore arise between the fetus (whose development is determined by both the mother and the father) and the fetal environment (which is created by the mother), especially if the parental populations differ in the pace and timing of fetal development.

If a fetus has European parents, it will come to term a week later, on average, than one with Sub-Saharan African parents (Rushton & Jensen, 2005, p. 264). Sub-Saharan African descent is a risk factor for spontaneous pre-term birth in both the United States and Africa, although “risk” here is defined in terms of European norms of pregnancy duration (Purisch & Gyamfi-Bannerman, 2017).

For a fetus, pre-term delivery is riskier than post-term delivery (Campbell et al., 1997; Fayed et al., 2022). Outcomes should therefore be worse for births to Black mothers and White fathers than for births to White mothers and Black fathers. Indeed, this has been a consistent finding of American studies over the past four decades.

Biracial births, Chicago, 1982-1983

The first study is of all singleton infants born to biracial couples in Chicago from 1982 to 1983 (n = 410 Black mothers, 739 White mothers).

Birthweights were low for 14% of births to Black mothers and White fathers, versus 9% of births to White mothers and Black fathers and 6% of births to White parents.

The authors adjusted the data to control for several risk factors: maternal age, education, marital status, parity (number of pregnancies that reached a viable gestational age), trimester of prenatal care initiation, and median family income of mother's census tract. This adjustment eliminated the difference in risk between White mother/Black father births and White mother/White father births, while maintaining the higher risk of Black mother/White father births:

We found that infants born to Black mothers and White fathers had a 40% higher chance of low-birthweight than did White infants, independent of other risk factors. In contrast, infants born to White mothers and Black fathers had odds of low birthweight equal to those of infants in the general White population when measured risk factors were controlled. (Collins & David, 1993)

The authors concluded that “traditional sociodemographic risk factors do not explain the birthweight disadvantage of biracial infants born to Black mothers. Indeed, our findings suggest that these infants may also be at increased risk for prematurity and intrauterine growth retardation. Other unmeasured variables appear to be exerting a negative effect only in biracial births in which the Black parent is the mother” (Collins & David, 1993)

Biracial births, California, 1992

The next study is of all births to Black, White and biracial couples in California in 1992 (n = 203,815 couples).

The data were adjusted to control for a wider range of risk factors: maternal characteristics (race, age, education, marital status, parity, obstetric history, tobacco use, medical complications, medical insurance and use of prenatal care); paternal characteristics (race, age and education); infant characteristics (gestational age and gender); and community characteristics (median household income from the 1990 US Census).

These adjustments reduced the overall race differences, particularly for very low birthweight infants, but there was now a relatively larger difference between Black mother/White father births and White mother/Black father births: “In the adjusted models analyzing moderately low birth weight infants, only black maternal race was a statistically significant risk factor.” Indeed, having a Black mother seems to explain the low weights of Black mother/White father births just as well as the low weights of Black mother/Black father births.

Yet, for the authors, the cause could not be genetic: “These results imply that black race per se is not a causal risk factor for low birth weight infants but rather that race is a surrogate marker for other causal risk factors, both measured and unmeasured” (Hessol et al., 1998, pp. 819-820).

Biracial births, US, 1978–1997

The next study is of all singleton births to American couples between 1978 and 1997 (n = 50 M).

Between 1978 and 1997, outcomes improved for White mother/Black father births relative to single-race White births, but not for Black mother/White father births relative to single-race Black births (Parker, 2000). The latter finding was unexpected, given the generally better circumstances of biracial infants with White fathers:

By 1994–1997, based on these indices, interracial births to black mothers had lower-risk demographic profiles than interracial births to white mothers. Black mothers with interracial births had more education, were older, were more likely to be foreign born, and were much more likely to be married than white mothers with interracial births. (Parker, 2000)

The author concluded:

This difference between black and white mothers suggests that factors specifically affecting interracial birth differ by maternal race and have changed little for black mothers. Indeed, although demographically a lower-risk group, black mother/white father infants had consistently higher risks of LBW [low birthweight] and VLBW [very low birthweight] than white mother/black father infants. This finding may be due to persistent effects of racial discrimination throughout childhood and adulthood or factors unique to black women, regardless of the race of their infant’s father. (Parker, 2000)

White, Black, Hispanic, and mixed-race births, Colorado, 1989–2000

The next study is of all singleton births to couples in Colorado between 1989 and 2000 (n = 91,061). Mothers were healthy, non-smoking, non-substance-using women who were pregnant for the first time.

Infants with Black mothers were smaller for gestational age (SGA) than infants with White mothers, while father’s race had no effect:

It is also noteworthy that paternal race/ethnicity entered the model for infant birth weight but not the model for SGA delivery. This suggests that paternal race/ethnicity contributes to normal variation in fetal growth but not to the clinically significant aberrations in fetal growth that are responsible for racial/ethnic disparities in neonatal morbidity and mortality in the US. (Sheeder et al., 2006, p. 387)

As in the previous study, the authors ruled out a genetic cause: “The negative findings, with respect to the effect of paternal race/ ethnicity on the risk of SGA delivery, also increase confidence that the effects of maternal race/ethnicity reflect acquired, not genetic, traits” (Sheeder et al., 2006, p. 387)

Twin biracial births, US, 1995–1997

The next study is of all twins born to biracial parents in the US from 1995 to 1997 (n = 702 Black mothers, 1,923 White mothers).

Black mother/White father infants were at higher risk than White mother/Black father infants for very low birthweight, very pre-term birth, fetal growth restriction, and both fetal and infant mortality.

Unfortunately, the authors did not distinguish between identical and fraternal twins, nor did they adjust for socioeconomic risk factors. This study therefore cannot tell us whether genetics or environment better explains the differences between Black and White maternal inheritance (Tan et al., 2004a; Tan et al., 2004b).

Biracial births, US, 1991

The next study is of all births to biracial American couples in 1991 (n = 50,980).

Low birthweight was 31% more frequent among Black mother/White father births (8.4%) than among White mother/Black father births (6.4%). The difference was smaller in the Northeast of the country, perhaps because a higher proportion of the White mothers were Puerto Ricans with some Black admixture (Polednak & King, 1998).

Biracial fetal deaths, California, 1998–2002

The next study is of all stillbirths to biracial couples in California between 1998 and 2002 (n = 0.01 million Black mothers, 0.03 million White mothers).

Risk of death was higher for biracial fetuses with Black mothers/White fathers than for those with White mothers/Black fathers (Gold et al., 2010).

Biracial births, US, 2001

The last study is of all births to biracial American couples in 2001 (n = 10,955 Black mothers, 42,609 White mothers).

“In black-white couples, the odds of preterm birth were greater if the mother was black … than if the father was black” (Simhan & Krohn, 2008).

Discussion

In sum, biracial infants with Black mothers have worse birth outcomes than those with White mothers, i.e., lower birthweight, higher rate of premature birth, poorer fetal growth, and higher fetal and infant mortality. These differences remain after adjustment for socioeconomic risk factors, even though the same adjustment equalizes birth outcomes between infants born to White mothers and Black fathers and those born to White mothers and White fathers.

If the cause is an environmental factor that adversely affects Black mothers, why did it remain unchanged between 1978 and 1997? That period saw major social and economic gains for Black women, particularly those in interracial relationships.

A review article has pointed to a possible genetic factor: the maternal C825T allele of the GNB3 gene (Van Den Oord & Rowe, 2001). This allele lowers birth weight and is much less frequent in Europeans (about 30%) than in Africans and African Americans (up to 80%). Because infants receive less paternal care in the highly polygynous societies of Sub-Saharan Africa, a developing fetus may have a better chance of survival if it matures faster:

On average, Black babies are born a week earlier than White babies, yet they are more mature as measured by pulmonary function, amniotic fluid, and bone development. In the United States, 51% of Black children have been born by week 39 of pregnancy compared with 33% of White children. Black African babies, even those born to mothers in the professional classes, are also born earlier than White babies. (Rushton & Jensen, 2005, p. 264)

If the cause is a genetic factor, a biracial fetus should mature more slowly than a Black one, since one of the parents has genes for slower fetal development. A conflict may thus arise between a biracial fetus and its womb if the mother is Black. Specifically, the womb may initiate the birth process before the fetus can survive on its own, thus increasing the risks of stillbirth and infant mortality.

This is suggested by the first of the above studies on biracial infants. Those born to Black mothers and White fathers, as opposed to White mothers and Black fathers, were not simply born earlier. They were also at increased risk for prematurity, uterine growth retardation, and other adverse effects (Collins & David, 1993).

A biracial fetus in a White womb has fewer risks. It is already mature enough to survive on its own when the womb initiates the birth process. Consequently, post-term births incur fewer risks than pre-term births (Marroun et al., 2012), with the possible exception of a higher risk of obesity for adolescent post-term boys (Beltrand et al., 2012; Schierding et al., 2014).

What about infants born to African immigrants?

Some authors still lean toward an environmental cause for the lower birthweight and higher pre-term birth rate of African Americans. This view is supported by two studies that found a higher birthweight and lower pre-term birth rate for infants born to African immigrants than for those born to African Americans. Since both groups presumably come from the same gene pool, any differences between the two must be environmental in nature.

The first study showed that infants of African immigrants weighed more than those of African Americans, although both groups still weighed less than infants of Euro Americans (David & Collins, 1997). Clearly, low African American birthweights have a partly environmental cause. This was shown by Hessol et al. (1998) when they adjusted their data for socioeconomic factors and thereby reduced the difference in risk for moderately low birthweight between African American births and Euro American births. Nonetheless, the first group still had a 60% higher risk than the second. This residual probably corresponds to the maternal genetic factor that likewise explains the earlier timing and lower weight of Black mother/White father births.

The second study showed that the pre-term birth rate did not differ at all between infants of Euro Americans and those of African immigrants, after adjustment for socioeconomic differences (Braveman et al., 2024). This study, however, had participants from the entire African continent, including Algeria, Djibouti, Egypt, Eritrea, Ethiopia, and Libya. No information is given on the number of participants from each African country, but we do have information on the African immigrant community in California, where the study was conducted. The community’s leading country of origin is Ethiopia, followed by Nigeria, Egypt, and South Africa (Reimers, 2005, p. 243). Thus, most of the participants were probably born outside West Africa, the main ancestral region of African Americans. So we are not looking at the same gene pool.

In addition, most African immigrants to the U.S. have some college education, 69% vs. 63% for Americans in general, partly because many come as students and partly because immigrant selection favors the highly educated (Anderson & Connor, 2018). African immigrants may thus be selected for slower life history and, hence, slower fetal development. This selection would not fully disappear after adjustment for socioeconomic factors, since this adjustment is based on data from American society. In African societies, the college-educated differ much more from the general population because proportionately fewer people go on to higher education.

The same point applies to the other studies that adjusted for socioeconomic factors. Such adjustment gives more weight to those African Americans who have slower life history and, hence, a different genetic makeup. Race differences are therefore reduced, since Euro Americans are now being compared with African Americans who have a more similar life history.

Avenues for future research

Conflicts between maternal and fetal programming may affect fetal development in other ways. According to a study of malformed infants born to biracial couples in California between 1989 and 2000 (n = 50 Black mothers, 150 White mothers), biracial infants are at higher risk for cleft palates, hypospadias, and polydactyly when the mother is Black than when she is White. On the other hand, the risk of hypertrophic pyloric stenosis is lower when she is Black (Yang et al., 2004). This is admittedly a single study with a limited number of cases.

Although a post-term birth is generally less risky for a newborn baby than a pre-term birth, adolescent post-term boys may be at higher risk for obesity (Beltrand et al., 2012; Schierding et al., 2014). It would be interesting to see whether this outcome is worse for biracial boys with White mothers than for those with Black mothers.

Finally, fetal development may be disrupted not only by maternal/fetal conflicts but also by maternal/paternal conflicts, perhaps in genes that influence appetite and metabolism. A California/Hawaii study found higher levels of obesity in ethnically mixed individuals than in either parental population, even after controlling for psychosocial and lifestyle factors, i.e., age, marital status, education, and smoking status:

For instance, the prevalence of overweight/obesity in five ethnic admixtures—Asian/white, Hawaiian/white, Hawaiian/Asian, Latina/white, and Hawaiian/Asian/white ethnic admixtures—was significantly higher (P < 0.0001) than the average of the prevalence estimates for their component ethnic groups.

… Controlling for psychosocial and lifestyle factors did not attenuate the differences in prevalence between ethnic admixtures and monorace adults. However, a high caloric intake (e.g., calories from fat and alcohol) and exercise did modestly decrease this difference, and could be important factors for future interventions to control obesity in mixed-race individuals. (Albright et al., 2008)

This sort of mismatch may explain why fertility seems to peak in marriages between third or fourth cousins and then falls progressively for parents who are decreasingly less related to each other. As genetic distance increases, so does the risk of maternal/paternal mismatches in a developing embryo (Frost, 2024).

References

Albright, C. L., Steffen, A., Wilkens, L. R., Henderson, B. E., & Kolonel, L. N. (2008). The prevalence of obesity in ethnic admixture adults. Obesity, 16 (5), 1138–1143. https://doi.org/10.1038/oby.2008.31

Anderson, M., & Connor, P. (2018). Sub-Saharan African Immigrants in the U.S. Are Often More Educated Than Those in Top European Destinations, Pew Research Center, April 24. https://www.pewresearch.org/internet/2018/04/24/sub-saharan-african-immigrants-in-the-u-s-are-often-more-educated-than-those-in-top-european-destinations/

Beltrand, J., Soboleva, T. K., Shorten, P. R., Derraik, J. G., Hofman, P., Albertsson-Wikland, K., ... & Cutfield, W. S. (2012). Post-term birth is associated with greater risk of obesity in adolescent males. The Journal of Pediatrics, 160(5), 769-773. https://doi.org/10.1016/j.jpeds.2011.10.030

Braveman, P., Heck, K., Dominguez, T. P., Marchi, K., Burke, W., & Holm, N. (2024). African immigrants’ favorable preterm birth rates challenge genetic etiology of the Black-White disparity in preterm birth. Frontiers in Public Health, 11, 1321331. https://doi.org/10.3389/fpubh.2023.1321331

Campbell, M. K., Østbye, T., & Irgens, L. M. (1997). Post-term birth: risk factors and outcomes in a 10-year cohort of Norwegian births. Obstetrics & Gynecology, 89(4), 543-548. https://doi.org/10.1016/S0029-7844(97)00049-5

Clausson, B., Lichtenstein, P., & Cnattingius, S. (2005). Genetic influence on birthweight and gestational length determined by studies in offspring of twins. BJOG, 107, 375-381. https://doi.org/10.1111/j.1471-0528.2000.tb13234.x

Collins, Jr, J.W., & David, R.J. (1993). Race and birthweight in biracial infants. American Journal of Public Health, 83(8), 1125-1129. https://doi.org/10.2105/AJPH.83.8.1125     

David, R. J., & Collins Jr, J. W. (1997). Differing birth weight among infants of US-born blacks, African-born blacks, and US-born whites. New England Journal of Medicine, 337(17), 1209-1214. https://doi.org/10.1056/nejm199710233371706

Fayed, A., Wahabi, H. A., Esmaeil, S., Elmorshedy, H., & AlAniezy, H. (2022). Preterm, early term, and post-term infants from Riyadh mother and baby multicenter cohort study: The cohort profile. Frontiers in Public Health, 10, 928037. https://doi.org/10.3389/fpubh.2022.928037

Frost, P. (2024). Outbreeding depression: Avenues for further research. Aporia Magazine, February 28.

Gold, K.J., DeMonner, S.M., Lantz, P.M., & Hayward, R.A. (2010). Prematurity and low birth weight as potential mediators of higher stillbirth risk in mixed black/white race couples. Journal of Women's Health, 19(4), 767-773. https://doi.org/10.1089/jwh.2009.1561

Hessol, N. A., Fuentes-Afflick, E., & Bacchetti, P. (1998). Risk of low birth weight infants among black and white parents. Obstetrics & Gynecology, 92(5), 814-822. https://doi.org/10.1016/s0029-7844(98)00310-x

Marroun El, H., Zeegers, M., Steegers, E., van der Ende, J., Schenk, J. J., Hofman, B., Jaddoe, V., Verhulst, F., & Tiemeier, H. (2012). Post-term birth and the risk of behavioural and emotional problems in early childhood. International Journal of Epidemiology, 41(3), 773-781. https://doi.org/10.1093/ije/dys043

Parker, J.D. (2000). Birth Weight Trends Among Interracial Black and White Infants. Epidemiology, 11(3), 242-248. https://doi.org/10.1097/00001648-200005000-00003

Polednak, A.P., & King, G. (1998). Birth weight of US biracial (black-white) infants: regional differences. Ethnicity & Disease, 8(3), 340-349.

Purisch, S. E., & Gyamfi-Bannerman, C. (2017). Epidemiology of preterm birth. In: Seminars in perinatology (Vol. 41, No. 7, pp. 387-391). WB Saunders. https://doi.org/10.1053/j.semperi.2017.07.009

Reimers, D. (2005). Other Immigrants: the Global Origins of the American People, NYU Press.

Rushton, J.P., & Jensen, A.R. (2005). Thirty years of research on race differences in cognitive ability. Psychology, Public Policy, and Law, 11(2), 235-294. https://psycnet.apa.org/doi/10.1037/1076-8971.11.2.235  

Schierding, W., O’Sullivan, J. M., Derraik, J. B., & Cutfield, W. S. (2014). Genes and post-term birth: late for delivery. BMC Research Notes, 7, 1-5. https://doi.org/10.1186/1756-0500-7-720

Sheeder, J., Lezottte, D., & Stevens-Simon, C. (2006). Maternal age and the size of white, black, Hispanic, and mixed infants. Journal of Pediatric and Adolescent Gynecology, 19(6), 385-389. https://doi.org/10.1016/j.jpag.2006.09.012

Simhan, H.N., & Krohn, M.A. (2008). Paternal race and preterm birth. American Journal of Obstetrics and Gynecology, 198(6), 644-e1. https://doi.org/10.1016/j.ajog.2007.11.046

Tan, H., Wen, S.W., Walker, M., & Demissie, K. (2004a). Parental race, birth weight, gestational age, and fetal growth among twin infants in the United States. Early Human Development, 80(2), 153-160. https://doi.org/10.1016/j.earlhumdev.2004.06.005

Tan, H., Wen, S.W., Walker, M., & Demissie, K. (2004b). The effect of parental race on fetal and infant mortality in twin gestations. Journal of the National Medical Association, 96(10), 1337.

Van Den Oord, E.J., & Rowe, D.C. (2001). A step in another direction: Looking for maternal genetic and environmental effects on racial differences in birth weight. Demography, 38(4), 573-576. https://doi.org/10.1353/dem.2001.0040

Yang, J., Carmichael, S.L., Kaidarova, Z., & Shaw, G.M. (2004). Risks of selected congenital malformations among offspring of mixed race-ethnicity. Birth Defects Research Part A: Clinical and Molecular Teratology, 70(10), 820-824. https://doi.org/10.1002/bdra.20054

Franz Boas, c. 1915 (Canadian Museum of History)

The anthropologist Franz Boas (1858-1942) is seen as a key figure in the development of current views on race and racism. Historian Carl Degler and psychologist Kevin MacDonald describe him as leading a “life-long assault on the idea that race was a primary source of the differences to be found in the mental or social capabilities of human groups” (Degler, 1991, p. 61; MacDonald, 1998, p. 23).

This view has become widespread among critics and admirers alike. Yet it is largely false.

In fact, Boas argued repeatedly for the idea that mental capability differs among human groups, a position he abandoned only with the rise of Nazism in the 1930s. After his death in 1942, and the end of the Second World War, his life work would be used to justify the emerging antiracist consensus. He has since become a mythical founder of that consensus, and his real role in history — as a man searching for the truth — has been obscured to the point of misrepresentation.

Boas on race and racial differences

In 1858, Boas was born into a non-observant Jewish family in Germany. He first took an interest in human differences while studying geography, particularly during his fieldwork among the Inuit of Baffin Island (1883-1884). He then worked for three years at the Royal Ethnological Museum in Berlin before leaving for an academic position in the United States.

Boas first presented his views on racial differences in an 1894 speech on ‘Human Faculty as Determined by Race.’

It does not seem probable that the minds of races which show variations in their anatomical structure should act in exactly the same manner. Differences of structure must be accompanied by differences of function, physiological as well as psychological; and, as we found clear evidence of difference in structure between the races, so we must anticipate that differences in mental characteristics will be found. Thus, a smaller size or lesser number of nervous elements would probably entail loss of mental energy, and paucity of connections in the central nervous system would produce sluggishness of the mind. (Boas, 1894, p. 25)

He did not limit the race concept to continent-wide groups. For him, a “race” could simply be a clan that ends up diverging genetically from other clans through the reproductive success of its leading families:

Furthermore, as certain anatomical traits are found to be hereditary in certain families and hence in tribes and perhaps even in peoples, in the same manner mental traits characterize certain families and may prevail among tribes. (Boas, 1894, p. 25)

But were such traits passed on genetically or culturally?

It seems, however, an impossible undertaking to separate in a satisfactory manner the social and the hereditary features. Galton's attempt to establish the laws of hereditary genius points out a way of treatment for these questions which will prove useful in so far as it opens a method of determining the influence of heredity upon mental qualities (Boas, 1894, p. 25)

These mental differences are statistical in nature, with significant overlap between racial groups:

We have shown that the anatomical evidence is such, that we may expect to find the races not equally gifted. While we have no right to consider one more ape-like than the other, the differences are such that some have probably greater mental vigor than others. The variations are, however, such that we may expect many individuals of all races to be equally gifted, while the number of men and women of higher ability will differ. (Boas, 1894, p. 28)

The above points would be expanded upon in his 1901 article about ‘The Mind of Primitive Man.’

The thoughts and actions of civilized man and those found in more primitive forms of society prove that, in various groups of mankind, the mind responds quite differently when exposed to the same conditions. Lack of logical connection in its conclusions, lack of control of will, are apparently two of its fundamental characteristics in primitive society. In the formation of opinions, belief takes the place of logical demonstration. The emotional values of opinions is great, and consequently they quickly lead to action. The will appears unbalanced, there being a readiness to yield to strong emotions, and a stubborn resistance in trifling matters. (Boas, 1901, p. 1)

Boas believed that humans differ mentally and emotionally for two possible reasons: 1) their minds are organized differently or 2) their minds are subjected to differences in “habit” and “individual experience.” Between these two explanations, he steered a middle course:

A number of anatomical facts point to the conclusion that the races of Africa, Australia, and Melanesia are to a certain extent inferior to the races of Asia, America, and Europe. We find that on the average the size of the brain of the negroid races is less than the size of the brain of the other races; and the difference in favor of the mongoloid and white races is so great that we are justified in assuming a certain correlation between their mental ability and the increased size of their brain. At the same time it must be borne in mind that the variability of the mongoloid and white races on the one hand, and of the negroid races on the other, is so great that only a small number, comparatively speaking, of individuals belonging to the latter have brains smaller than any brains found among the former; and that, on the other hand, only a few individuals of the mongoloid races have brains so large that they would not occur at all among the black races …

The question whether the power to inhibit impulses is the same in all races of man is not so easily answered. It is an impression obtained by many travellers, and also based upon experiences gained in our own country, that primitive man and the less educated have in common a lack of control of emotions, that they give way more readily to an impulse than civilized man and the highly educated. I believe that this conception is based largely upon the neglect to consider the occasions on which a strong control of impulses is demanded in various forms of society …. [Taboos and food prohibitions] are very numerous, and prove that primitive man has the ability to control his impulses, but that this control is exerted on occasions which depend upon the character of the social life of the people, and which do not coincide with the occasions on which we expect and require control of impulses …

It is not impossible that the degree of development of these functions may differ somewhat among different types of man; but I do not believe that we are able at the present time to form a just valuation of the power of abstraction, of control, and of choice among different races. (Boas, 1901, pp. 3-6)

This article closes with a discussion of cultural relativism, particularly how human groups differ in their notions of right and wrong. Although such differences are now usually seen as culturally determined — each culture conditions its members to think in a certain way — Boas saw the determination as being both cultural (“traditions”) and innate (“equilibrium of emotion and reason”).

It is somewhat difficult for us to recognize that the value which we attribute to our own civilization is due to the fact that we participate in this civilization, and that it has been controlling all our actions since the time of our birth; but it is certainly conceivable that there may be other civilizations, based perhaps on different traditions and on a different equilibrium of emotion and reason, which are of no less value than ours … (Boas, 1901, p. 11)

This article was the basis for a book of the same title, first published in 1911. Many more people are familiar with the 1921 and 1938 editions, which downplay or eliminate the racialist statements of the first edition.

In 1905, Boas wrote an article on ‘The Negro and the Demands of Modern Life’ for a special issue of Charities, a magazine aimed at a “familiar cast of Christian reformers, social workers, and social scientists” (Baker, 2022). Even to that audience, he affirmed that the average negro brain is "smaller than that of other races" and that it was "plausible that certain differences of form of brain exist." He then qualified this statement by adding: “We must remember that individually the correlation ... is often overshadowed by other causes, and that we find a considerable number of great men with slight brain weight.” He nonetheless concluded: “We may, therefore, expect less average ability and also, on account of probable anatomical differences, somewhat different mental tendencies” (Williams, 1996, p. 17).

That same year, he discussed "the desirability of collecting more definite information in relation to certain traits of the Negro race that seem of fundamental importance in determining the policy to be pursued towards that race" (Williams, 1996, p. 18). In particular, he had two questions in mind:

(1) Is there an earlier arrest of mental and physical development in the Negro child, as compared with the white child? And, if so, is this arrest due to social causes or to anatomical and physiological conditions?

(2) What is the position of the mulatto child and of the adult mulatto in relation to the two races? Is he an intermediate type, or is there a tendency of reversion towards either race? So that particularly gifted mulattoes have to be considered as reversals of the white race. The question of the physical vigor of the mulatto could be taken up at the same time. (Williams, 1996, p. 19)

In a private letter, he spoke more openly:

You may be aware that in my opinion the assumption seems justifiable that on the average the mental capacity of the negro may be a little less than that of the white, but that the capacities of the bulk of both races are on the same level. (Williams, 1996, p. 19)

These views were later reiterated in his 1909 article on ‘Race Problems in America.’

I do not believe that the negro is, in his physical and mental make-up, the same as the European. The anatomical differences are so great that corresponding mental differences are plausible. There may exist differences in character and in the direction of specific aptitudes. There is, however, no proof whatever that these differences signify any appreciable degree of inferiority of the negro, notwithstanding the slightly inferior size, and perhaps lesser complexity of structure, of his brain; for these racial differences are much less than the range of variation found in either race considered by itself. (Boas, 1909, pp. 847-848)

Boas and anti-racism

Were the above statements anti-racist for their time? Not really. Many contemporaries of Boas held strong views on human equality, including some in his entourage. In Berlin, he had done post-graduate work under the anthropologist Rudolf Virchow, who rejected Darwinian evolution and attributed human differences to the direct action of the environment. Boas had also studied under the ethnologist Adolf Bastian, who argued for the “psychic unity of mankind,” i.e., that all humans have the same intellectual capacity, and all cultures are based on the same basic mental principles (Wikipedia, 2025). In the United States, Boas would work with academics who supported the goal of making African Americans free and equal citizens. Although Reconstruction had been mostly abandoned by the turn of the century, a large segment of Northern opinion remained sympathetic to its aims and hotly contested the disenfranchisement of Black Southerners.

Thus, during his long tenure at Columbia University (1896-1942), Boas often expressed his views to people who rejected the possibility of innate mental differences. In such situations, he tailored his message to his audience, like the African Americans who attended his 1906 commencement address at Atlanta University:

The physical inferiority of the Negro race, if it exists at all, is insignificant when compared to the wide range of individual variability in each race. There is no anatomical evidence available that would sustain the view that the bulk of the Negro race could not become as useful citizens as the members of any other race. That there may be slightly different hereditary traits seems plausible, but it is entirely arbitrary to assume that those of the Negro, because perhaps slightly different, must be of an inferior type. (Boas, 1906)

This address was made at the invitation of W.E.B. Du Bois, who had read the contribution by Boas to the special issue of Charities, having himself contributed to the same issue. It was the start of a relationship that would open many doors for the aspiring anthropologist, notably an invitation to publish in The Crisis, the newly created magazine of the National Association for the Advancement of Colored People (NAACP) (Baker, 2022). Its editor, W.E.B. Du Bois, asked him to write a feature article for the second issue. Even to that audience, and despite reassuring words to the contrary, Boas still argued for the existence of differences in “mental characteristics” between African and European Americans:

The existing differences are differences in kind, not in value. This implies that the biological evidence also does not sustain the view, which is so often proposed, that the mental power of the one race is higher than that of the other, although their mental qualities show, presumably, differences analogous to the existing anatomical and physiological differences …

Thus it may safely be said that there is no anthropological evidence showing inferiority of the Negro race as compared with the white race, although we may assume that differences in mental characteristics of the two races exist. (Boas, 1910)

Five years later, the same doublespeak appeared in his preface to a book by Mary White Ovington, an NAACP cofounder:

Many students of anthropology recognize that no proof can be given of any material inferiority of the Negro race; that without doubt the bulk of the individuals composing the race are equal in mental aptitude to the bulk of our own people; that, although their hereditary aptitudes may lie in slightly different directions, it is very improbable that the majority of individuals composing the white race should possess greater ability than the Negro race. (Ovington, 1915, p. vii)

These new publishing opportunities may explain certain changes in wording between the 1911 and 1921 editions of The Mind of Primitive Man. Although Boas became more careful in expressing himself on the subject, his actual views show few signs of change. As late as 1928, he was still arguing for the existence of racial differences in mental capability, as seen in his work Anthropology and Modern Life:

Nevertheless the distribution of individuals and of family lines in the various races differs. When we select among the Europeans a group with large brains, their frequency will be relatively high, while among the Negroes the frequency of occurrence of the corresponding group will be low. If, for instance, there are 50 per cent of a European population who have a brain weight of more than, let us say, 1,500 grams, there may be only 20 per cent of Negroes of the same class. Therefore 30 per cent of the large-brained Europeans cannot be matched by any corresponding group of Negroes. (Boas, 1962[1928], p. 38)

To be sure, he also emphasized the existence of overlap between the two populations, but this caveat had the effect of driving home his initial point:

The brain in each race is very variable in size and the "overlapping" of individuals in the races is marked. It is not possible to identify an individual as a Negro or White according to the size and form of the brain, but serially the Negro brain is less extremely human than that of the White. (Boas, 19621928], pp. 39-40)

The differences between races are so small that they lie within the narrow range in the limits of which all forms may function equally well. We cannot say that the ratio of inadequate brains and nervous systems, that function noticeably worse than the norm, is the same in every race, nor that those of rare excellence are equally frequent. It is not improbable that such differences may exist in the same way as we find different ranges of adjustability in other organs. (Boas, 1962[1928], p. 41)

This was a recurring pattern in his line of argument. He would first assert the existence of a racial difference in mental capability, then qualify his assertion by pointing to the existence of overlap and, finally, qualify his qualification by noting that a small statistical difference can produce large differences among the very smart and the very dull.

Boas may have had plans during the mid-1920s for a comparative IQ study of African Americans and Euro Americans. In 1924, one of his PhD students, a young Margaret Mead, prepared an article on different methodologies for testing racial differences in IQ. It was written before her Samoan fieldwork and is the sort of research assistance that she, as a PhD novice, would have provided her supervisor before beginning her own research.

The article has a neutral tone. Although Mead argued that racial differences in IQ may be illusory (i.e., due to differences in education or language familiarity), she did not rule out their existence. In particular, she pointed to admixture studies as the best means to distinguish between true racial differences and false ones: “If the genealogical method could be subjected to extensive verification and supplemented by some technique for holding the other factors constant it might be productive of valuable results” (Mead, 1926, p. 661). By no means a blank-slatist, she deplored “the present state of ignorance concerning the laws regulating the inheritance of mental traits” (Mead, 1926, p. 667).

The final decade (1932-1942)

The first three decades of the 20th century had seen a slow shift toward the idea that all human races possess the same mental capabilities. In the early 1930s, this shift suddenly accelerated in response to events in the very land where Boas had been born. Alarmed, he and others threw themselves into the fight against “racism” — a word coming into use as a synonym for Nazism and other “blood and soil” ideologies (Frost, 2015). To that end, he purged any remaining racialist statements from the 1938 edition of The Mind of Primitive Man, perhaps in the belief that the anti-Nazi struggle required a united front against all forms of racial discrimination.

This chapter of his life, which ended with his death in 1942, is summarized in one of his obituaries:

Dr. Boas, who had studied and written widely in all fields of anthropology devoted most of his researches during the past few years to the study of the "race question," especially so after the rise of the Nazis in Germany. Discussing his efforts to disprove what he called "this Nordic nonsense," Prof. Boas said upon his retirement from teaching in 1936 that "with the present condition of the world, I consider the race question a most important one. I will try to clean up some of the nonsense that is being spread about race those days. I think the question is particularly important for this country, too; as here also people are going crazy." (JTA, 1942)

After his death, the mantle of Boasian anthropology fell to his two leading students: Ruth Benedict and Margaret Mead. Neither they nor many other academics wished to end the war on racism when the Second World War ended. This stance had at least three justifications:

• Many felt that nationalism had caused the war and that any discourse on human differences, no matter how scholarly, would make nationalism more intractable. A lasting peace could be built only on internationalism.

• Antisemitism was far from dead in 1945. There were fears of a resurgence, and such a resurgence could not be fought alone. As during the war, this fight would require a united front against any belief in human differences.

• The postwar era would be dominated by a Cold War rivalry for the hearts and minds of emerging nations in Africa, Asia and elsewhere. Discriminatory practices at home, notably in education and immigration, became a source of embarrassment to the U.S. in its foreign relations.

Boas had sought to strike a balance between nature and nurture in the study of human populations. Events intervened, however, and Boasian anthropology was conscripted to fight the war against racism. Eight decades later, that war is still being fought.

References

Baker, L.D. (2022). W.E.B. Du Bois and American anthropology. In: The Oxford Handbook of W.E.B. Du Bois, edited by A.D. Morris, M. Schwarrz, C. Johnson-Odim, W. Allen, M.A. Hunter, et al. https://doi.org/10.1093/oxfordhb/9780190062767.001.0001

Boas, F. (1894). Human faculty as determined by race. Address by Franz Boas before the Section of Anthropology, American Association for the Advancement of Science, at the Brooklyn meeting. https://upload.wikimedia.org/wikipedia/commons/3/37/Human_faculty_as_determined_by_race_-_address_%28IA_101313603.nlm.nih.gov%29.pdf

Boas, F. (1901). The mind of primitive man. The Journal of American Folk-Lore, 14, (52), 1-11. https://doi.org/10.1126/science.13.321.281

Boas, F. (1904). The Negro and the Demands of Modern Life: Ethnic and Anatomical Considerations. Charities, 15(1), 85-88.

Boas, F. (1906). Commencement address at Atlanta University, May 31. Atlanta University Leaflet, No. 19. http://www.webdubois.org/BoasAtlantaCommencement.html

Boas, F. (1909). Race problems in America. Science, 29(752), 839-849. https://doi.org/10.1126/science.29.752.839

Boas, F. (1910). The real race problem. The Crisis, 1(2), 22-25. https://modjourn.org/issue/bdr507810/

Boas, F. (1911). The Mind of Primitive Man. 1st edition, New York: Macmillan

Boas, F. (1921). The Mind of Primitive Man. 2nd edition, New York: Macmillan.

Boas, F. (1962[1928]). Anthropology and Modern Life. New York: Norton & Co.

Boas, F. (1938). The Mind of Primitive Man. 3rd edition, New York: Macmillan.

Boas, F. (1974). A Franz Boas Reader. The Shaping of American Anthropology, 1883-1911. G.W. Stocking Jr. (ed.), Chicago: The University of Chicago Press.

Degler, C. (1991). In Search of Human Nature: The Decline and Revival of Darwinism in American Social Thought. New York: Oxford University Press.

Frost, P. (2015). Birth of a word. Evo and Proud, May 23. https://evoandproud.blogspot.com/2015/05/birth-of-word.html

JTA (1942). Dr. Franz Boas, Debunker of Nazi Racial Theories, Dies in New York, December 23. http://www.jta.org/1942/12/23/archive/dr-franz-boas-debunker-of-nazi-racial-theories-dies-in-new-york

MacDonald, K. (1998). The Culture of Critique. An Evolutionary Analysis of Jewish Involvement in Twentieth-Century Intellectual and Political Movements. Praeger Publishers.

Mead, M. (1926). The methodology of racial testing: its significance for sociology. American Journal of Sociology, 31(5), 577-720. https://doi.org/10.1086/213952

Ovington, M.W. (1911). Half a Man. The Status of the Negro in New York. Foreword by Dr. Franz Boas of Columbia University. New York: Longmans, Green, and Co.

Wikipedia (2025). Franz Boas. http://en.wikipedia.org/wiki/Franz_Boas

Williams Jr., V.J. (1996). Rethinking Race: Franz Boas and His Contemporaries. University Press of Kentucky. https://books.google.ca/books?id=MKnIOfHNxXMC&printsec=frontcover&hl=fr#v=onepage&q&f=true

Cure juvenile delinquency in the slums. Federal Art Project, New York, 1936.

We may adopt a line of reasoning because it fits the current consensus, and not because it makes sense.

By framing the immigration issue in terms of "culture," Philippe Lemoine leads us to conclude that non-European immigrants are no real problem as long as they "assimilate" (Lemoine, 2025).

Is this really true? I understand why people fall for this line of reasoning — we all believe in the existence of culture and cultural differences. But there is much less consensus on genetics and, uh.... Look! A squirrel!

In reality, crime rates are higher among the more assimilated, non-practicing Muslims than among the less assimilated, practicing ones. This is the case in France (Yildiz, 2017). The two groups also show similar differences in non-criminal, but still undesirable aspects of behavior. This should be no surprise. Religion, like culture in general, serves the purpose of controlling innate desires and channeling them in the right direction.

So when non-European immigrants "assimilate" they abandon their cultural restraints and adopt those of Europe. Unfortunately, the latter are much weaker than the former. This isn't the Europe of the 1950s (let alone the one that existed before the Enlightenment). This is a Europe where cultural restraints have largely ceased to exist.

Non-European immigrants are thus leaving an environment where they are closely controlled by family, community, and religion and entering one where they have much more freedom to express their inner drives, desires, and impulses. This is the modern cultural environment — in France and throughout the West. Because the constraining influence of culture is relatively weak, differences in behavior, be they among individuals or between groups, are influenced much more by genetics.

I understand the reluctance to discuss genetics in this context. No need to sit me down and patiently explain to me the 20th century. I know all that. But I also know that the 21st century will not be a walk in the park if we ignore genetic factors and their role in human differences.

It is not simply because of "culture" that some people are far more willing to resolve personal differences through violence. It is not simply because of "culture" that some people are more focused on the present and seem unconcerned about the future. It is not simply because of "culture" that some people are better at thinking through a problem and solving it.

Well, let me restate the last paragraph. These things are wholly due to culture, but not as you might think. Humans adapt to their way of life, just as they adapt to their natural environment — more so, in fact. Culture creates a "template" for natural selection. It favors the survival and reproduction of those who can best fit into their culture and exploit its possibilities.

So, over time, a behavior that starts off being 100% cultural will become more and more genetic. This is the kind of evolutionary trajectory that humans have followed, and it has differed from one population to another.

If none of the above makes sense, well, I gave it my best try. Perhaps we are condemned to believe in the magic of "culture" and "assimilation." Perhaps we'll have to learn some things the hard way.

Note

La pratique de l’islam est perçue de manière quasi unanime comme un canalisateur de la délinquance. Nous avons pu observer une corrélation entre pratique religieuse et abandon de la délinquance sans, bien évidemment, établir de lien de causalité entre non-pratique religieuse et délinquance. Ce constat tend à confirmer l’hypothèse selon laquelle la religion permet une forme de réhabilitation pour certains jeunes délinquants.

À la lumière de notre recherche, nous pouvons souligner qu’il s’agit avant tout d’une canalisation de la petite délinquance. Même si elle n’est pas démontrée, cette vision de l’influence de la religion est partagée par la grande majorité des interviewés.

Nos travaux vont ainsi dans le sens des études démontrant un impact positif de la religiosité sur la délinquance (par exemple celles de Stark, 1997 ; Jensen et Erickson, 1979 ou Higgins et Albretch, 1977) : l’islam est perçu comme permettant d’inhiber les conduites déviantes. Dans nos entretiens, nous constatons que les musulmans délinquants sont les plus superficiellement religieux. (Yildiz, 2017)

References

Lemoine, P. (2025). Immigration and French Politics. Aporia Magazine, April 13.

Yildiz, T. (2017). Islam et lien social dans les quartiers défavorisés d’Île-de-France. Sociétés, 138(4), 73-84. https://doi.org/10.3917/soc.138.0073

Saturn devouring his son (1819-1823). Francisco de Goya (1746-1828) – Wikicommons (Soerfm)

Before the State imposed its monopoly on violence, a man would use violent means to defend himself, his family, and his kin from perceived threats. The resulting man-to-man conflict would create intense emotion in him and then “bloodlust” — the desire to see, feel, and taste his opponent’s blood. The more he satisfied that desire, the stronger it would become, and the more an initially defensive action would turn into a frenzy.

Once seen as normal, such “hemomania” became less acceptable as the State monopolized the use of violence. The new social environment discriminated against those individuals who responded positively to the sight, feel, and taste of blood — a response now deemed abnormal — while discriminating in favor of those who responded negatively. In this way, and in line with the degree of State pacification, hemomania was selected out of the population and hemophobia was selected in.

For more, please read:

Frost, P. (2025). Bloodlust: Altered or Ecstatic States of Consciousness. In: Shackelford, T.K. (ed.), Encyclopedia of Religious Psychology and Behavior, Springer. https://doi.org/10.1007/978-3-031-38971-9_2043-1

Centaur figure in a mosaic at Hadrian’s villa (Wikicommons)

Women are the “fair sex” because their skin has less melanin and less hemoglobin. They thus look pale in comparison to men, who are darker and ruddier. The cause is hormonal, as shown by studies of normal, castrated, and ovariectomized individuals, as well as by a digit ratio study (Edwards & Duntley, 1939; Edwards & Duntley, 1949; Edwards et al., 1941; Manning et al., 2004).

This sex difference once accounted for most of the variation in skin color within any human group. There is thus a strong cross-cultural tendency to associate darker skin with men and lighter skin with women (Frost, 2010; Frost, 2011; Frost, 2023; van den Berghe & Frost, 1986). This is also why many cultures have independently developed the artistic convention of using a lighter shade for women and a darker one for men (Capart 1905, pp. 26-27; Eaverly 2013; Pallottino 1952, pp. 34, 45, 73, 76-77, 87, 93, 95, 105, 107, 115; Soustelle 1970, p. 130; Tegner 1992; Wagatsuma 1967).

Even AI picks up on this sex difference (Cutler, 2025).

We still associate darker skin with men and lighter skin with women, if only subconsciously. Moreover, as shown by several controlled studies, we use this sex difference to tell male and female faces apart, especially if face shape is poorly visible. The main visual cue is the skin’s brownness and ruddiness. If the face is too far away or the lighting too dim, our mind switches to the more time-consuming but accurate cue of brightness. Facial color is thus perceived through a mechanism that served originally to tell men and women apart, even though ethnicity increasingly explains the differences we see among people we regularly meet (Dupuis-Roy et al. 2009; Dupuis-Roy et al., 2019; Jones et al., 2015; Nestor & Tarr, 2008a; Nestor & Tarr, 2008b; Tarr et al., 2001; Tarr et al., 2002; Yip & Sinha, 2002).

In sum, we are hardwired to see darker faces as male and lighter faces as female. This is seen in studies of people who are shown faces and asked to perform certain tasks:

• Make pictures of faces more attractive by varying their skin color. Male faces are made darker and ruddier than female ones (Carrito et al., 2016).

• Assess the skin color of faces. Female faces are judged to be lighter-skinned than male ones, even though most of the participants deny that skin color differs between men and women (Carrito & Semin, 2019).

• View several faces (of identical skin color). Then use your memory to identify each of them from a row of faces (which vary only in skin color). People respond by choosing lighter versions of the original female faces and darker versions of the original male faces, even though the original faces were all the same color (Carrito & Semin, 2019).

Even inanimate objects are unthinkingly gendered in the same way:

• Identify personal names by gender. Names are identified faster by gender when male names are presented in black and female names in white (Semin et al., 2018).

• Classify briefly appearing blobs by gender. Black blobs are classified predominantly as male and white blobs as female (Semin et al., 2018).

• Observe light and dark objects, while having your eye movements tracked. Observation is longer and fixation more frequent when a dark object is associated with a male character and a light object with a female character (Semin et al., 2018).

This hardwired mechanism has two complementary functions: evaluate the person you observe and modify your emotional state accordingly. The second function is best understood by understanding how women came to be lighter-skinned than men.

A lighter skin color is one of several infant traits that the adult female body seems to mimic, the others being a smaller nose and chin, a smoother and more pliable skin texture, and a higher pitch of voice. This is what ethologist Konrad Lorenz called the Kindchenschema — a set of visual, tactile, and auditory cues that identify an infant to an adult, who responds by feeling less aggressive and more willing to defend and nurture. In a word, the infant seems “cute” (Frost, 2010, pp. 134-135; Lorenz, 1971, pp. 154-164).

An infant’s pale skin is especially apparent in societies where adults are dark-skinned. A new Kenyan mother may tell her neighbors to come and see her mzungu, i.e., “European” (Walentowitz, 2008). When an anthropologist asked Zambian girls to describe how Africans look, some wrote: “At birth African children are born like Europeans, but after a few months the color changes to the color of an African” (Powdermaker, 1956).

The newborn’s light color is often attributed to a previous spirit life:

There is a rather widespread concept in Black Africa, according to which human beings, before “coming” into this world, dwell in heaven, where they are white. For, heaven itself is white and all the beings dwelling there are also white. Therefore the whiter a child is at birth, the more splendid it is. In other words, at that particular moment in a person's life, special importance is attached to the whiteness of his colour, which is endowed with exceptional qualities. (Zahan, 1974, p. 385)

As long as this light coloration persists, a newborn African is considered soft and vulnerable, and hence at risk of returning to heaven.

… it is also claimed that a newborn baby is not only white but also a soft being during the time between his birth and his acceptance into the society. (Zahan, 1974, p. 386-387)

This infant coloration is also apparent in many nonhuman primates. Langurs, baboons, and macaques are born pink and later darken, becoming almost black in adulthood. Their fur, too, is lighter at birth and darkens with age. The light coloration seems to make nearby adults more caring and protective. As the infant darkens with age, its mother no longer seeks it out to hold and fondle (Alley, 1980; Alley, 2014; Booth, 1962; Guthrie, 1970; Jay, 1962).

This change in emotional state has been studied in humans by a Chinese research team. The participants were shown a series of attractive or unattractive Chinese faces that varied in skin color and seemed to be either in pain (with a syringe needle sticking into a cheek) or not in pain (with a Q-tip touching a cheek). For each face, the participant had to judge whether it was experiencing pain. Reaction time was measured, as was brain activity on an electroencephalogram (EEG).

When the faces were attractive and seemingly in pain, the lighter ones were judged to be suffering more than the darker ones. Also, reaction time was shorter, and the EEGs showed a higher level of empathy. But this difference was absent for unattractive faces. Only when the face was both light-skinned and attractive did the observer feel more empathy if the face seemed to be in pain (Yang et al., 2022).

Why would an observer feel more empathy for a light-skinned person in pain? This may have originated as a response to seeing an infant or a woman in danger. In both cases, light skin is a visual cue for the Kindchenschema. Darker and ruddier skin would conversely evoke the image of an adult male, and make the observer less empathetic.

But why would light skin have this effect only if the face is attractive? The empathic response is probably stimulated by the entire Kindchenschema — not only lighter skin, but also bigger eyes and a less protruding chin and jaw. These visual cues probably interact with each other: a “cute” facial feature would increase the empathic effect of other “cute” features, whereas an “ugly” feature would have the reverse effect.

The lighter faces were judged to be in greater pain. Reaction time was also shorter, and the EEGs indicated more empathy (Yang et al., 2022)

This relationship between skin color and the observer’s emotional state is consistent with previous research findings. Let’s review the literature to understand how the different skin tones of men and women affect an observer’s emotional state.

General perceptions of male redness

We will first consider the redness of a male body. How does it affect the observer? This question has sparked much research since a high-profile finding two decades ago.

At the 2004 Olympic Games, athletes won more often with red uniforms than with blue ones. In man-to-man contests of boxing, taekwondo, Greco-Roman wrestling, and freestyle wrestling, contestants with red uniforms won 16 out of 21 rounds. Contestants with blue uniforms won only 4 rounds. The red uniforms seem to have been more intimidating, perhaps because dominant males have more testosterone and, hence, a ruddier color (Hill & Barton, 2005). This finding has been corroborated by two other studies, which likewise conclude that red clothing makes a man seem more dominant and threatening (Feltman & Elliot, 2011; Wiedemann et al., 2015).

Even redness by itself seems to make an observer more cautious, more compliant, more aggressive, or more submissive. This has been shown by studies of people who are asked to perform certain tasks with differently colored objects or surroundings:

• Identify the color of words on a computer screen. Men take longer than women to answer when the words are in red (Ioan et al., 2007).

• Identify the incorrect idiom from a list of idioms. The error rate is higher when the idioms are in red than when they are in blue (Shi et al., 2015).

• Exercise on a stationary bicycle in a red environment. Covered distance and heart rate are lower in a red environment than in a green or gray one, while enjoyment is higher in a green environment than in a red one (Briki et al., 2015).

• Make bids at an auction and make price offers during negotiations. A red background is associated with higher jumps in bids at an auction and lower offers during negotiations (Bagchi & Cheema, 2012).

• Respond to the message in an advertisement. Compliance is higher when a message is shown against a red background (Kareklas et al., 2019).

The redness effect varies with context. It is least apparent when redness is presented alone or overwhelmed by a much stronger effect; for example, English soccer teams have the same home advantage with or without red shirts (Allen & Jones, 2014). It becomes more apparent when redness is associated with a facial photo, rather than a simplified drawing of a face, and even more apparent when the face is attractive (Buechner et al., 2014; Pazda et al., 2023; Schwarz & Singer, 2013; Wen et al., 2014; Young, 2015).

Redness also affects the perceived passage of time. When a red screen and a blue screen are seen for the same duration, the red screen is perceived as lasting longer. The overestimation is greater for men than for women. People also react faster to the red screen than to the blue screen, and the difference in reaction time correlates with the tendency to overestimate red-screen duration. It seems that “the arousal induced by red increases the speed of the internal clock” (Shibasaki & Masataka, 2014).

Finally, redness can help distinguish between facially similar emotions, such as anger versus disgust, surprise versus fear, and sadness versus happiness (Thorstenson et al., 2019a; Thorstenson et al., 2021). When faces are viewed against a red background, people take less time to identify a face as angry than they do to identify a face as happy or fearful (Young et al., 2013).

Lack of redness, as when a face turns pale with fright, has the reverse effect: it appeases the observer and de-escalates conflict. When people are asked to rate their propensity for turning pale, the highest scores come from those who fear blood and injury. Women score higher on average than men (Drummond, 1997). 

What about a face turning red with embarrassment? Like anger, blushing serves to keep the observer at a distance, being a sort of controlled anger due to embarrassment (Drummond, 1997). If a person blushes while making an apology, the observer will more likely perceive it as sincere and be more forgiving (Thorstenson et al., 2019b). Blushing seems to be heritable. Darwin cited the case of a father, mother, and ten children, “all of whom, without exception, were prone to blush to a most painful degree” (Darwin, 1872, p. 312).

Left: Olympic contests won by color of uniform (red or blue) for all sports, for boxing, for taekwondo, for Greco-Roman wrestling, and for freestyle wrestling. Right: Proportion of contests won by color of uniform (red or blue) and by difference in skill and strength (degree of asymmetry). The color difference seems to matter most for contestants of equal skill and strength (Hill & Barton, 2005)

General perceptions of male darkness

We will now consider the darkness of a male body. The relevant literature is mostly about color symbolism in different cultures.

The ancient Greeks believed that a woman’s light color incarnated her “helplessness and need of protection” and a man’s dark color his “courage and the ability to fight well” (Irwin 1974, p. 121). A brave man had a “black rump”, and a coward a “white rump.” The same mental association was projected onto the internal organs and ultimately onto the soul. A “black heart” denoted strong emotions, and a “white heart” indifference or a refusal to act. A coward had a “white liver,” a metaphor that survives in the expression “lily-livered” (Irwin 1974, pp. 129-155).

Cross-culturally, whiteness evokes weakness and blackness strength (Gergen 1967, p. 397; Osgood 1960, p. 165). According to a Turkish study, blackness is associated with power, fear, courage, and eternity, and redness with courage, enthusiasm, fun, and anxiety. In contrast, whiteness is associated with cleanliness, honesty, peace, and hope (Demir, 2020). A study with participants from China, Germany, Greece, and the UK found that “BLACK and RED evoked more consistent colour–emotion association patterns than other colour terms. BLACK and RED colour terms were also more commonly associated with strong emotions, and evoked a larger number of associated emotions than many other colour terms” (Jonauskaite et al., 2019).

There is a similar pattern in a study of rural French Canadians. Darker men were perceived as being stronger and more virile in physique and character. If a man was too dark, he would be considered quick-tempered, arrogant, and malicious. Women were expected to be fairer-skinned, although their lighter color was likewise associated with weakness of physique and character (Frost, 2010, pp.149-152).

Female perceptions of male redness/darkness

We will now turn to gender-specific perceptions. How do women respond to the redness of a male body? How does it affect their emotional state?

According to one study, women associate high levels of male ruddiness with aggression, medium levels with dominance, and low levels with attractiveness (Stephen et al., 2012). According to another study, women judge men to be more attractive when viewed against a red background and in red clothing. This effect is confined to sexual attractiveness and does not change the man’s overall likability (Elliot et al., 2010). These findings seem contradictory: male attractiveness is associated with a low level of redness in one study and a high level in the other. Perhaps the word “attractiveness” was construed aesthetically by the women of the first study and erotically by those of the second.

How do women respond to the darkness of a male body? Their response seems to be hormonally influenced, as shown by two studies:

• Choose between two male faces that differ slightly in color. The darker face is more often preferred during the first two-thirds of the menstrual cycle than during the last third, although the lighter face is always the majority preference. During the first two-thirds of the cycle, the level of estrogen is high in relation to the level of progesterone (which acts as an anti-estrogen). During the last third, the ratio is reversed, with estrogen being low relative to progesterone. There is no cyclical change in preference when participants view female faces or have taken oral contraceptives (Frost, 1994).

• View male faces during a brain MRI. Women have a stronger neural response to masculinized male faces than to feminized ones. The intensity of the response correlates with the rise and fall of estrogen levels across the menstrual cycle. In a personal communication, the lead author stated that the faces were masculinized by making them darker and more robust in shape (Rupp et al., 2009).

An estrogenic effect is further suggested by a study of preschool children:

• Choose between two dolls that differ slightly in skin color. Among boys and girls below three years of age, body fat is greater in those choosing the darker doll than in those choosing the lighter doll. In that age range, estrogen is produced mostly in fatty tissues (Frost, 1989).

To some extent, the above preferences can be triggered by red or dark clothing, as shown by a study of preferences for men or women dressed in blue, green, yellow, red, white, or black. Women prefer men dressed in red or black. The same colors are preferred when men are asked to judge women or men. But neither red nor black is preferred when women are asked to judge other women (Roberts et al., 2010).

Male perceptions of female redness/darkness

We will now look at things from a more paradoxical perspective — a woman’s body with male pigmentation. How do men respond to a reddened and darkened female body? How does it affect their emotional state?

Let’s first consider redness. Women with red clothes seem more attractive to men (Elliot & Niesta, 2008), but this effect is conditional. A woman with red clothes seems more attractive if she is already attractive, but not if she is masculine-looking, older, or unattractive (Pazda et al., 2023; Schwarz & Singer, 2013; Wen et al., 2014; Young, 2015). This effect is confined to sexual attractiveness and does not affect overall likability (Wen et al., 2014). Redness seems to intensify a man’s interest in a woman; if there is nothing of interest to begin with, there is nothing to intensify.

This finding is consistent with popular culture. The “lady in red” is a femme fatale:

In literature, red has repeatedly been associated with female sexuality, especially illicit sexuality, most famously in Nathaniel Hawthorne’s classic work The Scarlet Letter. Likewise, in popular stage and film, there are many instances in which red clothing, especially a red dress, has been used to represent passion or sexuality …

Red is paired with hearts on Valentine’s Day to symbolize romantic affection and is a highly popular color for women’s lingerie. Red has been used for centuries to signal sexual availability or “open for business” in red-light districts. Women commonly use red lipstick and rouge to heighten their attractiveness, a practice that has been in place at least since the time of the ancient Egyptians (Elliot & Niesta, 2008)

Let’s now consider how men respond to a darkened female body. When men are shown pictures of lighter-skinned and darker-skinned women, the two groups are judged to be equally attractive, but the men’s eye movements tell another story: the lighter-skinned women are viewed for a longer time than the darker-skinned ones (Garza et al., 2016). Given that the two groups of women arouse the same degree of sexual interest, the male response to darker women must be briefer and more intense. In other words, the same level of interest is concentrated within a shorter time.

This finding is consistent with popular culture. In English novels from the Victorian era, “the dark lady” is an “impetuous,” “ardent,” and “passionate” player in short-lived romances (Carpenter, 1936, p. 254). In French and German novels from that era, “the love incarnated by les brunes appears as the conceptual equivalent of a devouring femininity, thus making them similar to the mythical figure of Lilith.” (Atzenhoffer, 2011, p. 6). For one French writer, men who love les brunes “are generally taken by a passion that is more violent, more demonstrative, and more domineering, but also less profound, less tender, and less durable” (Briot, 2007).

But why, then, did women evolve lighter skin?

In sum, a dark, reddish color excites the observer much more than a lighter one. The effect is stronger if a dark or reddish color is associated with a human body and, even more so, with an attractive human face. Furthermore, it is as strong for a man observing a woman as it is for a woman observing a man. This excitation can increase sexual arousal, but at the cost of a steeper decline afterwards — perhaps because a higher level of arousal is exhausted more rapidly.

If a dark, reddish color increases sexual arousal not only when a woman is observing a man but also when a man is observing a woman, why, then, did women become less dark and less ruddy than men? Wouldn’t sexual selection have favored darker, ruddier women? For that matter, why is there any sex difference in skin color?

Perhaps the need to survive and procreate favored those women who could not only arouse sexual interest but also maintain it over the longer term. That goal would not be met if sexual interest were concentrated within a briefer but more intense interval. Thus, lighter female skin was favored not because it intensified male sexual interest but rather because it stabilized the pair bond through a decrease in male aggressiveness and an increase in male provisioning — by mimicking the Kindchenschema.

This is sexual selection in a broad sense: the pressure of selection extends beyond the moment of mate choice to the much longer period of cohabitation after mating.

References

Allen, M. S., & Jones, M. V. (2014). The home advantage over the first 20 seasons of the English Premier League: Effects of shirt colour, team ability and time trends. International Journal of Sport and Exercise Psychology, 12(1), 10-18. https://doi.org/10.1080/1612197X.2012.756230

Alley, T. R. (1980). Infantile colouration as an elicitor of caretaking behaviour in Old World primates. Primates, 21(3), 416-429. https://doi.org/10.1007/BF02390470  

Alley, T.R. (2014[1986]). An ecological analysis of the protection of primate infants. In V. McCabe & G.J. Balzano (Eds.) Event Cognition: An Ecological Perspective. Routledge.

Atzenhoffer, R. (2011). Les hommes préfèrent les blondes. Les lectrices aussi. Effet de psychologie, horizons idéologiques et valeurs morales des héroïnes dans l'œuvre romanesque de H. Courths-Mahler. Colloque national (CNRIUT), Villeneuve d'Ascq, 8-10 juin 2009, https://web.archive.org/web/20210921193948/http://cnriut09.univ-lille1.fr/articles/Articles/Fulltext/10a.pdf

Bagchi, R., & Cheema, A. (2013). The effect of red background color on willingness-to-pay: the moderating role of selling mechanism. Journal of Consumer Research, 39(5), 947-960. https://doi.org/10.1086/666466

Booth, C. (1962). Some observations on behavior of Cercopithecus monkeys. Annals of the New York Academy of Sciences, 102(2), 477-487. https://doi.org/10.1111/j.1749-6632.1962.tb13654.x   

Briki, W., Rinaldi, K., Riera, F., Trong, T. T., & Hue, O. (2015). Perceiving red decreases motor performance over time: A pilot study. European Review of Applied Psychology, 65(6), 301-305. https://doi.org/10.1016/j.erap.2015.09.001

Briot, E. (2007). Couleurs de peau, odeurs de peau : le parfum de la femme et ses typologies au xixe siècle. Corps, 2(3), 57-63. https://doi.org/10.3917/corp.003.0057

Buechner, V. L., Maier, M. A., Lichtenfeld, S., & Schwarz, S. (2014). Red-take a closer look. PloS one, 9(9), e108111. https://doi.org/10.1371/journal.pone.0108111

Capart, J. (1905). Primitive Art in Egypt. London: H. Grevel.

Carpenter, F. I. (1936). Puritans preferred blondes. The heroines of Melville and Hawthorne. New England Quarterly, 9(2), 253-272. https://doi.org/10.2307/360391

Carrito, M.L., dos Santos, I.M.B., Lefevre, C.E., Whitehead, R.D., da Silva, C.F., & Perrett, D.I. (2016). The role of sexually dimorphic skin colour and shape in attractiveness of male faces. Evolution and Human Behavior, 37(2), 125-133. https://doi.org/10.1016/j.evolhumbehav.2015.09.006

Carrito, M. L., & Semin, G. R. (2019). When we don’t know what we know–Sex and skin color. Cognition, 191, 103972. https://doi.org/10.1016/j.cognition.2019.05.009

Cutler, A. (2025). Note.

Darwin, C. (1872). The Expression of the Emotions in Man and Animals. London: John Murray.

Demir, Ü. (2020). Investigation of color-emotion associations of the university students. Color Research & Application, 45(5), 871-884. https://doi.org/10.1002/col.22522

Drummond, P.D. (1997). Correlates of facial flushing and pallor in anger - provoking situations. Personality and Individual Differences, 23(4), 575-582. https://doi.org/10.1016/S0191-8869(97)00077-9  

Dupuis-Roy, N., Faghel-Soubeyrand, S., & Gosselin, F. (2019). Time course of the use of chromatic and achromatic facial information for sex categorization. Vision Research, 157, 36-43. https://doi.org/10.1016/j.visres.2018.08.004  

Dupuis-Roy, N., Fortin, I., Fiset, D., & Gosselin, F. (2009). Uncovering gender discrimination cues in a realistic setting. Journal of Vision, 9(2), 10, 1-8. https://doi.org/10.1167/9.2.10  

Eaverly, M.A. (2013). Tan Men/Pale Women. Color and Gender in Archaic Greece and Egypt, a Comparative Approach. Ann Arbor: The University of Michigan Press. https://doi.org/10.3998/mpub.3080238

Edwards, E.A., & Duntley, S.Q. (1939). The pigments and color of living human skin. American Journal of Anatomy, 65(1), 1-33. https://doi.org/10.1002/aja.1000650102

Edwards, E.A., & Duntley, S.Q. (1949). Cutaneous vascular changes in women in reference to the menstrual cycle and ovariectomy. American Journal of Obstetrics & Gynecology, 57(3), 501-509. https://doi.org/10.1016/0002-9378(49)90235-5

Edwards, E.A., Hamilton, J.B., Duntley, S.Q., & G. Hubert, G. (1941). Cutaneous vascular and pigmentary changes in castrate and eunuchoid men. Endocrinology, 28(1), 119-128. https://doi.org/10.1210/endo-28-1-119

Elliot, A. J., & Niesta, D. (2008). Romantic red: red enhances men's attraction to women. Journal of Personality and Social Psychology, 95(5), 1150. https://psycnet.apa.org/doi/10.1037/0022-3514.95.5.1150

Elliot, A. J., Niesta Kayser, D., Greitemeyer, T., Lichtenfeld, S., Gramzow, R. H., Maier, M. A., & Liu, H. (2010). Red, rank, and romance in women viewing men. Journal of Experimental Psychology: General, 139(3), 399–417. https://doi.org/10.1037/a0019689

Feltman, R., & Elliot, A. J. (2011). The influence of red on perceptions of relative dominance and threat in a competitive context. Journal of Sport and Exercise Psychology, 33(2), 308-314. https://doi.org/10.1123/jsep.33.2.308

Frost, P. (1989). Human skin color: the sexual differentiation of its social perception. Mankind Quarterly, 30, 3-16. https://doi.org/10.46469/mq.1989.30.1.1

Frost, P. (1994). Preference for darker faces in photographs at different phases of the menstrual cycle: Preliminary assessment of evidence for a hormonal relationship. Perceptual and Motor Skills, 79(1), 507-14. https://doi.org/10.2466/pms.1994.79.1.507

Frost, P. (2010). Femmes claires, hommes foncés. Les racines oubliées du colorisme. Quebec City: Les Presses de l'Université Laval, 202 p. https://www.pulaval.com/livres/femmes-claires-hommes-fonces-les-racines-oubliees-du-colorisme  

Frost, P. (2011). Hue and luminosity of human skin: a visual cue for gender recognition and other mental tasks. Human Ethology Bulletin, 26(2), 25-34. https://www.researchgate.net/publication/256296588_Hue_and_luminosity_of_human_skin_a_visual_cue_for_gender_recognition_and_other_mental_tasks

Frost, P. (2023). The original meaning of skin color. Peter Frost’s Newsletter, March 20. https://www.anthro1.net/p/the-original-meaning-of-skin-color  

Garza, R., Heredia, R.R., & Cieslicka, A.B. (2016). Male and female perception of physical attractiveness. An eye movement study. Evolutionary Psychology, 14(1), 1-16. https://doi.org/10.1177%2F1474704916631614

Gergen, K.J. (1967). The significance of skin color in human relations. Daedalus, 96, 390-406. http://www.jstor.org/stable/20027044  

Guthrie, R.D. (1970). Evolution of human threat display organs. In: Dobzhansky, T., Hecht, M.K., & Steere, W.C. (Eds.) Evolutionary Biology, 4, 257-302. New York: Appleton-Century Crofts.

Hill, R., & Barton, R. (2005). Red enhances human performance in contests. Nature, 435, 293. https://doi.org/10.1038/435293a  

Ioan, S., Sandulache, M., Avramescu, S., Ilie, A., & Neacsu, A. (2007). Red is a distractor for men in competition. Evolution and Human Behavior, 28, 285-293. https://doi.org/10.1016/j.evolhumbehav.2007.03.001  

Irwin, E. (1974). Colour Terms in Greek Poetry. Toronto: Hakkert.

Jay, P.C. (1962). Aspects of maternal behavior among langurs. Annals of the New York Academy of Sciences, 102(2), 468-476. https://doi.org/10.1111/j.1749-6632.1962.tb13653.x  

Jonauskaite, D., Wicker, J., Mohr, C., Dael, N., Havelka, J., Papadatou-Pastou, M., ... & Oberfeld, D. (2019). A machine learning approach to quantify the specificity of colour–emotion associations and their cultural differences. Royal Society Open Science, 6(9), 190741. https://doi.org/10.1098/rsos.190741

Jones, A.L., Russell, R., & Ward, R. (2015). Cosmetics alter biologically-based factors of beauty: evidence from facial contrast. Evolutionary Psychology, 13(1) https://doi.org/10.1177%2F147470491501300113  

Kareklas, I., Muehling, D. D., & King, S. (2019). The effect of color and self-view priming in persuasive communications. Journal of Business Research, 98, 33-49. https://doi.org/10.1016/j.jbusres.2019.01.022

Lorenz, K. (1971). Studies in Animal and Human Behaviour, vol. 2. London: Methuen & Co.

Manning, J.T., Bundred, P.E., & Mather, F.M. (2004). Second to fourth digit ratio, sexual selection, and skin colour. Evolution and Human Behavior, 25(1), 38-50. https://doi.org/10.1016/s1090-5138(03)00082-5

Nestor, A., & Tarr, M.J. (2008a). The segmental structure of faces and its use in gender recognition. Journal of Vision, 8(7), 7, 1-12. https://doi.org/10.1167/8.7.7  

Nestor, A., & Tarr, M.J. (2008b). Gender recognition of human faces using color. Psychological Science, 19(12), 1242-1246. https://doi.org/10.1111/j.1467-9280.2008.02232.x  

Osgood, C.E. (1960). The cross-cultural generality of visual-verbal synesthetic tendencies. Behavioral Science, 5(2), 146-169. https://doi.org/10.1002/bs.3830050204  

Pallottino, M. (1952). Etruscan Painting. Lausanne: Skira.

Pazda, A. D., & Elliot, A. J. (2017). Processing the word red can enhance women’s perceptions of men’s attractiveness. Current Psychology, 36, 316-323. https://doi.org/10.1007/s12144-016-9420-8

Pazda, A. D., Thorstenson, C. A., & Elliot, A. J. (2023). The effect of red on attractiveness for highly attractive women. Current Psychology, 42(10), 8066-8073. https://psycnet.apa.org/doi/10.1007/s12144-021-02045-3

Powdermaker, H. (1956). Social change through imagery and values of teen-age Africans in Northern Rhodesia. American Anthropologist, 58(5), 783-813 https://doi.org/10.1525/aa.1956.58.5.02a00030

Roberts, S. C., Owen, R. C., & Havlicek, J. (2010). Distinguishing between perceiver and wearer effects in clothing color-associated attributions. Evolutionary Psychology, 8, 350–364. https://doi.org/10.1177/147470491000800304

Rupp, H.A., James, T.W., Ketterson, E.D., Sengelaub, D.R., Janssen, E., & Heiman, J.R. (2009). Neural activation in women in response to masculinized male faces: mediation by hormones and psychosexual factors. Evolution and Human Behavior, 30(1), 1-10. https://doi.org/10.1016/j.evolhumbehav.2008.08.006

Schwarz, S., & Singer, M. (2013). Romantic red revisited: Red enhances men’s attraction to young, but not menopausal women. Journal of Experimental Social Psychology, 49, 161–164. https://doi.org/10.1016/j.jesp.2012.08.004

Semin, G.R., Palma, T., Acartürk, C., & Dziuba, A. (2018). Gender is not simply a matter of black and white, or is it? Philosophical Transactions of the Royal Society B Biological Sciences, 373(1752), 20170126. https://doi.org/10.1098/rstb.2017.0126   

Shi, J., Zhang, C., & Jiang, F. (2015). Does red undermine individuals' intellectual performance? A test in China. International Journal of Psychology, 50(1), 81-84. https://doi.org/10.1002/ijop.12076

Shibasaki, M., & Masataka, N. (2014). The color red distorts time perception for men, but not for women. Scientific Reports, 4, 5899. https://doi.org/10.1038/srep05899

Soustelle, J. (1970). The Daily Life of the Aztecs. Stanford, California: Stanford University Press.

Stephen, I.D., Oldham, F.H., Perrett, D.I., & Barton, R.A. (2012). Redness enhances perceived aggression, dominance and attractiveness in men's faces. Evolutionary Psychology, 10(3) https://doi.org/10.1177%2F147470491201000312

Tarr, M.J., Kersten, D., Cheng, Y., & Rossion, B. (2001). It's Pat! Sexing faces using only red and green. Journal of Vision, 1(3), 337, 337a. https://doi.org/10.1167/1.3.337  

Tarr, M. J., Rossion, B., & Doerschner, K. (2002). Men are from Mars, women are from Venus: Behavioral and neural correlates of face sexing using color. Journal of Vision, 2(7), 598, 598a. https://doi.org/10.1167/2.7.598  

Tegner, E. (1992). Sex differences in skin pigmentation illustrated in art. The American Journal of Dermatopathology, 14(3), 283-287. https://doi.org/10.1097/00000372-199206000-00016

Thorstenson, C.A. (2018). The social psychophysics of human face color: Review and recommendations. Social Cognition, 36(2), 247-273. https://doi.org/10.1521/soco.2018.36.2.247  

Thorstenson, C.A., Pazda, A.D., Young, S.G., & Elliot, A.J. (2019a). Face color facilitates the disambiguation of confusing emotion expressions: Toward a social functional account of face color in emotion communication. Emotion, 19(5), 799-807. https://doi.org/10.1037/emo0000485  

Thorstenson, C.A., Pazda, A., & Lichtenfeld, S. (2019b). Facial blushing influences perceived embarrassment and related social functional evaluations. Cognition and Emotion June 23, 1-14. https://doi.org/10.1080/02699931.2019.1634004   

Thorstenson, C.A., McPhetres, J., Pazda, A.D., & Young, S.G. (2021). The role of facial coloration in emotion disambiguation. Emotion, 22(7), 1604-1613. https://doi.org/10.1037/emo0000900  

van den Berghe, P.L., & Frost, P. (1986). Skin color preference, sexual dimorphism and sexual selection: A case of gene-culture co-evolution? Ethnic and Racial Studies, 9(1), 87-113. https://doi.org/10.1080/01419870.1986.9993516  

Wagatsuma, H. (1967). The social perception of skin color in Japan. Daedalus, 96(2), 407-443. http://www.jstor.org/stable/20027045

Walentowitz, S. (2008). Des êtres à peaufiner. Variations de la coloration et de la pigmentation du nouveau-né. In: J-P. Albert, B. Andrieu, P. Blanchard, G. Boëtsch, & D. Chevé (eds.) Coloris Corpus, (pp. 113-120). Paris: CNRS Éditions.

Wen, F., Zuo, B., Wu, Y., Sun, S., & Liu, K. (2014). Red is romantic, but only for feminine female faces: Sexual dimorphism moderates red effect on sexual attraction. Evolutionary Psychology, 12, 719–735. https://doi.org/10.1177/147470491401200404

Wiedemann, D., Burt, D. M., Hill, R. A., & Barton, R. A. (2015). Red clothing increases perceived dominance, aggression and anger. Biology Letters, 11(5), 20150166. https://doi.org/10.1098/rsbl.2015.0166

Yang, D., Li, X., Zhang, Y., Li, Z., & Meng, J. (2022). Skin color and attractiveness modulate empathy for pain: an event-related potential study. Frontiers in Psychology, 12, 780633. https://doi.org/10.3389/fpsyg.2021.780633

Yip, A.W., & Sinha, P. (2002). Contribution of color to face recognition. Perception, 31(8), 995-1003. https://doi.org/10.1068/p3376  

Young, S.G. (2015). The effect of red on male perceptions of female attractiveness: Moderation by baseline attractiveness of female faces. European Journal of Social Psychology, 45, 146–151. https://doi.org/10.1002/ejsp.2098

Young, S.G., Elliot, A.J., Feltman, R., & Ambady, N. (2013). Red enhances the processing of facial expressions of anger. Emotion, 13(3), 380–384. https://doi.org/10.1037/a0032471

Zahan, D. (1974). White, Red and Black: Colour Symbolism in Black Africa. In: A. Portmann & R. Ritsema (Eds.) The Realms of Colour, Eranos, 41(1972), 365-395, Leiden: Eranos. https://doi.org/10.2307/1573594

Kong Qiu (Confucius), c. 551- c. 479 BCE (Wikicommons)

Beginning some 10,000 years ago, hunting and gathering gave way to farming. This in turn gave way to ever more cultural change: population growth, year-round living in villages and towns, trade and specialization of labor, numeracy and literacy, the formation of states, the development of religion, the standardization of law and so on.

Humans were thus entering an ever-wider range of cultural environments, which, to varying degrees, placed new demands on the capacity to process information, recognize patterns and solve problems — in other words, cognitive ability.

The new demands varied from one region to another and from one time period to another, with the result that cognitive ability followed different trajectories of evolution in different populations. We can now chart this evolution by looking at DNA from human remains, specifically the various alleles associated with educational attainment (EA) and how their population frequencies have changed over time.

Several recent studies have charted this evolution in Europeans:

• No change during the long period of hunting and gathering, although this finding is uncertain because of small sample sizes.

• A steep rise between 9,000 and 7,000 years ago with the emergence of farming, in line with the cognitive demands of farming and also admixture from northern hunter-gatherers.

• A slower rise during the Neolithic and into recorded history, in line with the cognitive demands of increasing social complexity. With the stratification of society into classes, cognitive evolution became driven primarily by the higher fertility of upper-class families.

• A decline during the Imperial Era of Rome, in line with a decline in family formation and fertility among the upper classes.

• A renewed upward trend beginning in Late Antiquity, in line with the rise of Christianity and an increase in family formation among the upper classes.

• An especially rapid rise in Western Europe during late medieval and post-medieval times, in line with the demographic expansion of the middle class through higher fertility and lower mortality (Akbari et al., 2024; Frost, 2024; Kuijpers et al., 2022; Piffer et al., 2023; Piffer & Kirkegaard, 2024).

The other end of Eurasia

Europe is one of two large regions where mean cognitive ability has risen the most. The other is East Asia. How has cognitive ability evolved at the other end of Eurasia?

This question has been addressed by Davide Piffer in his recent study of 1,245 ancient genomes from eastern Eurasia over the last 12,000 years. For earlier periods, we have too few samples to draw worthwhile conclusions, as is also true for Europe.

He found an increase in mean cognitive ability over time. It was largest in southern China, followed by Mongolia and Southeast Asia. The size of the increase in southern China probably reflects not only cognitive evolution in situ but also the replacement of hunter-gatherers and simple farming societies by Han immigrants from northern China (Yang et al., 2020). Similarly, the increase in Mongolia may partly reflect gene flow from Han Chinese (10-51% of current Mongolian ancestry) and Europeans (6-40%) (Zhao et al., 2020). Southeast Asia has likewise seen extensive replacement of hunter-gatherers by Austronesian or Sino-Tibetan farmers (Lipson et al., 2018).

Mean cognitive ability did not rise as much elsewhere in eastern Eurasia. These smaller increases may reflect persistence of low social complexity (Arctic, North Siberia), failure to develop beyond a certain level (Central Asia and Tibet), or a complicated trajectory of positive and negative evolution (northeast China).

Increase in mean cognitive ability, as measured by alleles associated with educational attainment (EA), for each region of eastern Eurasia (Piffer, 2025a, Figure 4, p. 7)

A closer look

By pooling these regional results, Davide Piffer was able to track the cognitive evolution of eastern Eurasia in greater detail. His second analysis shows stagnation during the hunter-gatherer period, followed by a sustained upward trend beginning around 9,000 to 8,000 years ago, apparently with the advent of farming. This upward trend ended a little less than 1,500 years ago, near the beginning of the Tang Dynasty (618-907 CE).

The last finding may seem surprising because the Tang Dynasty saw great advances in art, literature and technology. Did this golden age have a darker side? Did the upper classes suffer a decline in fertility as they did in Imperial Rome?

This explanation has some support from historical sources. During the Tang Dynasty, the poet Wang Fanzhi affirmed that having one son was enough. The subsequent Song Dynasty (960–1279 CE) was a time when parents wanted to have only two sons or just one, according to a number of scholars. (Zhao, 2002, p. 756). Abortive drugs were sold during the Tang Dynasty and became even more available during the Song and Yuan (1206–1368 CE) dynasties. Such drugs were produced on a large scale in certain areas during the reign of Kang Xi (1662–1722 CE) and were widely used by the population — from members of the imperial family to ordinary citizens (Zhao, 2006, p. 17). When describing certain places in one of his poems, Wang Fan Zhi (c. 590–c. 660 CE) noted that the rich had few children while the poor had many (Zhao, 2006, p. 22).

Increase in mean cognitive ability, as measured by alleles associated with educational attainment (EA), over the last 12,000 years in eastern Eurasia. (Piffer, 2025, Figure 3, p. 6)

To confirm the stagnation of the Tang Dynasty and onwards, as shown by alleles associated with EA, Davide Piffer repeated his analysis but now used alleles associated with IQ. This method has the advantage of homing in on cognitive ability itself, rather than other traits that may assist success at school. But it also has the disadvantage of being based on much fewer alleles (only 434 independent SNPs, as compared to almost 4,000 for EA).

These results differ from the previous ones in two ways:

• Instead of remaining unchanged during the hunter-gatherer period, mean cognitive ability fell and continued to fall into the farming period, bottoming out only 6,000 years ago. The decline seems to mirror a decline in brain size that began in Eurasian populations after the last ice age, perhaps because their brains no longer had to store, process, and manipulate huge amounts of spatiotemporal data for hunting over large expanses of territory (Frost, 2019; Hawks, 2011). This loss of spatiotemporal ability may therefore show up more in IQ than in EA.

• Instead of remaining unchanged during the last 1,500 years, mean cognitive ability fell somewhat. This decline may show up more in IQ because EA captures not only cognitive ability but also certain propensities — rule following, submissiveness, resistance to boredom — that are less related to cognitive ability and more to classroom performance.

Increase in mean cognitive ability, as measured by alleles associated with IQ, over the last 12,000 years in eastern Eurasia (Piffer, 2025a, Figure 6, p. 8)

Conclusion

The “archeology of the mind” is shedding new light on the historical process. In particular, we’re learning that history is a reciprocal relationship between the human mind and the culture it creates. We make culture, and it remakes us — by favoring those who better fit in and exploit its possibilities. Humans and culture have thus been remaking each other along trajectories that differ from one population to another.

Culture especially favors those individuals who have certain properties of mind and behavior, notably cognitive ability. “Cultural selection” has gone farther in some parts of the world than in others, and has gone especially far in Europe and East Asia.

These two trajectories of cognitive evolution share certain similarities:

• An upward trend that began with the emergence of farming and continued as social complexity progressively increased.

• A halt and even reversal at a fairly high level of social complexity — during the periods of Imperial Rome and Tang China — perhaps due to a fertility decline among the upper classes.

But this reversal would play out differently in the two regions. In the Roman Empire, mean cognitive ability fell and then rose again a few centuries later. This “reboot” was associated with an ideological change: the rise of Christianity and its preeminent role in enforcing morality, particularly in the areas of sexual behavior and family formation. The new ideology would survive the collapse of Rome and define what would be called “Christendom” (Frost, 2024).

In China, the ensuing trajectory is less clear-cut. Whereas alleles associated with EA suggest a flat-lining of cognitive evolution, those associated with IQ suggest a decline. Perhaps the decline in upper-class fertility was partially offset by the growing importance of the Imperial Examination for social and economic advancement (Frost, 2011; Wen et al., 2024). This examination favored not only high cognitive ability but also certain other mental traits: rule following, submissiveness, and resistance to boredom — in short, the ability to sit still at a desk and do tedious assignments.

Like Imperial Rome, Tang China experienced a movement for social and moral reform. This movement was based on the teachings of Kong Qiu (Confucius) and gained ground during the Tang Dynasty through the work of pronatalist scholars who condemned abortion and advocated control of sexual desires (Hsiung, 2011). Their work led to the emergence of neo-Confucianism, which resembled Christianity in some respects, particularly in its efforts to regulate sexual behavior and family formation. But, unlike Christianity, it did not create an organizational network to impose its norms on the population.

In sum, East Asia failed to develop a regulator of public morality that could act independently of the elites and make them conform to certain sexual and reproductive norms. This point is broadly made by Francis Fukuyama in his work The Origins of Political Order.

Hopefully, we will see more studies of this kind, especially on the last 1,500 years of cognitive evolution. In particular, three points need to be cleared up:

• The higher values for Mongolians in relation to those for northeast Chinese. The apparently high cognitive ability of Mongolians may seem surprising, given the evidence that IQ is 5 points lower among them than among Han Chinese (Lynn, 2007).

• The apparent contradiction between Davide Piffer’s findings and Ron Unz’s. The latter has argued that mean cognitive ability continued to rise in China until the 20th century through higher fertility among the wealthy (Unz, 2013).

• The point in time when East Asians surpassed Europeans in mean cognitive ability. The crossover point seems to be just before the time of Christ, according to Piffer’s comparison of ancient DNA from those two regions (see below). This is a rough estimate because the data are aggregated into 2,000 year “bins.” In fact, Europeans may have surpassed East Asians c. 1900, given that the last bin includes both the peak of the Tang Dynasty (thus overstating recent East Asian evolution) and the nadir of Rome’s Imperial Era (thus understating recent European evolution).

Increase in mean cognitive ability, as measured by alleles associated with educational attainment, over the last 12,000 years in eastern Eurasia and Europe. (Piffer, 2025b)

References

Akbari, A., Barton, A.R., Gazal, S., Li, Z., Kariminejad, M., Perry, A., Zeng, Y., Mittnik, A., Patterson, N., Mah, M., Zhou, X., Price, A.L., Lander, E.S., Pinhasi, R., Rohland, N., Mallick, S., & Reich, D. (2024). Pervasive findings of directional selection realize the promise of ancient DNA to elucidate human adaptation. bioRxiv https://doi.org/10.1101/2024.09.14.613021

Frost, P. (2011). East Asian intelligence, Evo and Proud, February 18. https://evoandproud.blogspot.com/2011/02/east-asian-intelligence.html

Frost, P. (2019). The Original Industrial Revolution. Did Cold Winters Select for Cognitive Ability? Psych, 1(1), 166-181. https://doi.org/10.3390/psych1010012

Frost, P. (2024). How Christianity rebooted cognitive evolution. Aporia Magazine, October 10.

Fukuyama, F. (2011). The Origins of Political Order. From Prehuman times to the French Revolution. Profile Books.

Hawks, J. (2011). Selection for smaller brains in Holocene human evolution. arXiv:1102.5604 [q-bio.PE] https://arxiv.org/abs/1102.5604

Hsiung, P. C. (2011). More or Less; Marital Fertility and Physical Management in Late Imperial China. Journal of Archaeology and Anthropology, (74), 119-167.

Kuijpers, Y., Domínguez-Andrés, J., Bakker, O.B., Gupta, M.K., Grasshoff, M., Xu, C.J., Joosten, L.A.B., Bertranpetit, J., Netea, M.G., & Li, Y. (2022). Evolutionary Trajectories of Complex Traits in European Populations of Modern Humans. Frontiers in Genetics, 13, 833190. https://doi.org/10.3389/fgene.2022.833190

Lipson, M., Cheronet, O., Mallick, S., Rohland, N., Oxenham, M., Pietrusewsky, M., ... & Reich, D. (2018). Ancient genomes document multiple waves of migration in Southeast Asian prehistory. Science, 361(6397), 92-95. https://doi.org/10.1126/science.aat3188

Lynn, R. (2007). IQ of Mongolians. Mankind Quarterly, 47(3), 91. http://doi.org/10.46469/mq.2007.47.3.5

Piffer, D. (2025a). Directional Selection and Evolution of Polygenic Traits in Eastern Eurasia: Insights from Ancient DNA. Twin Research and Human Genetics, First view, 1-20. https://doi.org/10.1017/thg.2024.49

Piffer, D. (2025b). Temporal trends for EAS and EUR. https://x.com/DavidePiffer/status/1886032099779502459

Piffer D, Dutton, E., & Kirkegaard, E.O.W. (2023). Intelligence Trends in Ancient Rome: The Rise and Fall of Roman Polygenic Scores. OpenPsych. Published online July 21, 2023. https://doi.org/10.26775/OP.2023.07.21

Piffer, D., & Kirkegaard, E.O.W. (2024). Evolutionary Trends of Polygenic Scores in European Populations from the Paleolithic to Modern Times. Twin Research and Human Genetics, 27(1), 30-49 https://doi.org/10.1017/thg.2024.8

Unz, R. (2013). How Social Darwinism Made Modern China. The American Conservative, March 11. https://www.theamericanconservative.com/how-social-darwinism-made-modern-china-248/

Wen, F., Wang, E. H., & Hout, M. (2024). Social mobility in the Tang Dynasty as the Imperial Examination rose and aristocratic family pedigree declined, 618–907 CE. Proceedings of the National Academy of Sciences, 121(4), e2305564121. https://doi.org/10.1073/pnas.2305564121

Yang, M. A., Fan, X., Sun, B., Chen, C., Lang, J., Ko, Y. C., ... & Fu, Q. (2020). Ancient DNA indicates human population shifts and admixture in northern and southern China. Science, 369(6501), 282-288. https://doi.org/10.1126/science.aba0909

Zhao, J., Wurigemule, Sun, J., Xia, Z., He, G., Yang, X., et al. (2020). Genetic substructure and admixture of Mongolians and Kazakhs inferred from genome-wide array genotyping, Annals of Human Biology, 47, 7-8. https://doi.org/10.1080/03014460.2020.1837952

Zhao, Z. (2002). Fertility control in China's past. Population and Development Review, 28(4), 751-757. https://doi.org/10.1111/j.1728-4457.2002.00751.x

Zhao, Z. (2006). Towards a better understanding of past fertility regimes: the ideas and practice of controlling family size in Chinese history. Continuity and Change, 21(1), 9-35. https://doi.org/10.1017/S0268416006005777

Bernard Saladin d’Anglure, 2014, Livre 1. Enfance et vocation, Les possédés et leur monde. YouTube

As a graduate student at Université Laval, I was fascinated by Bernard Saladin d’Anglure. He had a talent for showmanship. Once he taught our class with the room completely in darkness, and it was only at the end that he slowly turned the lights on, to prevent us from being “blinded by the light.” It was his way of showing how a new way of thinking can completely overturn a society.

He also had a talent for fieldwork. Unlike many professors, he never left the “field.” He remained in contact with the Inuit of the far north and worked with them to promote their culture through films like Atanarjuat, the fast runner and novels like Sanaaq. He felt that Western culture was destroying Inuit communities by creating the wrong norms and expectations. Culture should belong to the people who live it, and not to outsiders.

Finally, he had a talent for noticing patterns, as shown by his work in structuralism. By examining the cultural production of a people — stories, beliefs, artefacts — one could, so he thought, identify “structures” that ultimately reflect structures in the mind itself. I wanted to push this analysis further, in the belief that these mental structures were created by a coevolution between a population and its culture. But Bernard himself never went that far.

He was better known than any other professor in our department, but that renown was the one thing I had mixed feelings about. To increase the reach of his ideas, he would partner with key groups and individuals whose ideas seemed to overlap with his own … more or less.

This was particularly true for his work on “the third gender.” Initially, he was interested in transgendering among the Inuit: Inuit newborns were traditionally named after a deceased ancestor or relative and raised according to their namesake’s gender. If a boy was named after a woman, he would be raised as a girl until puberty — which meant being dressed in girl’s clothing and taught how to sew. Conversely, if a girl was named after a man, she would be raised until puberty as a boy — which meant being dressed in boy’s clothing and taught how to hunt.

As Bernard saw it, such children were a “third gender” that challenged the male-female dichotomy. He went on to argue that this ternary group existed in all human societies, being encouraged in some and repressed in others, particularly in our own.

The rest of the anthropology department disagreed [this was in the late 1980s]. As they saw it, there are only two sexes, and some variability exists within each of them. Some men tend to be feminine, and some women tend to be masculine. But these men and women don’t form a group of their own. In fact, they differ as much from each other as do men as a whole and women as a whole.

These counter-arguments didn’t seem to bother Bernard, and I couldn’t help but wonder whether his term “the third gender” was simply a marketing ploy.

It certainly won him a lot of attention, particularly from the emerging discipline of “queer studies.” Initially, he enjoyed the attention. His work became better known, especially at colleges and universities across the English-speaking world.

A few years ago, however, during my last conversation with him, he expressed concern about his work being misinterpreted. He outlined several points that bothered him:

• These Inuit children were not exercising a right to self-expression; in fact, they were fulfilling the wishes of their deceased relatives and ancestors.

• These children were also expected to assume their biological gender upon reaching puberty. This is not at all the case in modern Western societies, where gender identity is increasingly seen as a personal choice to be made at any age and for any reason … or no reason at all.

• Finally, Inuit transgenderism was always confined to a minority of children. Among adults, it was completely absent. The situation was not unlike that of Western societies until the early 20th century, when little boys were commonly dressed as girls.

Franklin D. Roosevelt, at the age of two and a half (Wikicommons)

There seems to be a cross-cultural tendency to see little boys and girls as neither male nor female, probably because they share a similar basic phenotype that can pass for female. This is indeed the “third gender” that Western societies have sought to erase over the past century.

This erasure has been accompanied by a sexualization of little boys and girls to a degree that would seem perverse in other times and places and which has led to sexually immature children wondering about themselves. It doesn’t take much to convince a 5-year-old boy that he isn’t a man, since he clearly isn’t.

Bernard went on to say that sexuality was being turned into an act of individualism, whereas traditionally it had been the bond that held a society together. The same point came up in one of his last interviews:

… around that [the issue of sexual orientation], [they] want to bring together all the minorities who feel oppressed by an oppressive majority … So everyone can choose. It’s a mixture of individualism and … we are very far from the social bond that constitutes the basis of a society. [In their opinion] it is the individual who chooses, who wants to be [an individual], and in the name of human rights. But human rights are again an individual vision. At the beginning of the century, the first sociologists and anthropologists — Mauss, Durkheim, etc. — saw the social bond as something essential. (Saladin d’Anglure, 2014).

Soon afterwards, his health took a turn for the worse, and his public life came to an end. Today, he is remembered as a supporter of certain ideas that he actually did not support.

I understand the difficulties of being a public intellectual. To promote your ideas, you have to align yourself with influential groups. But such alignment comes at a price. You must make compromises between your ideas and theirs …

Yes, my dear readers, I hear you! “Don’t worry Peter! When I get to be famous, I’ll set things straight and say what I really think!”

But what if that day never comes? Or what if it comes, and you no longer have the strength — physically or mentally — to speak your mind?”

References

Saladin d'Anglure, B. (1986). Du fœtus au chamane: la construction d'un «troisième sexe» inuit. Etudes/inuit/studies, 25-113. http://www.jstor.org/stable/42869539

Saladin d’Anglure, B. (2007). Troisième sexe social, atome familial et médiations chamaniques: pour une anthropologie holiste: Entretien avec Bernard Saladin d’Anglure. Anthropologie et sociétés, 31(3), 165-184. https://doi.org/10.7202/018381ar

Saladin d’Anglure, B. (2012). The “Third Gender”. Revue du MAUSS, 39(1), 197-217. https://shs.cairn.info/journal-revue-du-mauss-2012-1-page-197?lang=en

Saladin d’Anglure, B. (2014). Livre 19. Les dangers du rationalisme moderne et de la biologisation: l’anthropologie de la naissance chez les Inuit, le cerveau reptilien et le travestissement. Les possédés et leurs mondes, March 28. Interviewer: Frédéric Benjamin Laugrand.

Kostel na kopci, Jindrich Štýrský (1921) - Wikicommons

Unlike other animals, humans have adapted much more to culture than to nature —things like climate, soil, and vegetation. Consider the rate of change to the human genome. This rate increased more than a hundred-fold some 10,000 years ago, when our ancestors had already spread from the Tropics to the Arctic and were no longer adapting to new natural environments. But they were adapting to new cultural environments. Hunting and gathering gave way to farming, which then brought many other changes:

• rapid growth of population

• shift to sedentary living in villages and towns

• emergence of the state and its monopoly on violence

• increase in peaceful interactions with non-kin

• resolution of disputes through codified law

• growth of trade, specialization of labor, literacy and numeracy, social stratification, and much more.

When farming replaced hunting and gathering, the rate of change to the human genome increased more than a hundred-fold (Hawks et al., 2007). 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 acceleration has been due to the quickening pace of cultural change over those 10,000 years. Like nature, culture can decide who will survive to pass on their genes. Unlike nature, it is a human creation, being in fact the human-made portion of our environment. We make culture, and it remakes us — by favoring those who better fit in and exploit its possibilities. Humans and culture have thus been remaking each other along trajectories that differ from one population to another.

But didn’t cultural evolution replace genetic evolution? Didn’t one make the other unnecessary? That used to be the accepted view, but the truth now seems to be the opposite. When humans gave up hunting and gathering, they had to adapt to a faster-changing cultural environment, and not simply a slow-changing natural one. So new adaptations appeared at a faster rate. Yes, many were nongenetic, but the sharp rise in adaptive change of any sort — whether learned behaviors or innate capabilities — necessarily spurred the pace of genetic change. Instead of slowing down, genetic evolution began to go even faster! (Cochran & Harpending, 2009; Hawks et al., 2007; Rinaldi, 2017).

What exactly was causing this acceleration? In part, humans were adapting to new diets, particularly grains instead of fruits, tubers, and vegetables. They were also adapting to infectious diseases — smallpox, typhus, cholera, bubonic plague — which could spread like wildfire through towns and cities.

Above all, our ancestors were adapting to new ways of using their minds. As society became more complex, they had to evaluate a larger number of people who were neither close kin nor close friends. They also had to decode a greater amount of encoded information — not only in speech but also on surfaces of clay, papyrus, and parchment. Finally, they had to make a wider range of objects from a wider range of materials, first wood and stone, then metal, glass, and gems.

All of this required an increase in the capacity to process information, recognize patterns, and solve problems — in other words, an increase in cognitive ability.

First cognitive advance: the peopling of northern Eurasia

We will focus here on ancestral Europeans, beginning with those modern humans who entered northern Eurasia some 30,000 years ago, long before farming and the cultural changes it caused.

At that time, cognitive ability still served to meet the challenges of hunting and gathering. Those challenges were especially great in cold environments, as shown by the inverse correlation between temperature and technological complexity in present-day hunter-gatherers.

Figure 1: Inverse correlation between technological complexity and temperature in present-day hunter-gatherer groups (Hoffecker, 2002, p. 10).

The most challenging environments were in northern Eurasia until the end of the last ice age some 10,000 years ago. Food was potentially abundant but almost wholly “meat on the hoof” — herds of wandering reindeer and other herbivores. That food was inaccessible unless one could:

• collect, process, and remember huge quantities of spatiotemporal data to track the herds, predict their movements, and find one’s way back to camp;

• monitor untended traps and snares to capture solitary animals, which were dispersed over a larger territory than in warmer environments and took too much time to hunt;

• make cold-resistant shelters and clothing;

• plan ahead to store fuel for winter and food for times of need.

Other cognitive challenges arose from a new sexual division of labor. Because women had few opportunities for food gathering, they specialized much more in higher-order tasks: weaving, needlework, leatherwork, garment making, pottery, kiln operation, shelter building, and food processing.

For all of the above reasons, northern Eurasians were selected for higher cognitive ability, as shown by the positive correlation between latitude and brain size. Indeed, brain size correlates with latitude not only in humans and other hominids but also in any animal that remains active during winter. As a general rule, cognitive demands are higher in colder, more seasonal environments (Beals et al., 1984; Frost, 2019a; Kirkegaard, 2022).

Northern Eurasians were thus “pre-adapted” for the cognitive demands of complex societies in the future — including many that would arise among humans native to the temperate and tropical zones. Such societies would be repeatedly taken over and replaced by northern Eurasians who, though less adapted climatically, could better exploit the possibilities being created.

Figure 2: Human variation in cranial capacity. Black, 1,450 cc and over; checkerboard, 1,400-1,449 cc; crosshatching, 1,350-1,399 cc; horizontal striping, 1,300-1,349 cc; diagonal striping, 1,250-1,299 cc; dots, 1,200-1,249 cc; white areas, under 1,200 cc (Beals et al., 1984).

Second cognitive advance: the rise of farming

Hunting persisted after the ice age, but the cognitive demands were no longer the same. The tundra was now boreal forest, where hunters pursued solitary animals over short distances, rather than large herds over longer distances. So the human mind had less need to store and process huge quantities of spatiotemporal data. Shorter hunting distances, followed by the shift to farming and sedentism, may explain why brain size decreased during post-glacial times, a trend that has been shown for European and Chinese populations but not for tropical ones, like the Nubians (Hawks, 2011). Did the decrease in brain size reflect a decrease in cognitive ability? Or was cognition simply reoriented toward tasks that consume less memory? Perhaps the excess mental capacity made new tasks easier to create and develop.

About 12,000 years ago, farming replaced hunting and gathering in the Middle East, followed by southeast Europe three thousand years later. Over the next thousand years, this new form of subsistence would spread north and west to a line stretching from the Netherlands to the Black Sea. After a pause, it resumed its advance and was still spreading into northern Europe on the eve of recorded history.

We have retrieved enough DNA from European remains to chart the evolution of cognitive ability since the advent of farming (Akbari et al., 2024; Kuijpers et al., 2022; Piffer et al., 2023). To be precise, we can chart changes in the population frequencies of alleles associated with intelligence, fluid intelligence, educational attainment, and household income — all of which are highly correlated with cognitive ability. Of course, these measures may also track other mental abilities, like proactiveness and rule following. If you show initiative and follow the rules, you will go farther at school and earn more money, independently of your cognitive ability.

Not much DNA is available from the Paleolithic or the Mesolithic, but the little we have indicates no change in mean cognitive ability during the long period of hunting and gathering. If more DNA becomes available, we should break it down by region. Did cognitive ability differ between the Arctic-adapted hunter-gatherers of northern Europe and those of central and southern Europe?

Around 9,000 years ago, Europeans became sufficiently numerous to provide enough DNA for study. Mean cognitive ability was now rising, apparently with the spread of farming into Europe, and the rise was especially steep until 7,000 years ago. Was this initial surge due to the cognitive demands of farming? Perhaps. But we should not rule out the consequences of admixture from indigenous hunter-gatherers. Admixture certainly occurred, for the earliest farmers looked distinctly more African than the later ones (Angel, 1972; Brace et al., 2006; Frost, 2015). As farmers spread into Europe, they must have mixed more and more with the indigenous hunter-gatherers, including northern groups that had previously adapted to Arctic environments.

This admixture is difficult to quantify. We could assume that Middle Eastern ancestry corresponds to the genetic difference between farmers and hunter-gatherers, but that assumption would be wrong. Some of the difference would also be due to founder effects, and some to natural selection. For instance, haplogroup U is often cited as a genetic marker of indigenous hunter-gatherers, yet it disappeared from Europe long after the transition to farming, probably because it was no longer needed (Melchior et al., 2010). This haplogroup shifts the energy balance toward production of body heat — an advantage for people who sleep in makeshift shelters and pursue game in all kinds of weather (Balloux et al., 2009).

In sum, we over-estimate Middle Eastern ancestry if we assume that all genetic change across the hunter-gatherer/farmer time boundary is due to the latter replacing the former. This change also includes “false positives”:

• alleles that may have originated among farmers but then steadily increased through natural selection in the farming population — regardless of its degree of hunter-gatherer ancestry.

• founder effects among those hunter-gatherers who joined farming communities. These effects were random but a certain number would be "false positives" and wrongly indicate farmer ancestry.

• convergent evolution, where similar selection pressures produced similar genetic adaptations in those hunter-gatherers who adopted farming.

The initial cognitive surge may thus reflect an ongoing inflow of ex-hunter-gatherers into farming communities — as spouses, as captives, or even as landowners with a right to settle there. Anthropologists have observed all three scenarios, and it is impossible to say which one applies here. Whatever the scenario, these ex-hunter-gatherers became much more present in DNA samples from archeological sites, which, in most cases, had been farming communities. It may therefore be no coincidence that the steep rise in mean cognitive ability began roughly when the advancing farmers first encountered the hunter-gatherers of northern and eastern Europe — who had adapted to the high cognitive demands of Arctic environments.

So the subsequently slower rise should not mislead us. Cognitive evolution did not slow down; rather, with no hunter-gatherers remaining to be assimilated, this evolution would be driven only by the cognitive demands of farming and the progressive increase in social complexity.

Social stratification

Subsequent evolution would also be driven by a stronger linkage between cognitive ability and reproductive success. Farming created a storable food surplus that fell under the control of powerful families, thereby increasing their power further (Harris, 1959; Testart et al., 1982). Society thus became stratified, and this stratification would determine the pace of cognitive evolution through a continual replacement of lower social classes by higher ones:

• Alleles associated with high cognitive ability became concentrated in the dominant class through the rise of more capable individuals and the fall of less capable ones.

• Because this class enjoyed higher fertility and lower mortality, it produced more offspring than could be supported by niches at the same social level.

• Surplus offspring moved down into niches left vacant by the lower classes, which had lower fertility and higher mortality.

• The lower classes were thus continually replaced by the higher ones, with the result that mean cognitive ability steadily rose from one generation to the next.

The same “rinse and repeat” cycle has been described by Gregory Clark in his work on English demography from the Middle Ages onward (Clark, 2007; 2009; 2023).

Pacification of social relations

Stratification would also assist cognitive evolution through a less direct chain of causation:

• Emergence of the State - Powerful families initially asserted their power through physical strength, charisma, and eloquence. This power gave them control over key resources, particularly the food supply, which enabled them to increase their power further — through a self-reinforcing feedback loop. In particular, they could hire young men for a personal guard and, eventually, an army. This was how the State came to be.

• Imposition of a State monopoly on violence - To eliminate potential rivals, the State reserved the use of violence to itself alone and maintained this monopoly by eliminating any man who used violence on his own initiative.

• Pacification of social relations - By eliminating such men, the State pacified not only social relations but also, over time, the gene pool: alleles associated with male violence were selected out, while those associated with deference and submission were selected in.

• Expansion of trade and other activities with high cognitive input - Peaceful interactions became the norm. Eventually, most men could live their entire lives without having to defend themselves or their families. They were free to devote themselves to other activities, including many that required specialized skills. Trade in particular benefited, both between and within communities. There thus arose a new kind of man who was not only less aggressive but also better able to handle the cognitive demands now possible (Frost, 2010; Frost & Harpending, 2015).

Figure 3: Changes in European DNA over the last 9,000 years for alleles associated with intelligence, household income, and years of schooling. Note the steep rise when farming replaced hunting and gathering. Also note the decline during the time of Imperial Rome (Akbari et al., 2024).

First cognitive retreat: the Imperial Era of Rome

Mean cognitive ability rose throughout the Neolithic and into historic times — a process eventually culminating in the civilizations of Greece and Rome. In this flowering of material and intellectual culture, the most intelligent individuals could go far beyond satisfying their basic needs and ask questions about the world, thus laying the groundwork for philosophy, mathematics, history, geography, astronomy, and other fields of enquiry.

But it was not to last. According to ancient DNA from central Italy, mean cognitive ability rose under the Roman Republic but then fell sharply under the Roman Empire. It would then rise again from Late Antiquity to post-medieval times (Piffer et al., 2023). The decline of the Imperial Era seems to have been confined to Roman territory, as it does not clearly show up in the first two studies of DNA from Europe as a whole (Kuijpers et al., 2022; Piffer & Kirkegaard, 2024), while appearing in the latest study, which used a larger dataset (Akbari et al., 2024).

Figure 4: Rise, fall, and renewed rise of mean cognitive ability in central Italy, as measured by alleles associated with educational attainment (EA4) (Piffer et al, 2023, Figure 1)

The cognitive decline of Imperial Rome likely had three causes:

• Decrease in fertility and family formation among the elite (Caldwell, 2004; Hopkins, 1965; Roetzel, 2000, p. 234; Sullivan, 2009, pp. 27-28, 35-38).

• Hypogamy between elite men and lowborn women, often in the form of polygyny with slave women or newly freed women. Elite women thus lost reproductive importance (Perry, 2013).

• Increase in the slave population, particularly foreign slaves (Harris, 1999. Previously, if elite offspring failed to find niches within their social level, they could move farther down the social ladder, but that option disappeared once those niches had been filled with slaves. Thus ended the “rinse and repeat” cycle of cognitive evolution.

This new social environment changed the life goals of the elite, causing them to value material wealth over family life. Sexual needs, though still important, could now be satisfied in less entangling ways, with slaves or prostitutes.

Some emperors challenged the new normal but ran into resistance from the very people whose lineages were dying out (Sullivan, 2009). In 27 BCE, Augustus Caesar condemned childlessness in a speech to Rome’s equestrian class:

For you are committing murder in not begetting in the first place those who ought to be your descendants; you are committing sacrilege in putting an end to the names and honours of your ancestors; and you are guilty of impiety in that you are abolishing your families, which were instituted by the gods, and destroying the greatest of offerings to them — human life. [Cassius Dio. Roman History, 56.5.2]

A related problem was the growing dependence on slavery, particularly for agriculture, which made the Empire ever more vulnerable to disturbances from within and without. Because slaves wished to grow only enough food for their own needs, they had to be coerced to meet the greater needs of the Empire (Harper, 2011, pp. 176, 250-254). Whenever coercion weakened during times of unrest or invasion, production fell and Romans starved. The slaves knew this. Everyone knew this.

But the benefits of slavery were difficult to ignore, and they were not simply economic. There were also the personal benefits for slaveowners, especially the access to young women without the restrictions and obligations of marriage. Slaves knew this weakness, as attested by a character in a 5th-century comedy: “We all steal and no one turns traitor, for we are all in it together. We watch out for the master and divert him, for slaves and slave-girls are in alliance” (Harper, 2011, p. 251).

One outcome was a growing number of alumni — offspring of masters and slave girls, or any slave children for whom the master had developed affection. If a master had no children by his wife, he could leave his property to such alumni (Rawson, 1986, 2014, pp. 173-179). Consequently, elite men were not just having fewer offspring by their wives; they were also having more by lowborn mistresses or simply adopting any the latter already had. Either way, Rome’s elite was no longer passing on its characteristics to future generations.

Rome needed a moral revolution. When Constantine the Great legalized Christianity, he saw it as a means to revitalize the Empire. Was he himself a Christian? Doubtful. His edicts were more consistent with a pagan mindset. One edict ordered the execution of any woman who had become involved with or married a slave, as well as the burning of her slave partner. Another authorized torture of slaves as a means to ascertain the truth (Thompson, 2008).

Third cognitive advance: the rise of Christianity

The cognitive decline of Imperial Rome reversed with the recognition of Christianity as a legal faith in 313 and as the State religion in 380. This reversal coincided with the growing power of the Church to intervene in many areas of life, especially marriage and procreation. In particular, the Church aggressively pushed monogamy and forced men to focus on procreation with a lawful wife, usually of similar status, rather than on sex with prostitutes or slave women. Successful men were now more likely to pass on the qualities of mind and behavior that had made them successful.

Did the new faith have that goal in mind? To some extent, yes. Early Christians understood that some lifestyles are more procreative than others. They also understood the basics of population genetics: 1) humans inherit not only physical traits but also mental and behavioral ones; 2) the latter vary not only among individuals but also among populations; and 3) selection for such traits will progressively change a population over successive generations.

This understanding of genetics was put into writing by Origen of Alexandria (c. 185–253 CE), a theologian who wrote over two thousand treatises on many subjects. Of course, the term “genetics” did not yet exist. He spoke instead of logoi spermatikoi, “the principles of the seed.” Logoi is difficult to translate; it conveys the idea of information that brings order out of disorder.

As Origen explained it, “the seed of someone has within itself — still immobile and placed in reserve — the procreator’s logoi” (Origen, Commentary on the Gospel of John, XX, 3).

This seed, which organizes both body and spirit, is passed down from one’s ancestors:

Just as among the seeds of the body there emerges sometimes, from a great number, a seed endowed with a greater capacity for action, so may we observe the same phenomenon with the seeds of the spirit. What I have said will become clear after what follows: since the procreator has within himself logoi from his ancestors and his collateral lines, it is sometimes his logos that prevails — and the child that comes into the world resembles its procreator — or sometimes the logos of his brother, his father or his uncle, sometimes even his grandfather. This is why those who come into the world resemble one or another [of that line]. We may equally see the wife’s logos prevail, or that of her father, her brother or her grandfather … (Origen, Commentary on the Gospel of John, XX, 35-36)

These logoi vary among humans:

It is evident that men have not all come into human life with absolutely identical logoi spermatikoi seeded into their souls. (Origen, Commentary on the Gospel of John, XX, 5)

Try to imagine whether it is not without reason that God destroys certain seeds in order to prevent the bad ones from multiplying on the earth, once there have been sown those [seeds] whose tendencies are not from the best, and to cultivate the products of superior seeds. This is why the Flood happened, to wipe out the seed of Cain … (Origen, Commentary on the Gospel of John, XX, 25)

Another theologian, Augustine of Hippo (354–430 CE), likewise understood that people differ innately in body and spirit, particularly in the power to distinguish bad desires from good ones and the power to put the latter into effect. He initially believed that these powers could be lacking only in a non-Christian, but he later acknowledged that a Christian could lack them too, despite baptism (Pang-White, 2001). Thus, a sinful act may arise from a weakness of will, rather than from a will to do wrong:

Whatever these souls do, if they do it by nature, not by will, that is, if they lack a movement of the spirit [that is] free both to do and not to do, if finally, no power of abstaining from their action is given to them, we cannot consider the sin theirs. (Augustine, Retractationes 1,14, 6)

Although such people are not responsible for their weakness of will or their inability to understand rules, they should still be shunned and resisted:

… But they are an evil for us if we are enticed and led astray by them; if, however, we are on our guard against them and overcome them, it is an honorable and glorious thing. (Augustine, Retractationes 1,14, 7)

In sum, Christianity reversed the cognitive decline of Imperial Rome, thereby launching an upward trend that would last throughout the Christian Era. The reversal was achieved by increasing the reproductive importance of elite women, essentially by limiting polygyny between elite men and lowborn women. Polygyny was already forbidden by pagan Roman law, but that ban had been ignored and openly flouted. Thus, in this and other ways, the new faith changed the culture not so much by making new rules as by reviving old rules and enforcing them more effectively.

Indeed, Christianity was much more rules-based than paganism, particularly the watered-down version of the Imperial Era. Pagans were more interested in developing a transactional relationship with their favorite god: “do this for me, and I’ll do that in exchange.” Their “religion” was not one as we understand the term:

It might be less confusing to say that the pagans, before their competition with Christianity, had no religion at all in the sense in which that word is normally used today. They had no tradition of discourse about ritual or religious matters (apart from philosophical debate or antiquarian treatise), no organized system of beliefs to which they were asked to commit themselves, no authority-structure peculiar to the religious area, above all no commitment to a particular group of people or set of ideas other than their family and political context. (North, 2013)

Christianity may have therefore assisted cognitive evolution not only indirectly, by increasing the reproductive importance of elite women, but also directly — by rewarding those individuals who could better learn and follow rules, while punishing those who could not. Rule following requires cognitive ability, while soliciting other mental domains and their underlying genetic factors (O’Gorman et al., 2008; Sasaki et al., 2013).

Fourth cognitive advance: the rise of the middle class

In the late Middle Ages, northern and western Europe became consolidated into a few nation states that expanded outward, first within Europe and then beyond. Eventually, these states would dominate the world, as evidenced not only by their colonial empires but also by their ascendancy in trade, science, technology and almost any other field of activity.

Gregory Clark has argued that the medieval/post-medieval period saw English society become increasingly middle class, mentally and behaviorally. "Thrift, prudence, negotiation, and hard work were becoming values for communities that previously had been spendthrift, impulsive, violent, and leisure loving” (Clark, 2007; Clark, 2009; Clark, 2023; Frost, 2022c).

Georg Oesterdiekhoff has argued for a similar evolution across Western Europe as a whole. He views this evolution in terms of Jean Piaget’s stages of cognitive development, i.e., more and more people could go beyond preoperational thinking (egocentrism, anthropomorphism, animism) and achieve operational thinking (ability to understand probability, cause and effect and another person’s perspective) (Oesterdiekhoff, 2023; Rindermann, 2018, pp. 49, 86-87).

With the retrieval of DNA from human remains, we can now verify this model of recent Western European evolution by comparing genomes from different time periods. Two such studies have been done recently.

In the first study, present-day genomes were compared with genomes from Late Antiquity to the Middle Ages (n = 467). The comparison showed a substantial rise in mean cognitive ability over time—between one third and one half of a standard deviation (Frost, 2024a; Piffer & Kirkegaard, 2024).

The actual rise may have been even larger, since it is imperfectly measured by a comparison between medieval and present-day genomes. At one end of the timeline, according to Gregory Clark, cognitive ability had already begun to rise during the Middle Ages. At the other, it may have peaked in the Victorian Era and then declined thereafter (Frost, 2022d). Also, the rise in cognitive ability may have begun earlier in some regions than in others.

On a side note, the cognitive increase is underestimated by alleles associated with socioeconomic status (SES), perhaps because such alleles favor aspects of mind and behavior that contribute much more today to social and economic success than they did in the past.

Table 1: Size of the cognitive increase in S.D. (Cohen’s d) for England and for Europe as a whole, as measured by alleles associated with EA3, EA4, IQ, and SES (Piffer & Kirkegaard, 2024). EA3 is based on alleles identified by Lee et al. (2018). EA4 is based on alleles identified by Okbay et al. (2022).

To address these limitations, a second study was conducted with genomes from only one region of England (Cambridge and surrounding area, n = 269). The genomes are from the eleventh to nineteenth centuries (Piffer & Connor, 2025).

Table 2: Distribution of genomes by century and mean EA polygenic score by century (Piffer & Connor, 2025. Table 1).

For some of these centuries, notably the sixteenth to eighteenth, we have too few genomes for a century-by-century analysis. Nonetheless, the overall trend is clear: a rapid increase in mean cognitive ability from the 1300s onward.

Particularly impressive is the increase in the "smart fraction": the top 1% in 1850 was as smart as the top 0.1% in the year 1000. In the following table, EA scores have been converted into IQ scores for ease of interpretation.

Table 3. Estimated percentage of the population in the top 5% and top 1% as defined by 1850 norms and using the linear regression for the whole period (Piffer & Connor, 2025, Table 3).

No longer voices in the wilderness, intellectuals could now gather in salons or coffeehouses, come together in learned societies, and publish their ideas in academic journals. This synergy would give rise to the Enlightenment and the Industrial Revolution, ultimately energizing all areas of life—not only the sciences but also literature, music and the arts (de Courson et al., 2023).

Individualism and weak kinship

Cognitive ability was not the sole mental trait behind the rise of northwest Europeans to global dominance. Some human groups have never achieved the same dominance, while having comparable or even higher levels of cognitive ability. Think of the Parsis, the Igbo, the Ashkenazim, or even the Chinese. Something held them back while pushing northwest Europeans forward.

Was it colonialism? The slave trade? The printing press? The Reformation? All of them together?

Yet in the 14th century — well before any of the above — England and Holland were already surpassing the rest of the world in GDP per capita growth (Frost, 2022a; Greer, 2013a; Greer, 2013b). This trajectory can be traced back to the 7th century, when trade began to intensify among the North Sea peoples (Melleno, 2014). Their network of exchange steadily grew in scale and in scope — through the commercialization of land, labor, and capital that used to exist outside the market economy.

Those traders were able to transcend their ties of kinship and organize new and more flexible relationships with a potentially much larger population. In this, they were unlike other trading peoples for whom a “market” was merely a marketplace — a small island of exchange limited in time and space, beyond which a person would usually produce not for a market but for family, kin, and lord. The North Sea traders were the first to break free of this economic model. They and their descendants would extend the market principle to society as a whole, thereby making kinship obsolete.

Initially, this trajectory was not made possible by a high level of cognitive ability. Although IQ scores and PISA results tend to be higher in the north and west of Europe, and lower to the south and east, this gradient seems to be relatively recent — an outcome of cognitive evolution since the late Middle Ages (Frost, 2022b; Jensen & Kirkegaard, 2024; Piffer & Lynn, 2022). Selection for cognitive ability was a consequence of the rise of northwest Europe, rather than an initial cause.

The initial cause was a pre-existing mindset that expressed itself in greater individualism and weaker kinship. This behavioral pattern has prevailed, and still does, north and west of a line running from Trieste to St. Petersburg, the so-called “Hajnal Line.” It is characterized by:

• solitary living for at least part of adulthood, with many individuals remaining single their entire lives;

• departure from home on reaching adulthood, either to form a new household or to circulate among unrelated households, typically as servants;

• greater individualism, less loyalty to kin, and more willingness to trust strangers;

• impersonal pro-sociality, i.e., social interaction that is less emotionally intense while encompassing more people (Schulz et al., 2019; see also Frost, 2017; Frost, 2020; Hajnal, 1965; Hartman, 2004; hbd*chick, 2014; ICA, 2020; MacDonald, 2019; Seccombe, 1992, pp. 94-95, 150-153, 184-190).

The underlying mindset may be summarized as follows:

Moral universalism and moral absolutism. Rules are followed more willingly if framed in universal and absolute terms, as opposed to the situational and relativistic framing of kinship.

Moralized perception of non-kin. Help is provided to non-kin if they belong to the same community of rule followers. Continual rule breaking leads to expulsion from the community. Insiders are thus separated from outsiders by a line dividing the morally worthy from the morally worthless. Xenophobia is thus much more a moral judgment than a simple rejection of the “Other.”

Proneness to guilt. Rule breaking is punished much more by guilt than by shame. A rule breaker feels guilt even when no one else has witnessed the rule breaking. Shame is felt only when others have witnessed it. Guilt is thus internally generated, whereas shame is socially imposed (Benedict, 1946).

Broader range of targets for affective empathy. Affective empathy is extended from the mother-child relationship to all social relationships. Through this involuntary transfer of another person’s feelings to oneself, rule breaking is thus emotionally experienced as harm not only to the other person but also to oneself. Affective empathy and guilt may be two aspects of the same mental mechanism (Frost, 2017; Frost, 2020).

Christian or pre-Christian origin?

The “northwest European mindset” is attributed by some to Western Christianity and by others to earlier causes. The first scenario is situated in the early Middle Ages when the Western Church severely limited cousin marriages and thereby encouraged individualism through weaker kinship ties. Previously, only first-cousin marriages had been banned. That ban was initially extended two degrees further in the 7th century, when the Church adopted the anti-incest prohibitions of the Visigothic Code. Then, in the early 9th century, the Church began to calculate degrees of kinship through the so-called “Germanic system,” thus doubling the number of forbidden marriage partners. This extreme ban on cousin marriages may have therefore come from norms that had already existed among northwest Europeans, particularly the Germanic tribes (Frost, 2020; McCann, 2010, pp. 57-58; Policy Tensor, 2021; Schulz et al., 2019).

Those norms were certainly enforced more effectively by the Church, but their actual origin seems to be pre-Christian. In the 9th century, when the extreme ban on cousin marriages was introduced, French households were already small and nuclear, with 12% to 16% of adults being unmarried and with adults usually marrying in their mid to late twenties (Hallam, 1985, p. 56; Seccombe, 1992, p. 94). This pattern of delayed marriage seems to have been present even further back in time, as suggested by the writings of Julius Caesar and Tacitus on the Germanic tribes of Antiquity:

Those who have remained chaste for the longest time, receive the greatest commendation among their people: they think that by this the growth is promoted, by this the physical powers are increased and the sinews are strengthened. And to have had knowledge of a woman before the twentieth year they reckon among the most disgraceful acts. Caesar, De Bello Gallico 6: 21

Late comes love to the young men, and their first manhood is not enfeebled; nor for the girls is there any hot-house forcing; they pass their youth in the same way as the boys. Tacitus, Germania 20

We now come to the second scenario, which is situated before Christianity, specifically in post-glacial times among the fishing peoples of the North Sea and the Baltic. Unlike other hunter-gatherers, they lived in large semi-sedentary communities thanks to an abundance of fish, seals, and shellfish (Price, 1991). They thus had to solve the “large society problem” at an early date, and not necessarily as later peoples would. Perhaps they reduced clannishness by discouraging cousin marriages and by imposing the same universal rules on all community members. Kin-based morality would thus give way to moral universalism and moral absolutism. This cultural evolution could have then acted as a template for genetic evolution (Frost, 2020).

Whatever its origin, this mindset would enable northwest Europeans to organize themselves into ever-larger entities, to a degree that would be impossible elsewhere because of kinship rivalries and loyalties. Such entities included not only the market economy, but also the nation state and the Church as a parallel state.

Rise of the State, the Church, and the market economy

The rise of these larger entities — the State, the Church, and the market economy — would trigger the cognitive advance of late medieval and post-medieval times, first in England and then, increasingly, across the whole continent.

The High Middle Ages (1000 - 1300 CE) saw the rise of the modern Western State, particularly its power to subdue local strongmen and enforce the rule of law. Meanwhile, the Church came to accept the need to execute violent males so that good men and women could live in peace. There thus emerged a State-Church consensus in favor of capital punishment. Courts imposed it more and more often and, by the late Middle Ages, were condemning to death between 0.5 and 1.0% of all men of each generation, with perhaps just as many offenders dying at the scene of the crime or in prison while awaiting trial. As a result, the homicide rate fell from between 20 and 40 homicides per 100,000 in the late Middle Ages to between 0.5 and 1 per 100,000 in the mid-twentieth century. The pool of violent men dried up until most murders occurred under conditions of jealousy, intoxication, or extreme stress. (Frost & Harpending, 2015).

The outcome was an ever-larger space for peaceful production and exchange. This space offered niches for individuals with the requisite cognitive skills, notably literacy, numeracy, and planning, as well as a desire to advance peacefully in life through trade and work, rather than theft and plunder.

Their descendants — the future middle class — would grow in number with the expansion of the market economy, eventually representing, in the case of England, the majority of lineages in the population. The historical economist Gregory Clark has argued that this demographic change brought a behavioral change: "Thrift, prudence, negotiation, and hard work were becoming values for communities that previously had been spendthrift, impulsive, violent, and leisure loving” (Clark, 2007, p. 166; Clark, 2009; Clark, 2023; Frost, 2022b).

The middle class grew especially through a population boom among “proto-industrial” artisans — semi-rural cottagers who produced on contract for urban merchants. This boom was due to a combination of low mortality and high fertility. Mortality was low because these artisans lived away from urban areas, where population growth was held in check by disease and malnutrition. Fertility was high because they employed their own family members, and thus had only one means to expand their workforce:

The family economy of cottage proto-industrialists was fundamentally different from that of either urban artisans or self-sufficient peasant agriculturalists. … The labour team of proto-industrial households was normally comprised exclusively of family members. Parents could not afford to retain live-in servants or hire others … (Seccombe, 1992, p. 183)

Because these artisans operated in elastic, competitive markets, some of them did much better than others. And because of the tight linkage between economic and reproductive success, some of them had many more children than others — generally by marrying earlier and by helping their offspring do likewise.

This type of cottage industry emerged in the 15th century and reached its peak in the 17th, 18th, and early 19th centuries. It existed across most of Europe but was especially concentrated in the semi-rural areas of Westphalia, Saxony, the Zurich uplands, Piedmont, Lombardy, Alsace, Flanders, Brittany, Lancashire, Yorkshire, and Ulster (Seccombe, 1992, pp. 205, 217). These regions would contribute disproportionately to the gene pool of modern Europe, thereby creating a cognitive gradient between Western Europe and those regions where cottage industry appeared later and to a lesser extent (Frost, 2022c; Jensen & Kirkegaard, 2024; Piffer & Lynn, 2022).

This gradient can be seen today on the Italian peninsula. Students from the north consistently outperform those from the south on the INVALSI, an annual exam in Italian schools. Academic performance differs on average by a little over half a standard deviation (Piffer & Lynn, 2022).

The difference in academic performance is paralleled by a difference in alleles associated with educational attainment, as shown by a study of 1,076 individuals whose grandparents were born in the same Italian region. The data revealed “a clear north-south gradient, with central Italians occupying an intermediate position” (Piffer & Lynn, 2022). The cause is likely a post-medieval divergence: the south participated much less in the emerging market economy and, by the 18th century, was lagging far behind the north; its middle class was still small and economic relationships were still structured by paternalism and familialism (De Rosa, 1979).

Figure 6: Distribution of alleles associated with northern and southern Italians (Piffer & Lynn, 2022, Figure 2)

In sum, it was the tight linkage between economic and reproductive success in an expanding market economy that drove the fast cognitive advance of late medieval and post-medieval times. One could imagine a positive feedback loop: 1) as the market economy expanded, it created niches that could be filled only by individuals with a certain level of cognitive ability; 2) such individuals translated their economic success into reproductive success, thus raising mean cognitive ability; 3) as mean cognitive ability rose, the smart fraction could come up with better products and better ways of doing things, thus spurring further expansion of the market economy.

Ashkenazi Jews: a cognitive advance with similar causes

Beyond Western Europe, some cultures were more inclined than others to participate in the emerging market economy. Ashkenazi Jews, in particular, benefited from an increasingly dynamic economic environment from the late Middle Ages onward.

This is reflected in a population boom: the Ashkenazim grew in number from a little over 7 thousand in the 12th century to over 8 million by 1900. Successful traders and artisans were translating economic success into demographic success, partly because they saw procreation as a desirable goal in itself, and partly because they relied primarily on a family workforce. A successful entrepreneur was incentivized to marry earlier and have more children (Cochran et al., 2006; DellaPergola, 2001; Frost, 2022b; Murray, 2007).

This rapid population growth seems to have been associated with an equally rapid rise in mean cognitive ability. Of all human populations, Ashkenazim today have the highest frequencies of alleles associated with educational attainment, and this genetic profile corresponds closely to their IQ scores (Dunkel et al., 2019; Piffer, 2019).

This rapid cognitive evolution had an adverse side-effect: an increase in alleles associated with nine neurological disorders: Tay-Sachs (two unrelated alleles); Gaucher's (five unrelated alleles); Niemann-Pick; and Mucolipidosis Type IV. These disorders affect the capacity of neurons to store sphingolipids, which are vital to neuronal growth. Thus, nine mutations have independently reached high frequencies in the same metabolic pathway, in the same population, and in the same time period.

The proximate cause seems to be unusually strong natural selection operating through heterozygote advantage: if just one parent has the mutation, as more often happens, the child suffers no adverse effects; in fact, the child's neurons are better supplied with sphingolipids (Cochran et al., 2006; Diamond, 1994). Evidently, these alleles are a fraction of a much larger number of alleles that explain the higher cognitive ability of Ashkenazi Jews.

Figure 5: Distribution of IQ scores and alleles associated with educational attainment (PGS), Wisconsin Longitudinal Study (Dunkel et al., 2019, Figure 1)

Second cognitive retreat: the Era of Industrial Capitalism

The middle-class baby boom began to run out of steam in the late 19th century. Household workshops, where family members did the work, gave way to factories, where it was done by employees. A successful entrepreneur could simply hire more workers, instead of marrying earlier and having more children. Children thus became a net economic cost, rather than a net economic benefit — all the more so because this cost came on top of the many others now needed to maintain a middle-class lifestyle: a big home, a summer cottage, a luxury car, a college education for the kids, and so on.

For all these reasons, the middle class had fewer and fewer children. The result would be a decline not only in fertility but also in mean cognitive ability, as shown by several lines of evidence.

Alleles associated with educational attainment

We have already seen this measure. It is constructed from the human genome, specifically from alleles associated with differences in educational attainment (EA) — a good proxy for cognitive ability. 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). Although this score is still unreliable for measuring the IQ of any one individual, it is a much better measure of mean population IQ — the correlation being 90% between the two (Piffer, 2019).

The following studies concern successive age cohorts within a single population. In all three cases, the mean polygenic score fell during the 20th century from one cohort to the next.

Americans

This study used data taken from Americans shortly before and during retirement, specifically 11,822 Americans of European ancestry born between 1931 and 1953 (Beauchamp, 2016).

From 1931 to 1953, each successive age cohort had, on average, fewer alleles associated with educational attainment. The decline seemed to be due to lower fertility among those participants who had pursued higher education. The mean EA polygenic score was thus significantly higher among the childless than among individuals with one or more children.

This study, like the next two, may suffer from survivorship bias — if you manage to live to an old age, you are probably smarter than average (Gottfredson & Deary, 2004). The bias must be small, however, since 85% of the original participants were still alive in 2008, the last year of genotyping.

Figure 7: Mean EA polygenic score as a function of lifetime reproductive success (LRS) (Beauchamp, 2016, Fig. 2)

British

This study used UK Biobank data, specifically 409,629 British individuals of European origin from two successive generations. The median birth year was 1950 for the second generation and unknown for the first (Hugh-Jones & Abdellaoui, 2022).

The second generation had fewer alleles associated with educational attainment, particularly at lower incomes. Conversely, there were increases in the frequencies of alleles associated with obesity, attention deficit hyperactivity disorder, depressive disorder, schizophrenia, neuroticism, and male extraversion.

This generational change may be underestimated because the childless individuals of the first generation were necessarily excluded.

Icelanders

This study used data from a genealogical database (deCODE genetics), specifically 129,808 Icelanders born between 1910 and 1990 (Kong et al., 2017).

The mean EA polygenic score fell at a rate of about 0.010 standard deviation per decade, with one pause in the 1950s and another in the 1970s. The first pause coincided with the post-war boom and a corresponding improvement in the ability of middle-class couples to start a family early in life. The second pause might reflect a law passed in 1975 to liberalize access to abortion in cases of rape, mental disability of the mother, and “difficult family situation” (Wikipedia, 2025a)

This cognitive decline was due only in part to the more intelligent staying in school longer and postponing marriage. Some of it was independent of higher education, perhaps because smarter people — irrespective of whether they go to university — tend to plan ahead and postpone family formation until they are financially ready. Survivorship bias was not a major factor in this study, being low for the last five cohorts born since 1940 and especially low for those born in the 1970s and the 1980s.

Figure 8: Mean polygenic score of Icelanders by year of birth for alleles associated with educational attainment (Kong et al., 2017, Fig. 2)

Brain size

This measure correlates moderately with IQ (r = 0.2 to 0.4) and is obtained through MRIs, autopsies, and anthropometry (Lee et al., 2019).

Americans (Framingham Heart Study)

In Framingham, Massachusetts, three generations have been examined since 1948 to understand the causes of heart disease. In particular, brain MRIs were done each year between 1999 and 2019 on 4,506 individuals born between 1902 and 1985. For each successive decade of birth, brain volume increased by 1.7 cc for men and 1.2 cc for women (DeCarli et al., 2023).

This seems to be evidence for an increase in cognitive ability. However, the cohorts are not comparable: the older ones were mill and factory workers, while the younger ones tended to be college-educated nurses and doctors:

Requirements for entry [into the Framingham Heart Study] were an age between 30 and 62 years at the time of first examination, with no history of heart attack or stroke. Due to lukewarm interest at first, doctors, nurses and healthcare workers volunteered for the study to set an example for patients. (Wikipedia, 2025b).

The presence of these younger, better educated volunteers likely explains the correlation between brain size and decade of birth. Indeed, the proportion of college-educated participants rose by decade of birth, reaching a peak of nearly 92% for women born in 1970 (DeCarli et al., 2023).

Americans (Forensic Anthropology Data Bank)

According to data from the Forensic Anthropology Data Bank, mean brain size increased among Americans between 1820 and 1920 with no further improvement during the time of the Flynn effect (Jantz & Jantz, 2016). Was this increase due to better nutrition or less childhood disease? Infant mortality is a good proxy for both, and it did not begin to fall until the turn of the 20th century. Since we have no data from before 1820, the increase in brain size may be the tail-end of a longer-term increase that began in late medieval times with the growing contribution of the middle class to the gene pool.

British and Germans (autopsies)

Three autopsy datasets, one from the UK and two from Germany, show a rise in brain weight by year of birth from the mid-19th century to the mid-20th. In the British dataset, male brains increased by 52 grams and female brains by 23 grams. In the German dataset, male brains increased by 73 grams and female brains by 52 grams (Haug, 1984; Kretschmann et al., 1979; Miller & Corsellis, 1977; Woodley of Menie et al., 2016).

Again, this seems to be evidence for an increase in cognitive ability. Unfortunately, the age cohorts suffer from collection bias, particularly the earlier ones. During the 19th century, autopsies were done largely on charity cases and condemned criminals because of the stigma attached to cutting up a dead body: “dissection was considered a ‘degrading and sacrilegious practice’ and was something reserved as a form of extra punishment for executed criminals” (Nystrom, 2011).

Attitudes changed during the 20th century, and autopsies became more socially acceptable. The increase in brain size may therefore be due to a more representative sampling of the population.

Estonians (anthropometry)

In Soviet Estonia, cranial volume was measured for an anthropometric survey of students born between 1937 and 1962 (7,123 boys and 9,866 girls). Volunteer bias was nonexistent because the measurements were mandatory, and mortality bias was minimal because the participants were all younger than 20 (Hõrak & Valge, 2015; Valge et al., 2021; Valge et al., 2022).

Girls with larger crania were more likely to pursue higher education. They thus married later and had fewer children. Throughout the 20th century, Estonian women with only primary education bore 0.5 to 0.75 more children on average than women with tertiary education. In contrast, higher education had no effect on male fertility. The most fertile men had average-sized crania.

Reaction time

This measure is the speed at which the brain processes information. It has a moderate to high correlation with IQ (Der & Deary, 2017).

We seem to be taking longer to process the same amount of information. This is the conclusion that emerged when 14 studies, published since 1941, were compared with a study by Francis Galton from the late 19th century. With one exception, the post-1941 studies showed a lengthening of reaction time since Galton’s study (Silverman, 2010). The corresponding decline in IQ is estimated at 1.23 points per decade, for a total loss of 14 points since the Victorian era (Woodley et al., 2013).

Admittedly, the decline may be due to a more representative sampling of the population by later studies (hbd*chick, 2013). On the other hand, a similar lengthening of reaction time has been shown by controlled studies of Swedish, Scottish, and American participants, especially for cohorts born since 1980. The corresponding decline in IQ is likewise estimated at 1.3 to 1.7 points per decade (Madison, 2014; Madison et al., 2016).

Figure 9: Lengthening of reaction time of Swedish participants. The red squares are adjusted for participant age (Madison et al., 2016)

Piagetian testing

This measure is highly correlated with IQ (r=0.51). Both are measures of speed, working memory, and complex reasoning (Rindermann & Ackermann, 2020).

When a research team examined the results of Piagetian tests in England and Wales from 1975-1976 to 2006-2007, they found that test scores began to fall after 1993, at a time when conventional IQ test scores were still rising. Moreover, the smart fraction was shrinking faster than the proportion of students who were simply above average:

The Piagetian results are particularly ominous … the pool of those who reach the top level of cognitive performance is being decimated: fewer and fewer people attain the formal level at which they can think in terms of abstractions and develop their capacity for deductive logic and systematic planning. (Flynn & Shayer, 2018)

And yet IQ scores were rising!

Mean IQ nonetheless rose throughout the 20th century. This upward trend was pointed out by intelligence researcher James Flynn (1984), who calculated a rise of 13.8 points between 1932 and 1978 among European Americans. Named the Flynn effect in his honor, it was charted from 1909 to 2013 by Pietschnig and Voracek (2015), who found a total rise of 35 points. It can be broken down into five periods:

• a small increase between 1909 and 1919 (0.80 points/decade)

• a surge during the 1920s and early 1930s (7.2 points/decade)

• a slower pace of growth between 1935 and 1947 (2.1 points/decade)

• a faster one between 1948 and 1976 (3.0 points/decade)

• a slower pace thereafter (2.3 points/decade)

The Flynn effect is now ending throughout the Western world. In fact, it has ended altogether in Norway and Sweden and has begun to reverse in Denmark and Finland (Pietschnig & Voracek, 2015).

How real was this increase? As a child, I knew people who had gone to school before IQ scores began to rise and who, presumably, were more than a standard deviation less intelligent than my generation. I also knew the textbooks they had read, which our elementary school kept in a storeroom. How could they have understood those books if they had been so lacking in intelligence? How could they have handled the dense subject matter, the vocabulary, the detailed charts and figures, and the long, complex sentences?

The Flynn effect is more likely an increase in familiarity with standardized written tests, due to students staying in school longer and teachers abandoning other methods of evaluation, like essays and orally administered tests. Remember, almost a third of the total Flynn effect occurred during the 1920s and early 1930s. This was when IQ testing, and standardized written tests in general, became much more widespread in American schools:

World War I, in effect, set in motion the process that would result — in an incredibly short time — in national intelligence testing for American school children. By the end of the first decade after the war, standardized educational testing was becoming a fixture in the schools. (U.S. Congress, 1992, pp. 121-122)

After World War II, standardized written tests became even more widely used as more students went to college and university. This rise in postsecondary enrolment levelled off not long after the turn of the millennium — at about the same time that the Flynn effect levelled off.

Thoughts on the Great Decline

The end of the Flynn effect is showing us the real long-term trend of the 20th century: a decline in the genetic capacity for intelligence. This is most clearly shown by the genome studies from the United States, the United Kingdom, and Iceland, all of which show a decline in alleles associated with educational attainment.

The main cause seems to be the tendency of the more intelligent to pursue higher education and thus postpone marriage and childbearing: “In all countries [Australia, United States, Norway, Sweden], however, education is negatively associated with childbearing across partnerships, and the differentials increased from the 1970s to the 2000s” (Thomson et al., 2014). Yet that is not the whole story. The Icelandic genome study found lower fertility among the more intelligent even after controlling for higher education.

The resulting cognitive decline would become measurable by IQ tests only after the population had become fully familiar with such tests. But it had been happening all along.

This view is not shared by everyone. In a study of Norwegian siblings, Bratsberg and Rogeberg (2018) found that the recent reversal of the Flynn effect in Norway can be explained by “within-family variation.” In other words, IQ is declining more in younger siblings than in older ones. Since siblings share the same genetic background, the cause must be environmental.

But siblings in Norway are increasingly half-siblings. Among Norwegian women with only two children, 13.4% have had them by more than one man. The figure rises to 24.9% among those with three children, 36.2% among those with four children, and 41.2% among those with five children (Thomson et al., 2014). If siblings have different fathers, they cannot share the same genetic background. Keep in mind that the Bratsberg and Rogeberg study was limited, by its very nature, to families with at least two children and was skewed toward those with even more.

With what sort of men do Norwegian single mothers have more children? The sort of polygynous, low-IQ men that single mothers often end up with: “At the lowest educational level, 19.3 percent of those [Norwegian men] who had become fathers had children with more than one woman, compared to 6.1 percent of those at the highest educational level” (Lappegård et al., 2011).

Thus ends a trajectory that took northwest Europeans from relative unimportance to global dominance. It began in late medieval times with an emerging middle class that catered to an emerging market economy. These entrepreneurs relied on a family workforce and, if successful, would expand it by having more children. As their lineages became more numerous with each generation, the population shifted toward higher cognitive ability, stronger impulse control, and greater foresight. In this way, the West remade itself in the image of its growing middle class (Clark, 2007, 2009, 2023; Frost, 2022a; Frost, 2024; Piffer & Kirkegaard, 2024).

That evolution began to stall in the late 19th century, as middle-class couples ceased to translate their economic success into reproductive success. “Ma and pa” shops gave way to modern businesses, which could more readily increase or reduce their workforce by hiring or firing as needed. Fertility thus fell into a decline that would momentarily reverse during the post-war baby boom, when the fruits of economic growth were widely distributed and when cultural messaging was much more pro-family than it is today (Frost, 2019b; Frost, 2022b).

The baby boom then gave way to a baby bust and then a “family bust.” Today, the family unit is less concerned with perpetuating a lineage, and more with serving the interests of its current members.

Overview of European cognitive evolution

References

Akbari, A., Barton, A.R., Gazal, S., Li, Z., Kariminejad, M., Perry, A., Zeng, Y., Mittnik, A., Patterson, N., Mah, M., Zhou, X., Price, A.L., Lander, E.S., Pinhasi, R., Rohland, N., Mallick, S., & Reich, D. (2024). Pervasive findings of directional selection realize the promise of ancient DNA to elucidate human adaptation. bioRxiv. https://doi.org/10.1101/2024.09.14.613021

Angel, J.L. (1972). Biological relations of Egyptian and eastern Mediterranean populations during Pre-dynastic and Dynastic times. Journal of Human Evolution, 1(3), 307-313. https://doi.org/10.1016/0047-2484(72)90066-8

Augustine. (1968). The Fathers of the Church. Saint Augustine, vol. 60. The Retractations. Washington D.C.: The Catholic University of America Press.

Balloux F., Handley, L.J., Jombart, T., Liu, H., & Manica, A. (2009). Climate shaped the worldwide distribution of human mitochondrial DNA sequence variation. Proceedings of the Royal Society B. Biological Sciences, 276, 3447-3455. https://doi.org/10.1098/rspb.2009.0752

Beals, K.L., Smith, C.L., Dodd, S.M., Angel, J.L., Armstrong, E., Blumenberg, B., Girgis, F.G., Turkel, S., Gibson, K.R., Henneberg, M., et al. (1984). Brain size, cranial morphology, climate, and time machines. Current Anthropology, 25, 301–330. https://psycnet.apa.org/doi/10.1086/203138

Beauchamp, J.P. (2016). Genetic evidence for natural selection in humans in the contemporary United States. Proceedings of the National Academy of Sciences, USA, 113, 7774-7779. https://doi.org/10.1073/pnas.1600398113

Benedict, R. (1946 [2005]). The Chrysanthemum and the Sword. Patterns of Japanese Culture. First Mariner Books.

Brace, C.L., Seguchi, N., Quintyn, C.B., Fox, S.C., Nelson, A.R., Manolis, S.K., & Qifeng, P. (2006). The questionable contribution of the Neolithic and the Bronze Age to European craniofacial form. Proceedings of the National Academy of Sciences, USA, 103, 242-247. https://doi.org/10.1073/pnas.0509801102

Bratsberg, B., & Rogeberg, O. (2018). Flynn effect and its reversal are both environmentally caused. Proceedings of the National Academy of Sciences, USA, 115, 6674-6678. https://doi.org/10.1073/pnas.1718793115

Caesar, J. (1915). De Bello Gallico & Other Commentaries of Caius Julius Caesar (Translated by W.A. Macdevitt). London: J.M. Dent.

Caldwell, J.C. (2004). Fertility control in the classical world: Was there an ancient fertility transition? Journal of Population Research, 21, 1. https://doi.org/10.1007/BF03032208

Clark, G. (2007). A Farewell to Alms. A Brief Economic History of the World. Princeton University Press: Princeton. https://press.princeton.edu/books/paperback/9780691141282/a-farewell-to-alms

Clark, G. (2009). The domestication of man: the social implications of Darwin. ArtefaCToS, 2, 64-80. https://www.researchgate.net/publication/277275046_The_Domestication_of_Man_The_Social_Implications_of_Darwin

Clark, G. (2023). The inheritance of social status: England, 1600 to 2022. Proceedings of the National Academy of Sciences, U.S.A., 120(27), e2300926120 https://doi.org/10.1073/pnas.2300926120

Cochran, G., Hardy, J., & Harpending, H. (2006). Natural history of Ashkenazi intelligence. Journal of Biosocial Science, 38(5), 659-693. https://doi.org/10.1017/S0021932005027069

Cochran, G. & Harpending, H. (2009). The 10,000 Year Explosion: How Civilization Accelerated Human Evolution. Basic Books: New York. https://www.amazon.ca/000-Year-Explosion-Civilization-Accelerated/dp/0465020429

DeCarli, C., Pase, M., Beiser, A., Kojis, D., Satizabal, C., et al. (2023). Secular Trends in Head Size and Cerebral Volumes in the Framingham Heart Study for Birth Years 1902-1985. Research Square, January 30. https://doi.org/10.21203/rs.3.rs-2524684/v1

de Courson, B., Thouzeau, V., & Baumard, N. (2023). Quantifying the scientific revolution. Evolutionary Human Sciences, 5, E19. https://doi.org/10.1017/ehs.2023.6

De Rosa, L. (1979). Property Rights, Institutional Change, and Economic Growth in Southern Italy in the XVIIIth and XIXth Centuries. Journal of European Economic History, 8(3), 531-551.

DellaPergola, S. (2001). Some fundamentals of Jewish Demographic History. In: S. DellaPergola, & J. Even (eds.) Papers in Jewish Demography 1997, (pp. 11-33), Jerusalem: The Hebrew University.

Der, G., & Deary, I.J. (2017). The relationship between intelligence and reaction time varies with age: Results from three representative narrow-age age cohorts at 30, 50 and 69 years. Intelligence, 64, 89-97. https://doi.org/10.1016/j.intell.2017.08.001

Diamond, J.M. (1994). Jewish Lysosomes. Nature, 368, 291-292. https://doi.org/10.1038/368291a0

Dunkel, C.S, Woodley of Menie, M.A., Pallesen, J., & Kirkegaard, E.O.W. (2019). Polygenic scores mediate the Jewish phenotypic advantage in educational attainment and cognitive ability compared with Catholics and Lutherans. Evolutionary Behavioral Sciences, 13(4), 366-375. https://psycnet.apa.org/doi/10.1037/ebs0000158

Flynn, J.R. (1984). The mean IQ of Americans: Massive gains 1932–1978. Psychological Bulletin, 95(1), 29–51. https://psycnet.apa.org/doi/10.1037/0033-2909.95.1.29

Flynn, J.R. & M. Shayer. (2018). IQ decline and Piaget: Does the rot start at the top? Intelligence, 66, 112-121. https://doi.org/10.1016/j.intell.2017.11.010

Frost, P. (2010). The Roman State and genetic pacification. Evolutionary Psychology, 8(3), 376-389. https://doi.org/10.1177%2F147470491000800306

Frost, P. (2015). The Past is Another Country. The Unz Review, August 15. https://www.unz.com/pfrost/the-past-is-another-country/

Frost, P. (2017). The Hajnal line and gene-culture coevolution in northwest Europe. Advances in Anthropology, 7, 154-174. https://doi.org/10.4236/aa.2017.73011

Frost, P. (2019a). The Original Industrial Revolution. Did Cold Winters Select for Cognitive Ability? Psych, 1(1), 166-181. https://doi.org/10.3390/psych1010012

Frost, P. (2019b). Demise of the West. Evo and Proud, January 7. https://evoandproud.blogspot.com/2019/01/demise-of-west.html

Frost, P. (2020). The large society problem in Northwest Europe and East Asia. Advances in Anthropology, 10(3), 214-134. https://doi.org/10.4236/aa.2020.103012

Frost, P. (2022a). When did Europe pull ahead? And why? Peter Frost’s Newsletter, November 21. https://www.anthro1.net/p/when-did-europe-pull-ahead-and-why

Frost, P. (2022b). Ashkenazi Jews and recent cognitive evolution. Peter Frost’s Newsletter, December 5. https://www.anthro1.net/p/ashkenazi-jews-and-recent-cognitive

Frost, P. (2022c). Europeans and recent cognitive evolution. Peter Frost’s Newsletter, December 12. https://www.anthro1.net/p/europeans-and-recent-cognitive-evolution

Frost, P. (2023). 1. Genetic pacification in Western Europe from late medieval to early modern times. My wish list for research in 2024. Peter Frost’s Newsletter, December 19. https://www.anthro1.net/p/1-genetic-pacification-in-western

Frost, P. (2024). Cognitive evolution in Europe: Two new studies. Peter Frost’s Newsletter, March 14. https://www.anthro1.net/p/cognitive-evolution-in-europe-two

Frost P., & Harpending, H. (2015). Western Europe, state formation, and genetic pacification. Evolutionary Psychology, 13(1), 230-243. https://doi.org/10.1177%2F147470491501300114

Gottfredson, L.S., & Deary, I.J. (2004). Intelligence Predicts Health and Longevity, but Why? Current Directions in Psychological Science, 13(1), 1–4. https://doi.org/10.1111/j.0963-7214.2004.01301001.x

Greer, T. (2013a). The Rise of the West: Asking the Right Questions. July 7, The Scholar's Stage. https://scholars-stage.org/the-rise-of-the-west-asking-the-right-questions/

Greer, T. (2013b). Another look at the 'Rise of the West' - but with better numbers. November 20. The Scholar's Stage. https://scholars-stage.org/another-look-at-the-rise-of-the-west-but-with-better-numbers/

Hajnal, J. (1965). European marriage pattern in perspective. In: D.V. Glass and D.E.C. Eversley (eds). Population in History. Essays in Historical Demography. London, Arnold.

Hallam, H.E. (1985). Age at First Marriage and Age at Death in the Lincolnshire Fenland, 1252-1478. Population Studies, 39, 55-69. https://doi.org/10.1080/0032472031000141276

Harper, K. (2011). Slavery in the Late Roman World, AD 275-425. Cambridge University Press.

Harris, M. (1959). The economy has no surplus. American Anthropologist, 61(2), 185-199. https://doi.org/10.1525/aa.1959.61.2.02a00010

Harris, W. (1999). Demography, Geography and the Sources of Roman Slaves. Journal of Roman Studies, 89, 62-75. https://doi.org/10.2307/300734

Hartman, M.S. (2004). The Household and the Making of History. A Subversive View of the Western Past. Cambridge, Cambridge University Press. https://doi.org/10.1017/CBO9780511818134

Haug, H. (1984). Der Einfluß der säkularen Acceleration auf das Hirngewicht des Menschen und dessen Änderung während der Alterung. Gegenbaurs Morphologisches Jahrbuch, 130, 481–500

Hawks, J. (2011). Selection for smaller brains in Holocene human evolution. arXiv: 1102.5604 [q-bio.PE]. https://doi.org/10.48550/arXiv.1102.5604

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, USA, 104, 20753-20758. https://doi.org/10.1073/pnas.0707650104

hbd*chick (2013). A response to a response to two critical commentaries on woodley, te nijenhuis and murphy. May 27. http://hbdchick.wordpress.com/2013/05/27/a-response-to-a-response-to-two-critical-commentaries-on-woodley-te-nijenhuis-murphy-2013/

hbd*chick (2014). Big summary post on the Hajnal Line. October 3. https://hbdchick.wordpress.com/2014/03/10/big-summary-post-on-the-hajnal-line/

Hoffecker, J.F. (2002). Desolate Landscapes. Ice-Age Settlement in Eastern Europe. New Brunswick: Rutgers University Press.

Hopkins, K. (1965). Contraception in the Roman Empire. Comparative Studies in Society and History, 8(1), 124-151. https://doi.org/10.1017/S0010417500003935

Hõrak, P., & Valge, M. (2015). Why did children grow so well at hard times? The ultimate importance of pathogen control during puberty. Evolution, Medicine, and Public Health, 1, 167–178. https://doi.org/10.1093/emph/eov017

Hugh-Jones, D., & Abdellaoui, A. (2022). Human Capital Mediates Natural Selection in Contemporary Humans. Behavior Genetics, 52, 205-234. https://doi.org/10.1007/s10519-022-10107-w

ICA (2020). Research Themes—Marriage Patterns. Institutions for Collective Action. https://web.archive.org/web/20190329070516/http://www.collective-action.info/_THE_MarriagePatterns_EMP

Jantz, R.L., & Jantz, L.M. (2016). The remarkable change in Euro-American cranial shape and size. Human Biology, 88(1), 56-64. https://doi.org/10.13110/humanbiology.88.1.0056

Jensen, S., & Kirkegaard, E.O.W. (2024). National IQs and Socioeconomic Development. PsyArXiv; June 7. https://doi.org/10.31234/osf.io/bx86g

Kirkegaard, E.O.W. (2022). Cold winters theory: a summary of the evidence and replies to objections. Just Emil Kirkegaard Things, December 28. https://www.emilkirkegaard.com/p/cold-winters-theory-a-summary-of

Kong, A., Frigge, M.L., Thorleifsson, G., Stefansson, H., Young, A.I., Zink, F., Jonsdottir, G.A., Okbay, A., Sulem, P., Masson, G., Gudbjartsson, D.F., Helgason, A., Bjornsdottir, G., Thorsteinsdottir, U., & Stefansson, K. (2017). Selection against variants in the genome associated with educational attainment. Proceedings of the National Academy of Sciences USA, 114(5), E727-E732. https://doi.org/10.1073/pnas.1612113114

Kretschmann, H.-J., Schleicher, A., Wingert, F., Zilles, K., & Löblich, K.-J. (1979). Human brain growth in the 19th and 20th century. Journal of the Neurological Sciences, 40, 169–188. https://doi.org/10.1016/0022-510X(79)90202-8

Kuijpers, Y., Domínguez-Andrés, J., Bakker, O.B., Gupta, M.K., Grasshoff, M., Xu, C.J., Joosten, L.A.B., Bertranpetit, J., Netea, M.G., & Li, Y. (2022). Evolutionary Trajectories of Complex Traits in European Populations of Modern Humans. Frontiers in Genetics, 13, 833190. https://doi.org/10.3389/fgene.2022.833190

Lappegård, T., Rønsen, M., & Skrede, K. (2011). Fatherhood and fertility. Fathering, 9(1), 103-120. https://doi.org/10.3149/fth.0901.103

Lee, J.J., McGue, M., Iacono, W.G., Michael, A.M., & Chabris, C.F. (2019). The causal influence of brain size on human intelligence: Evidence from within-family phenotypic associations and GWAS modeling. Intelligence, 75, 48-58. https://doi.org/10.1016%2Fj.intell.2019.01.011

Lee, J. J., Wedow, R., Okbay, A., Kong, E., Maghzian, O., Zacher, et al. (2018). Gene discovery and polygenic prediction from a genome-wide association study of educational attainment in 1.1 million individuals. Nature Genetics, 50(8), 1112-1121. https://doi.org/10.1038/s41588-018-0147-3

MacDonald, K. (2019). Individualism and the Western Liberal Tradition: Evolutionary Origins, History, and Prospects for the Future. Amazon. https://www.amazon.ca/Individualism-Western-Liberal-Tradition-Evolutionary/dp/1089691483

Madison, G. (2014). Increasing simple reaction times demonstrate decreasing genetic intelligence in Scotland and Sweden. London Conference on Intelligence. Psychological Comments, April 25, #LCI14 Conference proceedings. http://www.unz.com/jthompson/lci14-questions-on-intelligence/

Madison, G., Woodley of Menie, M.A., & Sänger, J. (2016). Secular Slowing of Auditory Simple Reaction Time in Sweden (1959-1985). Frontiers in Human Neuroscience, August 18. https://doi.org/10.3389/fnhum.2016.00407

McCann, C.A. (2010). Transgressing the Boundaries of Holiness: Sexual Deviance in the Early Medieval Penitential Handbooks of Ireland, England and France 500-1000. Theses, South Orange, NJ: Seton Hall University. https://scholarship.shu.edu/theses/76

Melchior, L., Lynnerup, N., Siegismund, H.R., Kivisild, T., & Dissing, J. (2010). Genetic diversity among ancient Nordic populations, PLoS ONE, 5(7), e11898. https://doi.org/10.1371/journal.pone.0011898

Melleno, D. (2014). North Sea networks: trade and communication from the seventh to the tenth century. Comitatus: A Journal of Medieval and Renaissance Studies, 45, 65-89. https://doi.org/10.1353/cjm.2014.0055

Miller, A.K.H., & Corsellis, J.A.N. (1977). Evidence for a secular increase in human brain weight during the past century. Annals of Human Biology, 4, 253–257. https://doi.org/10.1080/03014467700007142

Murray, C. (2007). Jewish Genius. Commentary, April. https://www.commentary.org/articles/charles-murray/jewish-genius/

North, J. (2013). The development of religious pluralism. In: The Jews among Pagans and Christians in the Roman Empire (pp. 174-193). Routledge.

Nystrom, K.C. (2011). Postmortem examinations and the embodiment of inequality in 19th century United States. International Journal of Paleopathology, 1(3-4), 164-172. https://doi.org/10.1016/j.ijpp.2012.02.003

Oesterdiekhoff, G.W. (2012). Was pre-modern man a child? The quintessence of the psychometric and developmental approaches. Intelligence 40, 470–478. https://doi.org/10.1016/j.intell.2012.05.005

O'Gorman, R., Wilson, D.S., & Miller, R.R. (2008). An evolved cognitive bias for social norms. Evolution and Human Behavior, 29(2), 71-78. https://doi.org/10.1016/j.evolhumbehav.2007.07.002

Okbay, A., Wu, Y., Wang, N., Jayashankar, H., Bennett, M., Nehzati, S.M., Sidorenko, J., Kweon, H., Goldman, G., Gjorgjieva, T., Jiang, Y., Hicks, B., Tian, C., Hinds, D.A., Ahlskog, R., Magnusson, P.K.E., Oskarsson, S., Hayward, C., Campbell, A., Porteous, D.J., … Young, A. I. (2022). Polygenic prediction of educational attainment within and between families from genome-wide association analyses in 3 million individuals. Nature Genetics, 54, 437–449. https://doi.org/10.1038/s41588-022-01016-z

Origen. (1982). Origène. Commentaire sur Saint-Jean. Tome IV (Livres XIX et XX). Paris: Les Éditions du Cerf.

Pang-White, A.A. (2001). The fall of humanity: Weakness of the will and moral responsibility in the later Augustine. Medieval Philosophy and Theology, 9, 51-67. https://doi.org/10.5840/medievalpt2000914

Perry, M.J. (2013). Gender, Manumission, and the Roman Freedwoman. Cambridge University Press.

Pietschnig, J., & Voracek, M. (2015). One Century of Global IQ Gains: A Formal Meta-Analysis of the Flynn Effect (1909-2013). Perspectives on Psychological Science, 10(3), 282-306. https://doi.org/10.1177/1745691615577701

Piffer, D. (2019). Evidence for Recent Polygenic Selection on Educational Attainment and Intelligence Inferred from Gwas Hits: A Replication of Previous Findings Using Recent Data. Psych, 1(1), 55-75. https://doi.org/10.3390/psych1010005

Piffer, D., & Connor, G. (2025). Genomic Evidence for Clark's Theory of the British Industrial Revolution, preprint, ResearchGate, June. https://www.researchgate.net/publication/392808200_Genomic_Evidence_for_Clark's_Theory_of_the_British_Industrial_Revolution

Piffer D, Dutton, E., & Kirkegaard, E.O.W. (2023). Intelligence Trends in Ancient Rome: The Rise and Fall of Roman Polygenic Scores. OpenPsych. Published online July 21, 2023. https://doi.org/10.26775/OP.2023.07.21

Piffer, D., & Kirkegaard, E.O.W. (2024). Evolutionary Trends of Polygenic Scores in European Populations from the Paleolithic to Modern Times. Twin Research and Human Genetics, 27(1), 30-49. https://doi.org/10.1017/thg.2024.8

Piffer, D., & Lynn, R. (2022). In Italy, North-South Differences in Student Performance Are Mirrored by Differences in Polygenic Scores for Educational Attainment. Mankind Quarterly, 62(4), Article 2. https://doi.org/10.46469/mq.2022.62.4.2

Policy Tensor. (2021). The Church’s crusade against cousin-marriage did not create the Western nuclear family. May 7.

Price, T. D. (1991). The Mesolithic of Northern Europe. Annual Review of Anthropology, 20, 211-233. https://doi.org/10.1146/annurev.an.20.100191.001235

Rawson, B. (1986). Children in the Roman familia. In: B. Rawson (ed.) The Family in Ancient Rome: New Perspectives, (pp. 170-200). Ithaca (New York): Cornell University Press.

Rinaldi, A. (2017). We're on a road to nowhere. Culture and adaptation to the environment are driving human evolution, but the destination of this journey is unpredictable. EMBO reports, 18, 2094-2100. https://doi.org/10.15252/embr.201745399

Rindermann, H. (2018). Cognitive Capitalism. Human Capital and the Wellbeing of Nations, 1st ed.; Cambridge University Press.

Rindermann, H., & Ackermann, A.L. (2021). Piagetian Tasks and Psychometric Intelligence: Different or Similar Constructs? Psychological Reports, 124(6), 2795-2821. https://doi.org/10.1177/0033294120965876

Roetzel, C.J. (2000). Sex and the single god: celibacy as social deviancy in the Roman period. In: S.G. Wilson & M. Desjardins (eds). Text and Artefact in the Religions of Mediterranean Antiquity. Essays in Honour of Peter Richardson (pp. 231-248), Wilfrid Laurier University Press.

Sasaki, J.Y., Kim, H.S., Mojaverian, T., Kelley, L.D. S., Park, I.Y., & Janušonis, S. (2013). Religion Priming Differentially Increases Prosocial Behavior among Variants of the Dopamine D4 Receptor (DRD4) Gene. Social Cognitive and Affective Neuroscience, 8, 209-215. https://doi.org/10.1093/scan/nsr089

Schulz, J.F., Bahrami-Rad, D., Beauchamp, J.P., & Henrich, J. (2019). The Church, intensive kinship, and global psychological variation. Science, 366(707), 1-12. https://doi.org/10.1126/science.aau5141

Seccombe, W. (1992). A Millennium of Family Change. Feudalism to Capitalism in Northwestern Europe. London: Verso. https://archive.org/details/millenniumoffami0000secc/page/n3/mode/2up

Silverman, I.W. (2010). Simple reaction time: It is not what it used to be. The American Journal of Psychology, 123(1), 39-50. https://doi.org/10.5406/amerjpsyc.123.1.0039

Sullivan, V. (2009). Increasing Fertility in the Roman Late Republic and Early Empire. MA thesis, History, North Carolina State University. https://repository.lib.ncsu.edu/handle/1840.16/812

Tacitus (1970). Agricola, Germania, Dialogus. Loeb Classical Library (Translated by M. Hutton & W. Peterson). Cambridge, MA: Harvard University Press.

Testart, A., Forbis, R. G., Hayden, B., Ingold, T., Perlman, S. M., Pokotylo, D. L., ... & Stuart, D. E. (1982). The significance of food storage among hunter-gatherers: Residence patterns, population densities, and social inequalities [and comments and reply]. Current anthropology, 23(5), 523-537. https://doi.org/10.1086/202894

Thompson, G.L. (2008). Works of Constantine I (the Great). Fourth-Century Christianity. Wisconsin Lutheran College. https://www.fourthcentury.com/works-of-constantine/

Thomson, E., Lappegård, T., Carlson, M., Evans, A., & Gray, E. (2014). Childbearing across partnerships in Australia, the United States, Norway, and Sweden. Demography, 51(2), 485-508. https://doi.org/10.1007/s13524-013-0273-6

U.S. Congress, Office of Technology Assessment. (1992). Testing in American Schools: Asking the Right Questions, OTA-SET-519, Washington D.C.: U.S. Government Printing Office.

Valge, M., Hõrak, P.. & Henshaw, J.M. (2021). Natural selection on anthropometric traits of Estonian girls. Evolution and Human Behavior, 42(2), 81-90. https://doi.org/10.1016/j.evolhumbehav.2020.07.013

Valge, M., Meitern, R., & Hõrak, P. (2022). Sexually antagonistic selection on educational attainment and body size in Estonian children. Annals of the New York Academy of Sciences, 1516(1), 271-285. https://doi.org/10.1111/nyas.14859

Wikipedia. (2025a). Abortion in Iceland. https://en.wikipedia.org/wiki/Abortion_in_Iceland

Wikipedia. (2025b). Framingham Heart Study. https://en.wikipedia.org/wiki/Framingham_Heart_Study

Woodley, M.A., Nijenhuis, J., & Murphy, R. (2013). Were the Victorians cleverer than us? The decline in general intelligence estimated from a meta-analysis of the slowing of simple reaction time. Intelligence, 41, 843-850. https://doi.org/10.1016/j.intell.2013.04.006

Woodley of Menie, M.A., Peñaherrera, M.A., Fernandes, H.B.F., Becker, D., & Flynn, J.R. (2016). It's getting bigger all the time: Estimating the Flynn effect from secular brain mass increases in Britain and Germany. Learning and Individual Differences, 45, 95-100. https://doi.org/10.1016/j.lindif.2015.11.004