Bayesian Investor Blog

I said in my review of WEIRDest People that the Flynn effect seems like a natural consequence of thinking styles that became more analytical, abstract, reductionist, and numerical.

I’ll expand here on some questions which I swept under the rug, so that I could keep that review focused on the book’s most important aspects.

Cultural Bias

After reading WEIRDest People, I find that the goal of a culture-neutral IQ test looks strange (and, of course, WEIRD). At least as strange as trying to fix basketball to stop favoring tall people.

I found after writing that review that Muthukrishna and Henrich discuss IQ in Innovation in the collective brain (hereinafter referred to as MH2016), saying roughly what I inferred from WEIRDest People.

MH2016 says: “IQ measures the abilities that are useful at school and work in these societies.”

There’s a good case to be made for removing accidental cultural bias. Doing so will increase the test’s ability to detect cognitive patterns that are oriented toward science and technology.

Similar cultural biases pervade psychology. E.g. there are real cultural differences about the wisdom of conformism, which lead me to doubt there’s a culture-neutral rule for what the correct answer is in the Asch conformity test.

Strong cultural biases are built into the central ideas behind what we call science. E.g. it’s somewhat WEIRD to believe that we should seek universal laws of nature, rather than expecting valuable knowledge to be context-dependent knowledge.

Could a genuinely culture-neutral IQ test be developed? I presume it’s possible to shift the emphasis so that it measures holistic reasoning more, and analytical reasoning less. But what would be a neutral mix of those two? You could average over all humans currently alive, but that would produce a rather different result than if you’d taken that average 1000 years ago.

Changes of this nature seem likely to reduce the value of IQ tests at satisfying the original goals behind testing, without satisfying critics.

If I really wanted to develop a culture-neutral IQ test, I’d start with the research that has been done on measuring cognitive abilities of nonhuman animals and software. I suspect that that research has so far remained an obscure niche due to confusion about whether it ought to measure something more universal than the skills needed in current high-status jobs.

Timing

The rise of WEIRD culture happened over many centuries. WEIRDest People doesn’t provide a lot of evidence that Western culture became more WEIRD during the 20th century, which is most of when the Flynn effect has been measured. Does this mean we’re just measuring the tail end of an effect that was stronger in prior centuries?

I’m pretty sure it started centuries before 1900, but it seems somewhat likely that the most dramatic effects happened after 1900.

A casual reading of the book might lead you to imagine culture spread fairly uniformly throughout a population. Yet I see hints that the book focuses more than Henrich admits on people in the most knowledge intensive professions.

The median man in 1900 needed muscle power much more than he needed brain power, but that shifted fairly steadily toward mostly needing brain power by 2000. To the extent which people optimize their cognitive abilities for their needs, I’d expect the median IQ to have risen mainly in the 20th century, even when some parts of the population experienced important IQ rises much earlier.

Some of the culture changes depended on reading in early childhood. I presume that practice spread fairly slowly, and didn’t reach diminishing returns until well after 1900. From MH2016:

[Tsimane] Children and adolescents with access to schooling showed a strong linear effect of age on IQ score (R^2 = 0.519), compared with no effect of age in those with no access to schooling (R^2 = 0.008). These results suggest that IQ increases with age not because of maturation, but because of the influences of a particular WEIRD cultural institution: formal schooling.

Also, nothing that Henrich says implies that changes in culture have slowed down. I presume he omitted discussion of most 20th century trends due to lack of time and/or space.

Nutrition

Henrich also points to evidence that the Flynn effect might be partly due to better nutrition, particularly better childhood nutrition.

Henrich cites Precocious albion: A new interpretation of the British Industrial Revolution, which argues that better human capital was a key driver of the industrial revolution, and nutrition contributed to that via improved health.

Diets had more meat, and several hundred more calories per day, in England than France in the late 1700s. English heights were 5 centimeters taller than French heights. Height is correlated with both nutrition and IQ.

That paper also shows evidence that English workers were more productive at standard common tasks such as reaping and threshing, and suggests this was due to health differences.

Is Childhood Meat Eating Associated with better later Adulthood Cognition in a Developing Population? suggests that meat improves cognition, at least compared to being too poor to afford meat (i.e. there’s some wealth related benefit, but it’s hard to say whether meat was a cause or a symptom).

Killing Me Softly: The Fetal Origins Hypothesis by Almond and Currie (nominative determinism alert!) summarizes evidence related to the earliest stage of nutrition: IQ was apparently not affected by the 1945 Dutch hunger winter, and there seem to be no studies directly showing that fetal nutrition affects IQ.

Yet they point to some studies saying something close to that. E.g. Health Capital and the Prenatal Environment:

the occurrence of Ramadan early in pregnancy nearly doubles the likelihood of a disability related to diminished cognitive function.

Iodine supplementation “can explain roughly one decade’s worth of the upward trend in IQ in the United States”.

Fogel argues that the biggest gains in nutrition started around 1900.

So I’m inferring from this somewhat sketchy evidence that there was a weak nutrition-led Flynn effect starting before 1800, which accelerated around 1900, and has been diminishing or maybe even reversing in developed countries more recently. But I’m guessing, based on the weakness of the evidence, that it’s not the primary cause of the Flynn effect.

Conclusion

These IQ differences are easy to confuse with genetic effects. Parental culture influences child culture, if only via influencing what community the child lives in. Maternal nutrition influences child nutrition. So the effects are somewhat inherited, even if not via genes, and partly have life-long effects.

Flynn explained the Flynn effect as a shift in how people view the world, from pre-scientific spectacles to post-scientific spectacles. That continues to look mostly right. Henrich mainly provides a richer description of that shift, and clarifies that it wasn’t a free lunch (it involved some loss of social bonds).

Have I learned anything about how IQ will change in the future? I certainly have a deeper understanding of why it’s hard to predict.

Our affluence has given many of us the luxury to return somewhat to a more forager-like culture, to which we’re more adapted.

But I still see a bit of pressure toward the kind of abstract thought and universal rules that have enabled cooperation within large populations.

An Age of Em might dramatically increase the returns to IQ-oriented cognitive styles, or an AI-run garden of Eden might enable a larger shift back toward forager cognitive styles.

I’ll weakly predict a trend of increasing IQ, but with large sectors of the population that decide to opt out and let their IQ slide.