Why did NW Europeans become WEIRD?

Ancient genomes show that cousin marriage was rare before the church bans

Joseph Henrich published two books based on the Hajnal line or WEIRD (Western Educated, Industrialized, Rich, Democratic) personality cluster idea (The Secret of Our Success 2016; The WEIRDest People in the World 2020). The model is this:

1. Low inbreeding, chiefly in form of low rates of cousin mating, decreases genetic relationships between distant kin, making it less genetically sensible to favor distant kin over non-kin (reducing the strength of kin selection in comparison with reciprocal altruism).

2. Favoring kin over non-kin is a kind of corruption for those in public office (courts, police, bureaucrats), so less inbred populations become less corrupt.

3. This decline of cousin inbreeding was due to the influence of the Christian church's bans.

I think (1-2) are mostly right, but (3) are mostly incorrect. First, let's look at the global pattern of inbreeding (per Wikipedia).

These are modern values, not historical values. However, inbreeding's genetic effects linger for some time in the gene pool until broken apart by recombination. What the above shows is that Muslims are high, as well as Indians (grey = no data). If we had genetic data for all countries, it would be possible to estimate the current level of inbreeding by population. Note, however, that this is tricky because small (effective) population size also causes inbreeding aside from cultural factors (if your 2nd cousin is the only woman of correct age around, you will choose her even if you prefer not to).

For good measure, here's a map of corruption (perception index) in 2023:

If you want a one variable explanation it's genetic distance to Denmark (large outliers: Chile, Uruguay, Japan and S. Korea). Or in the case of the "two Italies":

The evidence for the influence of the Christian church looks like this:

Top right plot shows years of Western church influence and current levels of cousin marriage.

There is a problem however. There is historical evidence that NW Europeans, specifically Germanic tribes, were already WEIRD before any influence of Christianity. Here's Roman historian Tacitus's Germania in about 98 AD, as summarized by Wikipedia:

In chapter 7, Tacitus describes their government and leadership as somewhat merit-based and egalitarian, with leadership by example rather than authority, and punishments are carried out by the priests. He mentions (chapter 8) that the opinions of women are given respect. In chapter 11, Tacitus describes a form of folk assembly rather similar to the public things recorded in later Germanic sources: in these public deliberations, the final decision rests with the men of the group as a whole.

Tacitus further discusses the role of women in chapters 7 and 8, mentioning that they often accompany the men to battle and offer encouragement. He says that the men are often motivated to fight for the women because of an extreme fear of losing them to captivity. Tacitus says (chapter 18) that the Germanic peoples are mainly content with one wife, except for a few political marriages, and specifically and explicitly compares this practice favorably to other cultures. He also records (chapter 19) that adultery is very rare, and that an adulterous woman is shunned afterward by the community regardless of her beauty. In chapter 45, Tacitus mentions that the people to the north of the Germanic peoples, the Sitones, "resemble [the Suevi Scandinavians] in all respects but one - woman is the ruling sex."^[4] "This," Tacitus comments, "is the measure of their decline, I will not say below freedom, but even below decent slavery."^[4]

Various other ancient sources describe similar things. This is about 1400 years before the medieval Christian church started the bans. For more discussion on ancient sources, Kevin MacDonald (yes, I know) wrote two criticisms of Henrich and colleagues work and there's another by Ricardo Duchesne. Their papers (21, 44, 52 pages long) cite various ancient sources showing that WEIRD personality seems to predate Christianity. MacDonald's more speculative counter-thesis is that WEIRD personality had something to do with the mixture of ancient Europeans (hunter-gatherers, farmers, Indo-Europeans). I am not an expert in ancient history, so read them and be your own judge.

Ancient historical sources can probably never give us definitive answers, but there is a method that can: genomics. Just like ancient genomes have conclusively proven and disproven various archaeological theories (Kossinna was right), so perhaps it can do for the origins of WEIRD models. The basics of the genetics of inbreeding is that when you marry your relatives, you will end up with 2 identical copies of long segments (strings of letters) of the genome. They are called runs of homozygosity, ROH, since a person will have no genetic variation for long stretches). This happens because your shared ancestor(s) passed that stretch of DNA along to both of you. Since recombination breaks up DNA stretches at random, the length and quantity of these overlaps tell us how long ago the inbreeding occurred. An illustrative example:

Thus, for this hypothetical case of cousin marriage and one pair of chromosomes, we see that the offspring of this marriage acquired two stretches of identical DNA on both of their chromosome copies (dark blue and dark purple). The reason inbreeding leads to disease is that most genetic disorders are recessive, that is, having 1 copy of a bad mutation does not cause (much) disease, but having 2 copies causes the disease. Since it is quite unlikely to acquire 2 rare mutations in the same gene by chance, when this occurs, it often results from relatively recent inbreeding (say, last 5 generations).

With high quality genetic data, finding all the ROHs in the genome is fairly easy and has been done at scale for modern worldwide populations. The results show that our estimates of cousin marriage are broadly accurate:

The rate of long ROH is substantially higher in the present-day Human Origins dataset; we inferred that 176 of 1941 modern individuals (9.1%, CI: 7.8–10.4%) have long ROH. In contrast to ancient data, several geographic clusters of long ROH are found, mainly in present-day Near East, North Africa, Central/South Asia, and South America (Supplementary Data 1). This signal was described previously [reviewed in ref. 2] and mirrors the estimated prevalence of cousin marriages1.

Unfortunately, they don't provide a map and also refuse to be precise in their wording. As mentioned above, one major complication is that effective population size causes inbreeding independently of any cousin marriage customs. For this reason, when a population is observed to have many ROHs, it is not immediately clear whether this is due to small population size (many distant relationships between mates) or cousin etc. marriages. One has to look at the distribution of the ROHs. ROHs that happened a longer time ago with loose but consistent inbreeding tend to be many, small ROHs. More recent inbreeding results in a different distribution, with larger segments. Thus, in theory with good data, it is possible to distinguish them.

A number of studies have examined ancient genomes to assess runs of homozygosity. Ancient genome data is usually sparse because the DNA has decayed and not every letter can be read reliably. This complicates matters because errors in reading cause an apparently break in a ROH segment, and thus have to be accounted for. I found at least one study that developed a method (2019 ROHan, authors like LOTR surely; also there's a 2022 method but wasn't applied to ancient data) to deal with this based on Bayesian hidden Markov models, but it wasn't applied recently as far as I can tell. However, this 2021 paper analyzed ~1800 ancient genomes and found:

Parental relatedness of present-day humans varies substantially across the globe, but little is known about the past. Here we analyze ancient DNA, leveraging that parental relatedness leaves genomic traces in the form of runs of homozygosity. We present an approach to identify such runs in low-coverage ancient DNA data aided by haplotype information from a modern phased reference panel. Simulation and experiments show that this method robustly detects runs of homozygosity longer than 4 centimorgan for ancient individuals with at least 0.3 × coverage. Analyzing genomic data from 1,785 ancient humans who lived in the last 45,000 years, we detect low rates of first cousin or closer unions across most ancient populations. Moreover, we find a marked decay in background parental relatedness co-occurring with or shortly after the advent of sedentary agriculture. We observe this signal, likely linked to increasing local population sizes, across several geographic transects worldwide.

So overall inbreeding has been declining and was higher in the past. We knew from historical records. However, this was also true in ancient times. The advent of farming tells us that this likely had a lot to do with effective population size increasing, not with customs changing as such. But moreover, they found that close inbreeding, such as cousins, was also rare in the past. Cousin marriage is a recent thing, not ancient: "we find that only 54 out of 1785 ancient individuals (3.0%, CI: 2.3–3.9%)" (there was one person resulting from parent-child/sibling inbreeding). Another 2021 paper backs up this conclusion.

In summary, cousin inbreeding was rare in ancient data. It is implausible that suddenly got common and was common in Europe until the Christian church banned it in 1500. A more likely explanation for their ban is that the church was trying to break up powerful noble families that the church had conflicts with. If we accept the model that WEIRD is caused by low rates of cousin inbreeding, it would seem non-WEIRD human populations evolved away from WEIRD relatively recently. This doesn't sound so plausible. Perhaps a better explanation is that neolocalism (bride and groom move away from their families), which is rare in ethnographic data, but common in NW Europeans. We don't know how far back this dates to, but perhaps far enough to fit with the Roman sources about Germanics.