Substack's obsession with simplistic narratives based on IQ
Bad incentives lead to an empirically and philosophically misguided approach
Substack writers tend to fit a very specific mold: analytical introvert who (a) positions themself as outside of the mainstream and (b) has a high estimation of their own intelligence. In fact, the two attributes are clearly linked–the underlying thematic impulse behind virtually every Substack I have ever read is: “Here is something important that the normies should be talking about more. Sadly, they aren’t because they are too dumb to understand it, much less recognize its importance. Fortunately there are smart, dedicated people like us around to think abstractly about these issues during our respective lunch breaks.” Notably, this underlying thematic homogeneity transcends big differences in Substack type and content (e.g., empirical “rationalists” vs. the sophisticated culture critics who despise them).
Another thing that I’ve noticed is that a disproportionate share of Substack writers seem to believe that innate (as in, genetic) intelligence is by far the strongest determinant of life outcomes / success. And they appear to believe that this innate intelligence is accurately measured by IQ.
Right off the bat, let me acknowledge that there is such a thing as innate cognitive ability that differs from individual to individual. It seems probable that greater innate cognitive ability will tend to position someone for greater professional (and other) success.
However, I believe that the simplistic narrative that high IQ = high innate intelligence = success is extremely overrated by a certain subset of Substackers.
Here’s the argument: Intelligence is hard to define so IQ tests are inherently limited. They do not and cannot distinguish between innate vs. developed intelligence. This is relevant because excess focus on innate intelligence is not just factually incorrect, it is likely to lead to worse individual and societal outcomes. After all that, IQ is not even as strong a determinant of success as many Substackers seem to believe. I suspect that the reason that this simplistic narrative persists on Substack is that contrarian writers who make their living by discussing models of the world may have certain incentives (both personal and professional) to favor IQ-driven explanations.
Intelligence is hard to define so IQ tests are inherently limited
What even is intelligence?
The two pertinent definitions in Merriam Webster are:
The ability to learn or understand or to deal with new or trying situations
The ability to apply knowledge to manipulate one's environment or to think abstractly as measured by objective criteria
These are pretty broad definitions. Indeed (2) seems worryingly circular insofar as it seems to define intelligence as “that thing that is measured by intelligence tests.” But it’s clear that intelligence is a measure of our ability to do things. What types of things would those be?
Solving chess puzzles, perhaps? Reading ten philosophical treatises and then writing a cogent synthesis? Coding a new program? Performing mental math quickly and accurately? Understanding and recalling a bunch of interrelated facts?
What about creativity–the ability to think originally rather than merely ‘plugging and playing’? Or Sherlock Holmes-style deduction? Giving a compelling impromptu speech? Selling a product to a reluctant customer? Orienteering in a wild and uncharted wilderness?
Based on the above, part of me thinks that the very notion of “intelligence” is a conceptual confusion. A misguided attempt to collapse a complex, multimodal set of qualities into a single, measurable variable. Different sorts of activities require different sorts of intelligence. Instead of conceiving of the human brain as a combustion engine that generates “intellectual horsepower,” it probably makes more sense to conceive of it as a team of like 100 different horses, each with their own jockey and unique training and doping regimen.
Because intelligence is conceptually complex and it is difficult to reduce to a clear, operational definition that captures everything we want it to, IQ tests are in a bit of a bind. If intelligence was just about e.g., solving chess puzzles, then it would be quite easy to test. You’d just have people do a bunch of chess puzzles and the people who solved the most challenging ones would be more intelligent than the people who failed to solve even the easy ones. (Even this is complicated by the consideration of speed–should the person who solves faster be considered more “intelligent?” What if there is a master who is quite slow, but brilliant?). The attempt to reduce a multimodal, conceptually ill-defined phenomenon to a single number seems…ill-considered.
IQ advocates argue that, in fact, all of these types of intelligence are so correlated that the test can measure everything without measuring everything. It could be the case that, on average, the ability to perform mental math is correlated with, say, the ability to write an essay. Maybe being good at one intelligence-requiring activity means that, on average, you’re more likely to be good at another. However, it also seems clear that these activities are not all demanding 100% identical cognitive input(s). We all know people who are excellent at mathematical reasoning but struggle with its verbal cousin or vice versa. IQ tests are probably measuring some types of intelligence. But it is impractical to ask them to measure Intelligence with a capital “I” across all of the nuanced dimensions that are captured by the word in common parlance.
IQ tests do not and cannot distinguish between innate vs. developed intelligence
For the sorts of intelligence that IQ is measuring, how can we even determine what is innate vs. a product of one’s environment and prior learning? Twin studies? But in that case you’d need one twin growing up in e.g., dire poverty and one twin growing up middle class.
Think about what a test is: there are words or visuals on paper that you must interpret and respond to. If I took an IQ test in Mandarin then I would flunk that test because I do not understand Mandarin. Doesn’t that imply that language skills–which are clearly, to a great extent, learned–are a pretty crucial factor in test performance that cannot be controlled for? What about reasoning skills? Those can clearly be developed, too, through practice. Think about the chess puzzle example above–you might be able to teach a chess prodigy the rules of the game on the day of the test, and they might prove to be a natural whiz. But the kid who’s been obsessively studying chess for years is probably going to do pretty well, too. In fact, it’s highly likely that the trained kid will do better than the kid with natural ability. How do you control for these learning effects, which inevitably bias the data? What is a product of intellectual curiosity + time + resources vs. innate ability? And we haven’t even discussed environmental effects from lead exposure, nutrition, trauma, etc. There are literally an infinite array of important inputs to test performance beyond genetics.
It is implausible that IQ tests have ever or will ever fully control for this array of inputs. Children are not blank slates at any point in their lives, so the notion that we can quantify their native ability without some major noise from other factors is flawed.
Moving beyond appeals to commonsense and logic, we can clearly see this in the data. First, there has been a “substantial and long-sustained increase in [population-level IQ scores] in many parts of the world over the 20th century” called the Flynn effect. For example, the average IQ of British children increased 14 IQ points from 1942 to 2008. That’s a massive average increase! It definitely appears to signal that IQ is far from a pure measure of heritable intelligence. But, what gets really interesting is comparisons that control for the Flynn effect. Here’s a long explanation from researcher Ron Unz published in the American Conservative:
“[When we look at the IQ data from Europe] what we immediately notice is a long list of enormous variations in the tested IQs of genetically indistinguishable European peoples across temporal, geographical, and political lines, variations so large as to raise severe doubts about the strongly genetic-deterministic model of IQ favored by Lynn and Vanhanen and perhaps also quietly held by many others. (Unless otherwise indicated, all the IQ data that follow are drawn from their work and incorporate their Flynn adjustments.)
Consider, for example, the results from Germany obtained prior to its 1991 reunification. Lynn and Vanhanen present four separate IQ studies from the former West Germany, all quite sizable, which indicate mean IQs in the range 99–107, with the oldest 1970 sample providing the low end of that range. Meanwhile, a 1967 sample of East German children produced a score of just 90, while two later East German studies in 1978 and 1984 came in at 97–99, much closer to the West German numbers.
These results seem anomalous from the perspective of strong genetic determinism for IQ. To a very good approximation, East Germans and West Germans are genetically indistinguishable, and an IQ gap as wide as 17 points between the two groups seems inexplicable, while the recorded rise in East German scores of 7–9 points in just half a generation seems even more difficult to explain.
The dreary communist regime of East Germany was certainly far poorer than its western counterpart and its population may indeed have been “culturally deprived” in some sense, but East Germans hardly suffered from severe dietary deficiencies during the 1960s or late 1950s when the group of especially low-scoring children were born and raised. The huge apparent testing gap between the wealthy West and the dingy East raises serious questions about the strict genetic interpretation favored by Lynn and Vanhanen.”
He goes on to outline the same trend across a variety of different genetically-identical groups of people from Greeks to the Irish to Slavs. This data is pretty damning for the idea that IQ is a measure of heritable intelligence. Often folks who are keen on the idea that IQ = native intelligence will complain that those who disagree with them are relying on emotion and dogma rather than empirical fact. Based on this data, the reality is probably the opposite.
We should acknowledge that IQ tests will never be able to fully distinguish innate from developed intelligence. Which, in turn, probably implies that even if we do want to use IQ as a proxy for intelligence (and, for the reasons articulated earlier in the piece, I’m quite skeptical of that approach as well), we should be very wary of claims that it is an unbiased indicator of inborn potential versus development.
Excess focus on innate intelligence causes problems
This is not just an abstract argument, either. I think that there are very good reasons to avoid over-indexing on the importance of innate intelligence, given that it seems to be nigh impossible to separate out from developed intelligence. This is true whether we are trying to do the separating with traditional IQ tests or with any other mechanism we might devise.
Leaving aside the historical associations between obsession with the belief that most intelligence is innate and the eugenics movement, Nazism, scientific racism, etc. (which already should make people feel at least a bit wary about the whole enterprise), there are clear reasons that this is not only an empirically shaky belief, but a counterproductive one.
Questions of intelligence are relevant to our societal decisionmaking. Arguably the most important policy area downstream of questions of intelligence is education, and the most pertinent policy area within the field of education is that of educational tracking. In other words: (how) should we tailor a student’s education to their ability?
Think about two ways of framing a policy of educational tracking:
Intelligence is innate. Therefore we need to test for natively-gifted children in kindergarten to separate them out from the dummies, who should be on a different track. That way we can provide the “gifted and talented” kids with the advanced resources and curriculum suited to their inborn smarts.
Intelligence is developed. Therefore, we should tailor a student’s education to where they are on their educational journey. Students that are further ahead on certain topics should be offered curricula that challenges and excites them. Students that are not there yet on certain topics should be given the remedial curricula and attention that they need to make sure that they aren’t left behind. At the end of each semester, we should see how our different students are progressing and ensure that they are being offered the educational resources that they need to learn and succeed.
Note that both proposals are essentially advocating for the same policy approach, just using different language. At the societal level, which narrative do you think is going to get more public buy-in? I suspect that the narrative based on an intellectual elite needing to be cloistered away from the vast majority of mouth-breathing morons is going to get less traction in a democracy than the one about catering to individual children’s educational needs. If you want to fund and expand opportunities for children with greater ability, obsessing over the importance of their “innate” intelligence is silly. Which perhaps explains why programs branded as for “gifted and talented” students are slowly being dismantled across the country (to everyone’s detriment, I might add).
Which narrative do you think will bring out the best in students? The one that tells the “gifted” students that ability is inherited so (a) trying is pointless and (b) any failure or challenge they experience is proof that they are an imposter of glaringly weak intellect? And that implicitly tells the non-”gifted” students that they are congenitally less intelligent and probably should learn to expect poor academic performance regardless of how much effort they put in? Or, the narrative that tells kids that hard work and dedication enables them to cultivate their abilities, and teaches them to leverage the support that they need to get to the next level? I bet the latter would be simultaneously a lot more motivating and a lot less anxiety-inducing. It would also encourage more intellectual risk-taking, effort, love of learning, and humility.
It comes down to the combustion engine vs. team of horses analogy–are we given a lifeless piece of machinery at birth or a bunch of young and wild ponies? Educational systems that help students to train their team of horses to ever greater speed and endurance are probably going to do a lot better than those that implicitly endorse the view that intellectual horsepower is a finite, fixed quantity.
Of course it is the case that people are born with a spectrum of potential. But the spectrum is important here! We should focus on enabling people to reach the apex of their spectrum. One of my horses might have been the runt of the litter, but let’s nurse that foal to health rather than writing it off as congenitally defective.
IQ is not as clear a determinant of success as many argue
Let’s say that IQ was a measure of native intelligence. Substackers use this idea to advocate for a bunch of different ideas (e.g., that differences in native intelligence are evidence that trying to mitigate social inequality is a fool’s errand).
The funny thing is that IQ is not as clear a determinant of success as they seem to believe. Sure, one’s “spectrum of cognitive ability” probably sets a ceiling on success. For example, say I was obsessed with theoretical physics. Like, obsessed. The foundational texts were my gospel, I read the latest papers before breakfast, and studied it—boy, did I study it—in college. Irrespective of all this, I would never become a physics genius on the order of a Feynman. In the same way that if I were obsessed with basketball and had spent 8 hours a day practicing my jumper since age 10, I would still not have made it to the NBA. A Feynman or a Giannis is just built different (literally in the case of a 7-foot tall athletic freak).
But there are two things to note here. First, actual genius may be exceedingly rare, but with enough effort most people who are not intellectually handicapped can probably achieve in most fields at some level. E.g., I might have been able to become a physics professor at a mid-tier university or to play DIII college ball. That’s kind of just a restatement of my assertion that we should focus on cultivating intelligence to our highest ability, rather than focusing on what ceiling that ability sets.
However, the more important point is this: the highest-IQ people generally do not become the most successful. Success does not increase linearly with IQ. Indeed, some high-IQ people probably have their prospects of success harmed by their high IQ. Why would that be?
The first reason is that IQ is not a complete measure of intelligence. It doesn’t capture some of the types of intelligence (especially interpersonal intelligence) that are foundational to success in most fields.
The second is that IQ does not capture a bunch of non-intelligence qualities that are absolutely crucial to success. For example, executive function—the mental skills that help you get things done. Like, avoiding procrastination, adhering to plans, switching between activities relatively seamlessly, being able to understand what you need and how to get it to accomplish a goal etc. Plus, the qualities of paramount importance in most jobs–conscientiousness, motivation, hard work. Ironically, an obsession with innate intelligence probably serves to diminish each of these important skills.
Speculation as to why certain Substack writers might have a tendency to espouse simplistic IQ narratives
They have an affinity for simple narratives that purport to be explanatory, in part because of their own intellectual proclivities and in part because these get a lot of engagement from readers.
They like narratives that run counter to prevailing beliefs / that are unpopular, in part because of their own intellectual proclivities and in part because these get a lot of engagement from readers.
They themselves have high IQs, so they fetishize IQ in a self-congratulatory fashion. They like the idea of this being the rationale for their success. Corollary: they may have relatively lower skill in the forms of intelligence not captured by IQ scores, and like the idea of these being less “important.”
They are partisans of nefarious political projects (e.g., fascism) that are often cloaked under pseudo-empirical conversations about IQ.
Conclusion
Intelligence cannot be adequately defined and whatever it is, it is certainly not merely genetic. Therefore, IQ is not and cannot be a measure of innate cognitive ability. It is also not as clear a determinant of success as some people like to believe. But the simple narrative that high IQ = high innate intelligence = success remains popular on Substack because contrarian writers who make their living by discussing models of the world may be incentivized (personally and professionally) to favor IQ-driven explanations.