Book review: WEIRDest People

Link post

Book review: The WEIRDest People in the World, by Joseph Henrich.


Henrich previously wrote one of the best books of the last decade. Normally, I expect such an author’s future books to, at best, exhibit regression toward the mean. But Henrich’s grand overview of humanity’s first few million years was merely a modest portion of the ideas that he originally tried to fit into this magnum opus. Henrich couldn’t quite explain in one volume how humanity got all the way to industrial empires, so he split the explanation into two books.

The cartoon version of the industrial revolution: Protestant culture made the West more autistic.

However, explaining the most important event in history makes up only about 25% of this book’s focus and value.

Henrich doesn’t want us to think of it as the most important event - because he views it not as a single event, but as a stage in a long process. Most books on the industrial revolution concentrate on some subset of the 1500-1800 time frame. WEIRDest People devotes a majority of its attention to the prior millennium.

When I last reviewed a book on the industrial revolution, I was pessimistic about ever getting enough evidence to distinguish between too many plausible hypotheses. Henrich found a solution: most of the proposed explanations describe features that contributed to the industrial revolution; they follow naturally from the way that WEIRD culture (Western, Educated, Industrialized, Rich, Democratic) altered our psychology.

Important aspects of this new culture /​ psychology include: analytic thinking, nonconformity, impersonal prosociality (trusting strangers, treating them fairly), and internal attributions (e.g. the idea that a good afterlife depends on internal mental states, rather than rituals).

Along the way, Henrich provides at least partial answers to a surprising number of questions that I carelessly neglected to ask, such as:

  • how do rules about cousin marriage affect conformity (as measured by the Asch Conformity Test)?

  • how strong is the correlation between how individualistic a society is and its rate of innovation?

  • why does Latin have 25 words for prostitute?

  • how did the peculiarities of rice cultivation affect the ability of southeastern China to develop science?

  • how do social safety nets influence a society’s rate of innovation?

Much of Henrich’s focus is on this key question: how do increasingly large groups of people cooperate enough to form increasingly large societies?

The Dunbar Number and its cousins

There’s a phenomenon that is somewhat well known among financial traders of a limit of about 20 stocks, beyond which traders can’t remain sufficiently aware of the details to be a competent market maker.

Henrich describes what is likely another manifestation of the same phenomenon in “truly individualistic” human cultures, such as the Matsigenka, where hamlets rarely get as large as 25 people before nuclear families decide they prefer to set off on their own.

I’ve also noticed a seemingly similar phenomenon in business, particularly in a dot-com where I worked that rapidly grew from 4 to 75 people. At some point between the 20 person size and 40 person size, it switched from feeling like users were part of the company’s community, to a feeling that users were distant people who were dealt with via specialists such as customer service. Also, internal politics went from not being detectable, to being important.

Sadly, Henrich doesn’t mention a name for this 20-25 entity limit. Nor does he name the Dunbar Number, despite providing important insights into how cultures manage to create social groups that are bigger than the Dunbar Number.

Dunbar-sized tribes often end up with rituals that artificially create interdependence and kin-like bonds that help hold the tribe together.

Switching from a system of bilineal descent to unilineal descent prevents some kinds of conflicts between extended families.

Some other norms that promote harmony in large villages include: arranged marriages, making entire clans responsible for harm caused by any clan member, and well-defined hierarchies.

The Evolution of Religion

There’s at least one more size limit, well above the Dunbar Number, where rituals that expand kinship become inadequate for further expansion. To overcome that, societies needed Big Gods who can command subjects to cooperate with distant co-religionists.

Belief in heaven and hell correlates with (and likely causes) a large increase in economic growth. Alas, belief in heaven alone doesn’t seem to be very valuable.

A similar pattern is seen for belief in supernatural punishment in societies before European contact.

The estimated probability of a historical transition to a complex chiefdom when no such punishment existed was—surprisingly—close to zero. By contrast, when ancestral communities already had beliefs in supernatural punishments for important moral violations, there was a roughly 40% chance of scaling up in complexity every three centuries or so.

Those religions succeeded better if they destroyed the kinship institutions that had previously been needed for scaling up past the Dunbar Number.

Why? Kin-based clans interfered with loyalty to larger, more abstract groups such as Christianity or nations. Here’s a quote from a politician in contemporary Pakistan that illustrates how kin-based group identity conflicts with newer, larger social groups:

I have been a Pashtun for six thousand years, a Muslim for thirteen hundred years, and a Pakistani for twenty-five.

How? By changing many rules involving marriage and family relations. That included banning the marriages between cousins (a ban which sometimes extended to sixth cousins), and requiring monogamy.

Don’t assume you can design your own religion:

the powerful Mughal emperor Akbar the Great tried to unify his Muslim and Hindu subjects by making his own highly tolerant religious creed … At its peak, the powerful emperor’s religion accumulated a total of only 18 prominent adherents before vanishing into history. My point is that throughout human history, rulers needed religion much more than religion needed rulers.

I sometimes got the feeling that the Western Christian church’s success at stamping out kin-based institutions had to be mostly due to careful planning and foresight, but Henrich implies that’s mostly hindsight bias, and calls the process “accidental genius”. Henrich has good arguments that cultural evolution includes an important amount of semi-blind trial and error, but I suspect he goes a bit overboard with this line of thought.

E.g. Deuteronomy 25:5-10 clearly describes levirate marriage as an obligation. How does unplanned exploration of cultural variation get from there to declaring levirate marriage a sin, while still treating the bible as the word of God?

Fukuyama has a better model for that than Henrich: in The Origins of Political Order:, he implies that the church had a fairly deliberate strategy of destroying the kinship ties that were hindering the church’s goal of inheriting property. Note that it’s fairly WEIRD of me to care about whether the church’s strategies were intentional.

Faster Cultural Evolution

Kin-based societies have strong constraints on who people can associate with. Accidents of birth, plus arranged marriages, mostly determine who a person will interact with. Cooperation between kin works fairly well, but distrust hinders most cooperation outside of the clan.

Early Western culture caused many clan-oriented social interactions to wither, creating some desire for new types of interactions. It also made people more independent and trusting of strangers, which enabled a wide variety of voluntary organizations to develop, such as guilds, charter towns, monasteries, and universities.

Monasteries morphed from clan businesses to NGO-like organizations. Guilds had to compete with goods from similar guilds in nearby towns. Most organizations competed for new members. Cities needed to experiment with better governance (including some democracy) in order to offset the urban graveyard effect.

That competition may not seem like much, but it seems to have been more than was possible in China /​ India /​ Islam. It likely contributed to faster experimentation and vetting of new cultural features.

Flynn Effect

The power of Henrich’s model can be illustrated by asking how it explains the big 20th century increase in IQ. Henrich doesn’t discuss this topic directly, but if I’d read WEIRDest People before learning of the Flynn Effect, I expect I would have found the Flynn Effect unsurprising. It seems like a natural consequence of thinking styles that became more analytical, abstract, reductionist, and numerical.

Moreover, Henrich’s model provides clues as to why low-IQ cultures are reluctant to adopt the changes that raise their IQs. It’s not that they’re lazy or held back by harmful mutations (Henrich doesn’t dismiss the existence of those problems; instead, he convinced me that WEIRD culture shifts are more powerful explanations).

An important insight is that people take cues from their environment early in life, and use those cues to invest in cognitive features that are expected to yield the most benefit.

WEIRD culture gets people to invest more in high-IQ cognitive features, at the cost of less investment in skills that foster social ties (e.g. learning to read at early age seems to impair facial recognition). “WEIRD people are bad friends”—beliefs such as impartial rules, and moral universalism have important social consequences. E.g. WEIRD people are less willing to lie in court to keep their friends out of jail. It looks hard to separate that effect from the cognitive styles that promote high IQ.

There are also trade-offs between analytical thinking and holistic thinking. IQ tests tend to favor the analytical approach that Western societies reward, while kin-based societies reward holistic thinking more (see Ecocultural basis of cognition: Farmers and fishermen are more holistic than herders).

Many other books on the industrial revolution now sound like the proverbial blind men and an elephant.

  • In The Origins of Political Order:, Fukuyama sees about 20% of what Henrich sees, and is the only person I’m aware of that traces the origins of key features further back than does Henrich. Fukuyama also explains better why the industrial revolution happened in Europe rather than China. This now looks like clearly the second most important book on the industrial revolution.

  • State, Economy, and the Great Divergence: Great Britain and China, 1680s − 1850s, by Peer Vries identifies a modest fraction of the cultural differences that Henrich discusses. Vries seems to disagree with Henrich about the mobility of the average British worker, but otherwise supports Henrich more than I recalled. Vries’ expertise as a historian lends credence to Henrich. I wish I had time to carefully recheck the extent to which Vries’ evidence supports Henrich, but Vries is hard to read.

  • Nick Szabo’s Book Consciousness explanation captures a medium-sized portion of Henrich’s vision, and the two complement each other fairly well.

  • A Farewell to Alms, by Gregory Clark seems a bit like WEIRDest People, with lots of tension between their details and tone. Bryan Caplan’s objection illustrates some of this:

    Why are rich countries so much richer than poor countries, according to Clark? Because they have lower-quality workers.

    Obvious objection: If that’s the problem, why does moving these low-quality workers to the West quickly raise their wages from $1/​day to $40/​day? Yes, that’s below average for the West, but it’s in the same ballpark. If that isn’t iron-clad proof that institutions/​policy matter a lot, what is?

    Henrich resolves this dispute by convincing me that “lower-quality workers” means workers whose culture and psychology are poorly adapted to most modern jobs. Clark’s arguments could be modified to accommodate that view without drastic changes.

  • The Birth of Plenty by William Bernstein presents a view of institutions circa 1800 that mostly agrees with what Henrich considers to be important for the industrial revolution, but Bernstein’s attempts at explaining how the West got there seem rather pathetic in comparison to Henrich’s.

  • The Great Divergence: China, Europe, and the Making of the Modern World Economy, by Kenneth Pomeranz raised the bar by pointing out that leading discussions of the industrial revolution failed to explain why Europe did better than China, but Pomeranz went overboard in claiming those two regions were similar. Pomeranz tried to argue against cultural explanations in general, but seemed confused as to how culture could explain more than differences in luxury goods. Were older cultural explanations (such as Max Weber’s?) really that weak?

  • How the West Won by Rodney Stark may share some of Henrich’s attitudes, but I’ve mostly forgotten what Stark wrote because he wasn’t very convincing.

  • Where Is My Flying Car? by J. Storrs Hall almost sounds like a sequel to Henrich’s books, explaining how Western culture is decaying due to problems such as Christianity being replaced by religions that are less well adapted to modernity. Henrich has increased my confidence that Where is my Flying Car? is more than half-right about the causes of the Great Stagnation.

  • I see a strange parallel to Marxism when Henrich describes the need for societies to go through different, partly opposing (dialectic), stages. But Henrich is the opposite of a Marxist in many other respects, such as the ability of intellectual leaders to predict those stages.

How Credible?

I’ll guess that the book is about 80% correct.

Henrich exaggerates and oversimplifies a modest amount, but he seems to be a pretty careful researcher, actively trying to find empirical tests of alternate causal models.

He often cites evidence that isn’t especially compelling, but he’s careful not to depend much on any one piece of evidence. E.g. he apologizes for only being able to cite one study each for the claims that the BIG-5 personality dimensions and endowment effect are not universal.

The book is somewhat limited by only having a sample size of one for some of his broadest claims, but Henrich manages to find a larger sample size for many interesting sub-points.

E.g. I had assumed that China’s one-child policy was tricky to evaluate because it was only imposed once. Yet Henrich points us to Sex ratios and crime: Evidence from China, showing that we can get sort-of-causal evidence from comparing provinces that implemented the policy in different years. Yes, it sure looks like the policy caused crime to increase (the policy may have also had desirable effects via weakening kinship ties—Henrich doesn’t express any overall opinion on the policy).

Henrich has a remarkable range of expertise (anthropology, evolutionary biology, engineering, psychology, and economics); these maybe make him better than a historian for the purpose of this book.

Historians are apparently upset at being bypassed, and at the inadequate nuance of a shorter version of WEIRDest People, but their disagreements don’t sound particularly important to me.

If you’re still undecided about whether to read the book, a fair amount of Henrich’s evidence is available in this preprint.


This book is essential reading for any serious scholar of human nature.

Not only does it demystify some of the most important processes of human history, but it also provides an unusually balanced view of how Western culture compares to other cultures. Henrich discredits both “all cultures are equal” worldviews, and most of the common claims of Western superiority. Western culture is genuinely superior in key respects, but that superiority comes with possibly large downsides.

Henrich is a master at organizing large amounts of evidence into an understandable package.

Please don’t treat this review as an adequate substitute for reading the book. I can’t describe enough of Henrich’s model to sound half as convincing as Henrich’s full description is. I only hope to whet your appetite enough to convince you to read the book.

P.S. I was maybe a bit misleading when I used the word autistic to describe the psychological changes that Henrich attributes to Christianity /​ Protestantism. I can’t confirm that he’s even familiar with autism. I find it to be a convenient label to approximate his more nuanced, but hard to summarize, description of Western psychology.