But you can use basic common sense. If everyone in Malawi knows that tall parents tend to have tall kids and smart parents tend to have smart kids and we know from Finland that IQ is nearly as heritable as height it takes a lot of perverse reasoning to remain resolutely agnostic as to whether height, or lactose intolerance, or IQ are largely heritable in Finland but largely environmental in Malawi. One of these is more likely than the other.
National IQ estimates tell us something about what is causing under-development, how optimistic we should be about current development interventions, how much we should start looking for other possible development interventions, how immigrants from a country are likely to perform when exported to another country, how many immigrants we should accept from other countries, how much we should address inequality of outcomes between people of one nationality by efforts to correct prejudice or compensating them because of a presumption they have been unfairly treated.
"Scientific racism" is obviously pejorative. There is nothing racist about noting that every two real normal distributions of non-integer values will have different means. If you spend all of your time obsessing about IQ and black vs. white, on the other hand, you just might be racist.
If we know there are 30 point differences between pre and post industrialization within the same societies shouldn't it be pretty self evident we see 30 point gaps between countries?
Are you referring to the pre- and post-agriculture confusion Scott talks about here?
"Reich had said that pre-agriculture Europeans were '2-3 standard deviations' below moderns. I had interpreted that as IQ deviations of 15 points, making them genetic IQ 55-70, which would have been pretty crazy. [Lyman] Stone tells me he actually meant PGS deviations, each of which was about 3-4 IQ points, so he’s claiming that pre-agriculture Europeans had genetic IQ of 90"
I'm skeptical of that summary of the finding. Any selection pressures that impacted the genes used to compose a PGS score probably impacted the frequency of a bunch of other genes for intelligence we haven't identified yet.
(This is why PGS tracks with observed group-level phenotypic IQ across racial/ethnic/religious lines so well, even though it only directly accounts for ~10% of the variance.)
Fair enough, but the "30-point differences" is still based on a misunderstanding of the point that was made (even if that point was wrong). Or does it come from somewhere else?
OK. I wasn't aware that worked out to a flat 30 points.
In fact, I doubt that it does represent a flat 30-point industrialization bonus for all societies, because societies are presumably starting from different baselines on all the factors that may contribute to the Flynn effect (poverty, mal-/nutrition, education, environmental toxins, etc.) So it really gets to the heart of the topic: Yes, pretty much everyone agrees that there's a very significant environmental contribution, but we don't know how big it is, or how it is distributed, among or within societies. It's a bit facile to say it is self-evidently 30 points.
(And as for MA_browsing's skepticism, it is interesting, but still not particularly relevant to my reply when I misunderstood your comment, and even less relevant to your original comment.)
I was generalizing a bit too much, sorry for that. The initial Flynn effect was the re-meaning of IQ scores because they increased by roughly 3 points per decade in the United States. The strength of gains vary according to country, intelligence domains, and the investigated time span. Some areas like sub-Saharan Africa have shown low or no gains while, countries like South Korea and Japan have had periods of 7.7 point per decade gains with 35+ point gains overall.
If my reading of the Flynn effect is correct, this suggests pre-agricultural Europeans had an IQ about in line with Americans in the 1980s? Seems to undermine basically any suggestion in my mind that this means anything whatsoever.
This speaks specifically to genetic IQ, whereas my understanding of the Flynn effect is that it is best understood as a raising of the mean largely due to environmental changes that raise the lower part of the distribution in the US (through better education, nutrition, less poverty, etc.) up into the 80s. Also the 3-point pr decade probably can’t easily be generalized to all other periods and places. It would be astounding if genetic IQ increased by something like 3 points (or even 3%) every decade – especially ever since the agricultural revolution.
Does genetic IQ make sense as a concept? Is this not just a renaming of g? I would guess the explanation would be what they'd score if they were brought up in the existing American system, but why is the educational system of the US privileged as the "normal" education system? It seems impossible to determine what the objective measure would be because everything is a mixture between genetic and educational components. Maybe the US has terrible education and very strong genetic advantage? Maybe the opposite.
In principle even pure g-factor can be lowered by environmental factors, like malnutrition or cyanide poisoning or viral infections of the nervous system and so on.
The Flynn effect doesn't represent a true intelligence gain because it isn't measure invariant. AFAIK there's no reliable way to norm test results over time so it's not really meaningful to say that modern Americans are 30 points smarter than 19th century Americans. Informally it seems pretty obvious that that can't be true because if the average IQ a century ago was 70 then we never could have modernized.
That only holds if IQ is actually normally distributed; if in the past there were more prevalent environmental limitations to most people's IQ but the distribution were more heavy-tailed, we'd still have enough smart people to modernize society.
(Yes, I know IQ is normally distributed by definition; just pretend I said some sort of "effective IQ").
This doesn't mean that your overall statement is incorrect, though.
I am inclined to believe the IQ test results are primarily a product of human adaptability at early ages. That is, if you grow up in a literate culture with lots of emphasis on the kinds of thought present in solving pen-and-paper puzzles, you will be much more capable of understanding them, and much less capable of understanding the things that you would understand were you brought up as an illiterate hunter-gatherer. Not explaining myself very well, but a friend told me about a study suggesting that literacy restructures our brain by overwriting a portion of it that would otherwise be used for recognizing faces. Assuming something like this is right, I would imagine it extends to the sorts of concepts involved in any puzzles, and so I probably wouldn't expect an illiterate hunter-gatherer of average intelligence to be capable of understanding anything he was shown on pen and paper, regardless of whether it involved words or not, to the same level as a literate person with a high school diploma.
While I agree the original statement is too strong, IQ test could measure something closer to intelligence in literate cultures and fail to measure that in illiterate cultures, through the process he said above. Especially considering IQ test already require normalization to work correctly, even within literate societies.
By the same token even intelligent people from literate societies often appear to hunter gatherers to be childlike in their inability to acquire basic observational and tracking skills so the „overwriting“ theory makes sense to me.
Some of the best hunters in the world are westerners who have just spent more time hunting than the rest of us. It's probably not anything to do about "overwriting." You can learn hunting skills even as a fully developed and literate adult.
I would strongly expect that all of those individuals were also taken hunting at a young age, or at the very least participated in types of play that used adjacent skills.
"Overwriting" might not be quite the right word, since it implies replacing skills that already existed, but it's certainly true that early development devotes more or less "space" for skills based on their usage, sometimes in in areas (evolutionarily) "intended" for other skills. (Of course, neuroplasticity doesn't entirely go away with age either, it just gets less effective)
Why wouldn't you just expect that hunting is a skill like any other?
I expect that hunting is like anything else -- if you devote a lot of effort to learning about it and practicing it, even starting as an adult, then you can become very good at it -- if not among the best in the world, then at least as good as the average tribesman.
Right, but *like any other skill*, the people who are the best in the world usually started practicing when they were young. Like I said, neuroplasticity is just less for adults, and at the top of the bell curve you usually have people with every advantage. And starting will usually give you a greater effect for less effort. (Plus in this specific case, 'dads taking their son hunting' is practically a stock story, so it's not like it's uncommon.)
Best hunters in the world with what tool package and education? Would these hunters really be comparable to the skillset developed by societies with a hunter-gatherer cultural package?
Yes, Henrich mentions that early on in the book. He claimed that literacy re-purposed the ability to process faces.
Unfortunately the claim is based on studies by Dehaene with very small Ns (~10) and interpretations of fMRI images (which are controversial).
There is a paper by Huettig, F., & Mishra, 2014, that concludes:
"Dehaene and colleagues (2010) have suggested that face perception abilities will suffer as individuals become literate because orthographic representations ‘invade’ the ventral visual system and interfere with the enlargement of faces and houses into the surrounding cortex.
There is no solid experimental evidence from behavioral studies to support such claims. Huettig et al. (in preparation) compared complete illiterates and literates from the same village and the same community of Indian society on the recognition of the emotions of the standardized Ekman (Ekman & Friesen, 1975) faces. There was no hint of a difference between the two groups. The Ekman faces are sometimes criticized for being culture-specific (i.e., Western), and thus, perhaps these results have to be interpreted with some caution. Orihuela, Carreiras, and Dunabeita (2013) have also investigated whether illiterates and literates differ in face perception.
These authors found in fact the opposite effect, illiterates made more errors than literates in a face recognition task of varying degrees of difficulty. Thus, as in the case of memory research, literacy may improve face perception rather than being detrimental."
I offer my own n=1 observation: I am pretty bad with faces (not terrible, but noticeably worse than most other people) and I learned to read very young.
An ordinary 21st century Westerner with a TV can easily recognise thousands of faces, more faces than an average hunter-gatherer will ever encounter in their whole lives. If literacy has invaded the facial-recognition part of our brain then what are we supposed to have lost?
The hunter gatherer will have learnt thousands of plants and animals, more than an ordinary 21st century Westerner will ever encounter in their whole lives.
Evolution gave us that capability and it wasn't so we could recognize Z-list celebs!
I was echoing Melvin's comment - the numbers aren't important. But hunter-gathering is memory/information intensive: you need to remember which trees produce edible fruit or nuts, where they grow, what time of year they produce, how to collect and how to prepare them to remove toxins - multiply this by all the other things you need to eat over the different seasons.
According to Aporia, the claim that Lynn described himself as a “scientific racist” is taken out of context. He said, “If we are talking about people who believe there are genetic differences between the races, then I am definitely a scientific racist.” I can’t speak for the other claims about his beliefs or who he was affiliated with, however.
I’m assuming he meant genetic differences *in intelligence* between the races, but I guess it’s unfair to attribute that broader context to this one quote
Even if it was it's still not racist. You're making a category error in conflating empirical claims with normative ones. If races differ in IQ for genetic reasons then they do. That's a scientific question that has an objective answer. It doesn't make anyone racist to have a good-faith opinion about objective reality.
In fact it's very dangerous to tie anti-racist views to arguments like this. What would you do if it became the indisputable scientific consensus that IQ is 90% genetic and definitely differs between races? Would you then be compelled to describe yourself (and everyone else) as a scientific racist?
I’m just describing how the concept is used; not commenting on its normative valence. For 95% of people, if you tell them you think Whites are genetically smarter than Blacks, they will call you racist.
“Okay, but you shouldn’t use that definition. Like I would refuse to use the word “Quran” to mean “the true word of Allah” even if societal developments lead it to be commonly defined that way because it’s unhelpful for many reasons. It’s better to use a different word without the assumption of “true” within the definition.”
I don't know if it's 95%, but you're pretty much right. But probably because they are assuming (incorrectly) that there is no good evidence, and (possibly with some justification depending on their social context) that people who say it don't like other races, otherwise they would respect the taboo and not investigate the topic, and/or just keep their mouths shut.
A non-trivial percent will also say stupid incoherent things like "racism = prejudice + power" and only whites (all whites, everywhere, even poor homeless ones, because they "participate in and benefit from whiteness" or something), so if a Korean, for example, says, "god I hate f***ing n***ers, and I intend to discriminate against them", that's not racist. People have all sorts of stupid and incoherent beliefs.
As Scott argues, people switch between different definitions without realizing it, but the only one that really makes sense, and also the one that people ultimately fall back on in practice is the definition from motives (being motivated by ill-will toward a particular race(s)).
+1 - you can acknowledge differences and still treat people without prejudice and not make categorical judgements about them. The term racism includes discrimination based on race.
I don’t have a perfect definition of racism. I just know that it can’t be that because scientific differences between races doesn’t justify cruelty, disrespect, lack of rights, etc. towards other races.
The word's meaning has swollen quite a bit beyond that limited definition, though. Almost all the people who fought to abolish slavery in the 19th century would be considered 'scientific racists' by modern standards.
Okay, but you shouldn’t use that definition. Like I would refuse to use the word “Quran” to mean “the true word of Allah” even if societal developments lead it to be commonly defined that way because it’s unhelpful for many reasons. It’s better to use a different word without the assumption of “true” within the definition.
The way I learned it at least is that racism was trying to allocate resources based on racial differences. Just positing a difference between/among races isn't enough.
So if I say "Koreans and Nigerians are different", that's not racist according to the definition I learned in college. Even if I say "Koreans and Nigerians are physiologically and biologically distinguishable", that's not racist. But if I say "Koreans and Nigerians are different, and therefore I won't let one group into my store", that's racism.
Once you tie some kind of resource allocation to being a member of a particular race, that's where you run into trouble.
From what I can see on the Wikipedia article, it mirrors what I learned in college. Wikipedia defines scientific racism as "is the pseudoscientific belief that the human species is divided into biologically distinct taxa called "races", AND that empirical evidence exists to support or justify racial discrimination, racial inferiority, or racial superiority" (emphasis mine). Without that second part, I don't think you have racism.
I think that many people (especially today) drop the second clause when attributing racism, and think that any highlighting of (especially biological) differences counts as racism.
Why? In general I think that there *had* been a decoupling of these beliefs, at least in terms of people willing to discus so publicly. If you imagine the quadrant, nearly all public discussion got sorted into the groups of (Differences + Discrimination) or (No Differences + No discrimination). This state of affairs lasted for roughly the long second half of the 20th Century. Highlighting major genetic or biological differences between groups of people that may gesture towards "race" being real was increasingly the domain of people who happened to also be racist or at least vaguely gestured in that direction.
It wasn't purely coincidental that this original sorting from 4 groups to 2 happened. The descriptive claim isn't purely independent of the normative claim. Consider racial justice or feminist campaigns of recent years. The ~roughly mainstream scientific (and Scott's?) view "racial differences are real but we can treat people equally" view is now more popular and acceptable, mostly due to the sheer amount of genetic evidence coming out every day. But it also saw the decoupling on the other side, people who thought (innate) racial differences weren't real BUT you should treat people differently.
You might correctly identify differences between groups, deny the biological explanation, and then attribute said differences to solely environmental factors. If you then think "and that means we should act differently to if all these groups had similar outcomes" (e.g. DEI) then it means your descriptive claim has influenced your normative claim. But if you instead think that "wait, no, a lot of these differences are actually due to *biology* or *genetics* (and those cluster in groups etc. etc.)" then you can see how the maintenance of "differences in people should entail differential treatment" (the same macro view as before!) becomes racist.
If the four views then are "90s liberal race blind", "technocentrist enlightened genetics liberal", "DEI advocate" and "dictionary definition racist", which of those today could get you labelled as a racist?
For the last 25 or 30 years, I've been publicly articulating the moderate, reasonable, empirically-backed position that the racial differences due to nature and nurture that we see when watching sports on TV are also seen in other aspects of society. (I realize that most of Scott's readers don't watch sportsball, but a whole lot of people really care about it.)
I've pointed out that the obviousness of racial differences in sports performance (e.g., a Kenyan is about a zillion times more likely to win the 3000m Steeplechase gold medal at the 2028 Los Angeles Olympics than a Bangladeshi) has not led to the genocide of the (literally) weaker or stronger races.
Instead, people really like sports, which leads them to having heroes of different races. White football fans really admire Lamar Jackson (a black QB) and black football fans really admire Josh Allen (a white QB). They might be modestly prejudiced in favor of the football hero of their own race, but ... that's okay.
When people ask me when I became aware of racial differences, I point to an October 13, 1968 college football game I attended at the Los Angeles Coliseum featuring two All-American running backs. #13 Oregon State's star was the huge Steve Enyart, who used his upper body strength to devastate #1 USC's defensive line in the first half. But in the second half, USC's superstar tailback used his lower body speed to overwhelm Oregon State. His name was O.J. Simpson.
But these obvious racial differences in sports didn't cause American elites to conclude that, well, sure, nature and/or nurture lead to racial differences. Instead, they concluded that anybody daring to mention the obvious was an evil enemy of society.
I've oversimplified, obviously there are "acceptable" ways of treating people differently due to their differences that don't get you labelled as a racist, high performance sports being one of them. Nozick chose Wilt Chamberlain for his famous thought experiment after all. If you interrogate this too deeply or take it toward logical conclusions (reward elite talent, but don't discriminate at the bottom end..?) it falls apart of course. So you could self deceive that you sit in one quadrant and not in the other.
Anyway this is more for public consumption, I don't need to explain this to you Steve. I do note that your own views (and self!) have become more mainstream as of late and this is largely due to that 2 quadrant equilibrium breaking down and the millions of sequenced human genomes we now have.
If you truly think that widespread acceptance of the idea that black people are less intelligent than whites would not lead to a massively more negative perception against black people from other races, you're deceiving yourself. Most people assume that white people are not that good at basketball, which doesn't do white basketball players any favors, but doesn't move the needle that much because basketball is a tiny part of life. If people assumed that black people had the skill level for jobs requiring intelligence as white people have at basketball, it would lead to a significant increase in bias against black people.
Professional athletes are outliers. Already I am seeing (unprompted by anything I choose to engage with) pretty racist memes on Instagram Reels and (especially) twitter based on the idea that black people are less intelligent. Maybe that theory is correct, but if it becomes widely accepted things will be much, much worse for a lot of people, and I don't believe you are being honest when you suggest otherwise.
That suggests a useful academic study: how much are white athletes hurt by stereotype threat and the like? Over a 20 year stretch that recently ended, no white players were regular starters at cornerback. How come? Well, blacks tend to be better cornerbacks? But 100% black with no whites or Samoans? Really? Probably something else like the old corporate saying: no executive ever got fired for buying IMB. Similarly, no coach ever got fired for starting a black cornerback, but many could imagine getting fired for starting a white cornerback and he gives up a touchdown.
So, here's the question that could be studied: how much different would be a world where nobody perceived race at all? How many more whites would have started at cornerback?
My guess is: a few. Not really all that many. Blacks really are much better at cornerback on average than nonblacks. But I could be wrong. Perhaps racial stereotyping kept dozens of whites and Samoans from playing cornerback.
Doesn't anybody want to know this?
Especially because it seems obviously useful to estimating the size of the harm you dread as people like Scott make it more socially respectable to be unignorant about race and IQ.
An interesting aspect of the widespread ignorance of how huge is the intelligence gap between Asians and blacks is that college admission affirmative action is highly unpopular with the public, even though affirmative action is the only way to have more than negligible black representation at our most elite academic institutions. I've often joked that when you become president of Harvard, they give you the key to unlock the Harvard President's Book of Secrets, which turns out to be a dog-eared copy of "The Bell Curve."
The point isn’t that what you’re saying isn’t true, that point is that for most people who have heard of Steve Sailer, Steve Sailer is the textbook example of a scientific racist. So I think we have a good working definition of scientific racism, which is just, whatever Steve Sailer says. I say this unironically with no disrespect; I also consider myself a scientific racist.
Kenyans, Ethiopians and Ugandans who do well in middle and long distance running grow up and train at altitude where the air is thin. The body adjusts to the conditions and becomes more efficient at handling oxygen. Uganda recently build a training centre at altitude open to athletes from all over the world. The distance running genes haven't been discovered (yet).
Kenyans and Ethiopians also do pretty well in distance running when they move to the US and compete at low altitude, or when their parents do.
For example, in the all-time great 1500 meter men's final in 2024 Olympics, the bronze metal was won by Yared Nuguse, an Ethiopian-American who was born in Arlington, Virginia and went to high school in Louisville.
Setting aside the obvious problems (that AA programs are textbook-racist under the above definition), I think it's going to be very difficult, in practice, to acknowledge the 'biologically distinct taxa' element without leading to the 'discrimination/hierarchy' element.
"We should always judge individuals on their individual merits" all sounds very fine and noble, but you can't evaluate individual merit unless you have perfect information about individuals, which is never *perfectly* true and often largely missing, especially from people who arrive on rubber dinghys or chuck out their passports on arrival. And in any case even carefully-screened individuals are bound to their background groups through loyalty to kinship networks and regression-to-the-mean effects, which we can't reliably predict at an individual level. It's not irrational or unjustified to fill in the blanks here with probabilistic information derived from an individual's background group, which is definitionally 'racist'.
The case for this can be made with or without reference to genetic factors, of course, but the case certainly gets stronger when genetics enters the picture.
This doesn't make, e.g, assimilation of some non-zero number of foreign migrants necessarily *impossible*, but to make it work you're talking about a multi-generational process with necessarily higher standards of scrutiny than would be required of native-born citizens to compensate for any background risk-factors, which... again, is definitionally 'racist'.
What changed during the Great Awokening versus the previous several decades is that expressions of racial animus against whites in the mainstream media became not just acceptable but fashionable.
I favor doing as much as is practical to get better information about individuals: e.g., colleges should go back to requiring the SAT or ACT.
But I am against adjusting scores for reasons of racial regression toward the mean. For example, sure, a black applicant who gets a perfect 1600 on the SAT the one time he takes it would be likely to average a little lower than an Asian applicant who got a 1600 the one time he takes it if they both took it nine more times. But let's not penalize him for that statistical pattern. He got a 1600. Good for him.
On the other hand, say your mother's car breaks down in the middle of a dark block and her phone isn't charged, so she has to decide in which direction to walk. Hanging out on one corner are a half dozen Asian-American teens and hanging out on the other corner are a half dozen African-American teens. I think she can be forgiven in this case for doing some racial profiling and walking toward the Asian-Americans.
I was referring to regression-to-the-mean (maybe a misnomer, I think Galton used the term 'regression-towards-mediocrity') as a phenomenon seen among the children of exceptional parents- e.g, the children of very short people tend to be taller, on average, and the same can be seen at both ends of the bell curve for any complex human trait, due to a combination of non-additive genetic and non-systematic environmental factors.
Notably, as you probably know, different racial groups regress toward different means for IQ (up until the advent of GWAS, this was one of the more powerful pieces of evidence for hereditarianism.) This is relevant to long-term migration policy, but also in the broader sense to any attempt at integrating exceptional individuals into meritocratic societies or institutions.
For example, the children of wealthy black parents do about as well on the SAT as white kids who grow up in trailer parks. And... I have to imagine this could strongly motivate parents of colour to push for the... shall we say, relaxation of meritocratic standards, when it comes to their child? Potentially true for any parent, of course, but more likely to be true the larger the regression effect. And this has to be a factor when considering whether to include such parents in the social milieu that decide standards of merit to begin with.
Looking back, I think that the definition I was using was from Omi and Winant. They added an "essentialism" component.
So that rescues most AA programs.
By way of example, let's say that the Nazis put all the Jews in camps and stole their property. You could, without being racist, say that the Jews should be compensated - not because of some essentialist characteristic of being Jewish, but because of their historical experience suffering in Nazi concentration camps.
As for the rest of it, you don't have perfect information to treat individuals the same. I don't run a background check on the people I meet, though I'm sure it would reveal lots of information about whom to trust and whom to shun. Instead, I just treat people like they're all normal, unless and until I get individualized data saying that someone is, say, a convicted felon, or something.
People make decisions based on imperfect information every day. Unfortunately, race is something that's really hard to not see!
I'm not sure I buy that AA programs can be justified on any non-essentialist / contingent grounds: the bar to clear for receiving benefits is "be black", right, not any feature equivalent to "experienced XYZ event or type of event"—or, in other words, if it were that being black was any sort of proxy, there'd be a better way to get at the thing we /really/ care about without need for a proxy.
(e.g., a middle-class black woman from a recent-immigrant family applies for some benefit under AA: well, she & her ancestors experienced no slavery, or none the U.S. was complicit in; she isn't poorer than other applicants; she isn't disadvantaged in any way relative to other applicants; etc... she's just black, and that's the sole criterion.)
Unless I'm wrong and AA in practice requires that one prove U.S. slave ancestry or other white American oppression against you & yours in particular, or something...? I think most proponents are against anything like this, though, even something like "just give help to poor people as a whole"—some have come up with various ways to try to justify making it race-based, but (if I may do some armchair psychologizing) I'd bet you $5 that nah, they see it in essentialist terms in their hearts of hearts. So to speak.
I understand your point, but the problem with allocating AA on the basis of, e.g, poverty is that poverty just doesn't predict social mobility across racial backgrounds, so it wouldn't have the required 'compensatory' effect.
So a couple points. First, I don't think your characterization of the facts is accurate. In California, where I live, the reparations task force said that reparations should be "based on lineage, determined by an individual being a Black descendant of a chattel enslaved person or a descendant of a free Black person living in the United States prior to the end of the 19th Century".
Second, with that fact in mind, I don't think the facts matter all that much for this discussion. I know literally zero people who would support (or oppose) a lineage-based reparation program but who would change their mind for just a race-based reparation program.
So I'll agree that most people who support AA/reparations are racist, albeit in a direction that runs contrary to the bulk of American history. I think that's why the people in charge of, say, California's reparations task force took pains to separate race from lineage, to avoid charges of racism.
Of course, that got them zero converts, because normal people aren't living based on what Omi and Winant were writing about. They're not trying to conform their lives to some ideology - they have vibes and make decisions based on those vibes.
But people who care about the ideology do, in fact, have answers to these questions. I just don't think they represent too many real people.
I agree that people injured through either malice or incompetence deserve compensation, but I also don't think you can just measure per-capita GDP disparities and use this to infer injury, so I want to know what level of compensation is meant to be sufficient. And I've written elsewhere why I don't think affirmative-action programs are an appropriate form of compensation under any circumstances.
If you mean 'treat the same' in the sense of 'receive the same nominal courtesies during casual conversation', I mean... sure, politeness doesn't cost much. But the risk/reward evaluation can become quite different if you're deciding who to rent an apartment to, or approve for a bank loan, or hire for a job. Once upon a time, you could, perhaps, rely on credentialling institutions or background checks or just asking face-to-face questions to fill in the blanks, but the left has increasingly made it *illegal* to seek out such individuating information, and has actively corrupted every credentialing institution to reduce the exposure of embarrassing facts.
Racism is like democracy or love--it has dozens of different meanings used in different contexts, and often even the speaker isn't aware when they switch from one meaning to another.
To be useful, we really need to distinguish the moral / social stuff from the factual claims. You can't learn anything about the truth or falsehood of a factual claim by its moral or political or social implications. It would be bad if blacks' lower average IQ in the US were due to genetic factors that couldn't be changed, just as it would be bad if burning the cheapest available fuels messed up the climate. But if you let "it would be bad if this were true" influence your evaluation of "is this claim true," you are sabotaging your brain.
I agree with that, although in some circumstances, it makes sense to circle back and check with your moral / social stuff before publishing the information. Privacy is an obvious example - if my social security number gets released, I'm mad about people obtaining truthful information about me.
Regardless, the bar to clear for "This is true, but for moral / social reasons it must be suppressed" is really, really, really high. I can understand people who want to suppress the study of racial intelligence, but I'm pretty skeptical of the whole endeavor (leaving aside the question of whether it does more harm than good, or whatever).
In that case it still would be bad to publish a false social security number, because that can impact a random person with that number.
A problem is that once you start lying about the truth, you have to expand that lie to deal with hard questions. And suddenly all kinds of useful science becomes off limits.
And you need other explanations for certain things, and in practice this often seems to boil down to racist conspiracy theories that are remarkable similar to those wielded against Jews with rather nasty consequences. So is it really preferable to have people believe those conspiracy theories?
Well said. Funny how no one ever makes this kind of argument to, say, deny the Holocaust or something. "But if you believe than that, then you must think humans are terrible and there's no hope for the future!" Well...maybe? Not necessarily, but regardless... "Therefore it didn't happen" Non sequitur! "and anyone who says it did is an evil person who needs to be shunned and banished from polite company."
Oh, so you claim there's "evidence" that the Holocaust was real? Why are you interested in that? Do you LIKE the idea of millions of Jews being gassed? Gross! What a vile and disgusting person you are!
What about something like "people of X race are stupid, but we shouldn't give people more or less resources just because they're stupid"?
And a somewhat related question: if you think people of X race are stupid, and that they should get disability benefits as a result, are you racist against X or for X?
You kinda have to pick your battles. By way of example, I'm a Dodgers fan. I think that Giants fans are awful and stupid. All right-thinking people agree. But the kind of discrimination that Giants fans have historically endured is pretty small potatoes compared to racial, religious, or gender-based discrimination. If we could stop allocating resources based on racial, religious or gender lines, we'd be doing really, really well. I'd be extremely happy to live in a country where race was just as likely to predict your future financial success as what baseball team you root for.
To your second question, I was remembering the definition of racism used by Omi and Winant and I think they snuck in some requirement of essentialism. So if you believe that people of X race are essentially stupid, and therefore they should get more resources, yeah, that's racist.
Of course, I don't see many historical examples where that ever happened, so I don't think there's any cause for alarm. If in the future racist people start being far too kind and friendly and caring to other races, I'm sure we can figure out a solution.
The question is, of course, /why/ race predicts future financial status. If, hypothetically, the answer were to be "because some races are better at deferring gratification and have more smarts, on average", ought we still engage in redistribution along purely racial lines?
I think we pretty much always end up better not using race as a proxy, regardless—you have to engage in various unintuitive contortions to rescue such a policy over "just help whoever appears to need it most" without worrying about the racial balance of the recipients—but, considering how many people are only too glad to engage in such contortions, and how academia looks upon those who find for the wrong side on questions like "is it mostly differential intellectual ability, or mostly evil whitie being all oppressive and shit?", I'm not hopeful for the future here.
I don't know that we do engage in redistribution along purely racial lines. But if we did - and we just let the chips fall where they may - I'm pretty sure we'd stop caring about the people in Ohio and let unlimited numbers of Chinese people immigrate here.
Strangely, the people who care the most about race and intelligence never seem content to let the people of West Virginia sink into deserved oblivion as the more intelligent Japanese and Chinese people move in next door and buy up everything.
Quite the contrary, they want to ban Chinese investment, like TikTok, rather than just acquiescing in the face of the Asians' superior brain power.
Would you be okay with getting rid of all race/nationality/citizenship distinctions and letting individual merit be the only factor considered for, say, housing, employment, voting rights, etc.?
If you think IQ differences are leading to differences in outcome that should be compensated for by welfare or something, how does race even come into that? You would want to do that by individuals, not by races--a black doctor or engineer doesn't need any compensation for having a low IQ, whereas a white or Asian janitor may need that compensation.
You can have *some* level of social safety net, but, but how do you propose to allocate equal resources to all people regardless of their own productivity? This was tried a couple times and didn't work out well.
Even the 'safety net' approach still requires a reasonable number of net tax payers, by the way, which in turn depends on the typical genetic endowments of your country's citizens.
Most redistribution works out great - human beings depend on redistribution to survive to their teenage years.
Literally no newborn human can support themselves. Not a single one. They depend on net contributions from someone - usually their parents. This strategy differs from almost every other animal or plant ever encountered. Yet human beings are the ones who've taken over the world.
I think it's totally fine to take care of our fellow human beings - after all, every single one of us was taken care of at some point in our lives.
There are many examples of social species that exhibit reciprocal aid and essentially all mammals and birds provide parental care, so this is hardly a unique human innovation.
Do you think that social animals are immune to Darwinism?
Nobody sensible is saying people of race X are stupid. If you can't distinguish "the mean IQ of people of race X is lower than the mean IQ of people of race Y" from "people of race X are stupid," you're basically doomed to talk nonsense in this whole conversation. Seriously, a vast amount of nonsense on this topic comes from people who can't keep the difference between the mean and the distribution straight in their heads.
The counter claim is usually that "race" is an arbitrary way of grouping people that doesn't map well onto real populations. You see the conceptual confusion Scott has in these articles in his treating of "black" as a natural kind.
I think you may be reading too much into it. That study looked at four broad racial categories and discovered that they could tell which of the participants was in what category by their genes— the participants being people who were pre-selected to be among those four categories with none of the participants "mixed-race".
African-Americans are a mixed race population, as are mestizo Hispanics. Discrete notions of race do break down if you don't have concepts like "mestizo" or "mulatto" (much less "quadroon"), which are more popular in Spanish than English-speaking settler colonies, but it's still a fact that there are genetic differences that multiple services can now use to portion out your ancestry.
I don't think that obviates the point, which is that the population breaks down into separate races as well genetically as it does socially; hence, it isn't arbitrary: these are real genetic categories, and the expectation would have been (and, previously, a common claim was that) "conventional" race /wouldn't/ map to genetics in this way. Compare the prediction for how well an actually arbitrary grouping would be sliced along genetic joints (e.g., "people who like lettuce" vs. "people who like spinach", or something, barring some SNP that turns out to mediate this by itself or whatever—you could probably eventually get okay at slotting participants into the correct category, but 99% would be surprising).
Cluster analysis supports this as well—the only arbitrary portion being how many clusters you choose, which many commonly misunderstand as somehow a relevant consideration—and it's also just what you'd expect when eyeballing the geographical population barriers (& hence structures) of the world (if something , like a mountain range or ocean, is in between two populations in different environments that thus don't do much mixing for X millennia, one would expect to end up with two genetically-distinct groups).
Is black, not the race Black, but the color black, a natural kind? If yes, why not the race? If not, how is Scott confused is he treats Black and black as scientific concepts and not natural kinds?
Black the color is the absence of reflected light, and basically every culture has words to distinguish black from white. As you go up in the number of colors, the percentage of cultures with distinct words for them goes down.
In terms of ancestral race, African vs non-African is one of the biggest divides, since all non-Africans are descended from a relatively small group of (presumably) east Africans who then appeared to have mixed with Neandertals in the Middle East before splitting again into west & east branches (with the east branch mixing with Denisovans). But supposedly even within Africa there are splits that go back even further between the majority of Africans vs Khoi-San or pygmies.
A complication is that certain Oceanians (such as the aptly named "Melanesians" as well as Australian Aborigenes) have been called "black", because they have dark skin. They aren't any more closely related to Africans than any other non-African, in fact their Denisovan ancestry makes them LESS related to Africans than Europeans are.
The point is that red, as distinguished from orange, will have many edge cases, and the dividing line between them will appear arbitrary. From this, it appears that either there is no Platonic ideal/natural kind that is “redness”, or there are infinitely many; either way as humans we rely on our concept of redness, somewhat arbitrarily defined. Yet there is a fact of the matter that apples are red, and bananas are not.
Similarly, the division of humans into races will be somewhat arbitrarily defined, but there is a fact of the matter that LeBron is Black and Messi is not, even if Black is not a natural kind.
You hear that all the time, but what are the implications?
For example, on the 2010 Census, Barack Obama counted as black and only black because that's what he chose to self-identify as (according to a White House press release). Of course, he had a white mother.
If you don't count as Barack Obama as black, would that make the white-black IQ gap smaller?
No.
We've got the DNA technology now to measure racial admixture to 3 decimal places. That's now been done with two databases of around ten thousand young people each for which we also have their IQ scores: the national ABCD database and the Philadelphia Neurodevelopment database. It turns out that the more white ancestry a self-identified African American has, the higher his IQ on average. The effect is not huge but it's definitely there.
In the US, the rule for being black for affirmative action purposes seems to be that you must have at least one parent who self-identified as a member of the black community and was socially recognized as a member of the black community.
So, you can't claim to deserve affirmative action just because one of your ancestors 1024 ancestors ten generations ago was black.
That actually works surprisingly well. We have strikingly few disputes over who is black for affirmative action purposes.
One issue would be if your parent was part black but decided to pass as white. Can you reclaim blackness? But, ever since the introduction of affirmative action in 1969, not that many people who could justify a claim to being black have chosen to self-identify as white. State and society now provide benefits to identifying as black, so most people who can reasonably claim to belong to the legally privileged race.
The rules for claiming American Indian ancestry are more specific for individual tribes and more vague for the whole race.
For being Hispanic, having one Hispanic grandparent seems to work, but several judges have felt that being 1/8th Hispanic was ridiculous.
I don't think there are genetic differences between races in any *meaningful* or *persistent* sense, only in a technical and temporary sense. Gene alleles aren't locked to races, but to individuals, and through combinations of intermarriage and geographical isolation over time, alleles that are common in individuals in one group can transfer into and become prevalent in another group, and over long periods of time racial groupings at one time can dissolve and resort into completely different racial groupings at another time (and in fact, by investigating ancient human DNA we have proof that this has in fact happened).
I think those alleles have persisted for a LONG time. You're right that intermarriage will result in recombination, but we're still far from panmixia (Brazil is probably the place where socially constructed race has diverged most form genetic ancestry, but most of the world isn't Brazil).
I'm not just talking about what will happen, but what has already happened. I highly recommend the book *Who We Are and How We Got Here* by David Reich. To quote the summary from wikipedia, "almost all human populations are mixtures resulting from multiple population migrations and gene flow. "
Yes, modern day populations are the results of even more ancient populations mixing. If you go back far enough it wouldn't make sense to talk of people in terms we use for modern populations. This doesn't change the fact that modern populations are distinct from each other.
Yeah, but you're talking about a process that takes centuries at minimum, if not thousands of years, and many of these large-scale migration+admixture events were heavily co-extensive with genocide (the Bantu expansion and Indo-European invasions, for example.)
I don't see how that really resolves the anxiety of anyone concerned with present-day race relations.
Why do people of African descent have dark skin? I always assumed it was a genetic thing, but if there aren't meaningful genetic differences between races, it must be something else.
No two real distributions of non-integer values will have the same mean value. IQ is a root summed square of a nearly astronomical number of genetic components and thus non-integer. Accordingly, every race has different mean IQ. So do men and women. Clearly the definition of "scientific racist" is not, "actually understands statistics." I think you need a new definition.
The formally correct response to your criticism "yes there might be differences but they are about 10^-8 order of magnitude and do not even align with observed phenotype differences"
Couple of nitpicks: if the genetic components were actually "nearly astronomical", with no systematic difference between the groups, the quantity of resulting from their averaged impact would actually have negligible variance (1/sqrt(N)). To get appreciable differences, you need either a small number of random unbiased factors (such that 1/sqrt(N) is not too small), or a bias. But it would be a miracle if selection pressures for intelligence and for the other effects of the involved genes had been identical for all population groups.
The question is, are the differences large enough to be reliably detectable, and relevant for practical purposes. That is an empirical question that cannot be answered from theoretical arguments.
I am not sure that systematic is the correct word there. I also believe that individual genes aren't exactly identical in their contribution (there is no reason to believe they would be so not 1/sqrt(N)). On the other hand, obviously the collection of genes present in various populations are neither identical nor randomly distributed and thus the variation in populations without heavy crossbreeding are not merely noise.
I was making the more narrow point that nobody should be surprised that there are variances between distributions (it's not Racist to observe fact). The presence or absence of racism shows up more in an individual's obsession and need to shoehorn it into every conversation.
Scientific racism is the idea that there are genetic differences between the races, and that these differences result in certain races being smarter or more law-abiding and that sort of thing. Basically, that it makes sense to judge people on their race. As opposed to things like the genes that control skin color being different so you can easily identify their race, but it's totally useless to figure out which of two people you should hire to do a certain job.
So, I understand the definition you're using, but isn't it kind-of odd that scientific racists like Sailer and Murray advocate for hiring / admitting people to school entirely on merit, whereas antiracists advocate for hiring/admitting to school based on race?
One last comment on changes: I could not find anywhere that any proponents of Lynn’s data had ever shown what it shows about changes. I had to download the record-level data (not the country aggregates they estimate) and calculate it myself. My sense is that the researchers who favor this data will probably just respond by saying: all the studies are noisy proxies for stable underlying traits, the apparent changes are fake, etc. And they will point out that we observe the least change in the best measured countries. My response to that would be: if you think these are all noisy proxies for a latent trait you should be calculated national IQs with a latent class model not what you are actually doing, and also, isn’t this a huge own-goal on the credibility of the data that you don’t even think your own selected estimates within the same country are properly comparable? If same country estimates can’t be compared, how can we trust the cross country estimates? But I will be interested to see what responses you get!
I published an article about 15 or 20 years ago for which I graphed all the changes over time of national IQ scores in Lynn's database to see if the Flynn Effect was closing gaps. You could seem some growth in scores in northeast Asia. But it mostly looked like the gaps were about the same.
I couldn't tell you about in the 21st Century.
But you might remember when the PISA test was fashionable in the press around 2013. But there isn't much change in results every three years, so the media has gotten bored with it.
Since you're here, are you really confident that the PGS score differences between ancient and modern europeans only imply a ~10-IQ point difference? Any selection pressures that impacted the genes used to compose a PGS score probably impacted the frequency of a bunch of other genes for intelligence we haven't identified yet.
Is a third grader IQ really as low as 60? I thought once someone can read and understand pictures, their measured IQ should more or less lock in, in this case to the average 100?
I think there's a normalization thing going on here. My kids took IQ tests at age four, and got scores much higher than 60, but they weren't the same IQ tests that you would give an adult.
So maybe a third grader scores 60 on an adult IQ test, but 100 on the version of the test calibrated for third graders.
I thought about that but I don't think third grader would be that different to adult (certainly compared to 4 years old). Is there any non-normalized data somewhere?
Not what you're asking for but maybe intuitively helpful is the original way of calculating IQ scores, which worked something like this:
Suppose IQ doesn't increase past age 16. Someone 16 years old or older that scored like a typical 9 year old on an IQ test would have an IQ of 9/16 = (0.)56
So typical 9 year old would be equivalent to an adult with an IQ of 56 in terms of test scores.
I think this sort of calculation was used to justify Marilyn Vos Savant's 220+ IQ for inclusion in the Guinness Book of Records before they removed that category. She supposedly tested at the mental age of a 22 year old when she was 10. This is all from vague memory though so take it with a grain of salt.
Yeah, but if IQ doesn't increase much after 16 that's obviously silly. Like if a 10-year old is as smart as the typical (half-senile) 90-year old, that would give him an IQ of 900.
I used an item response model to generate IQ scores from raw response data in the Louisville Twin Study. Subject to many limitations and assumptions, it looked like 15 year olds were roughly as smart as adults, and 7 year olds were ~2 SDs worse on the most age-sensitive tasks and ~1 SD worse on the least age-sensitive. So I'd expect 3rd graders to be more like 1.5 SDs worse overall (or 78 IQ on the adult scale).
I think a lot of that is emotional control, accumulated knowledge, and other "maturity" aspects. My understanding is in accordance with yours that a third grader simply does not have the abstract reasoning capabilities to major in, say, theoretical physics or something. But there are plenty of majors that I think a fairly (non-child genius) third grader could get through if they had sufficient maturity and self-control.
I was offered placement in college when I was in second grade grade (after the school "accidentally" gave me an IQ test instead of the test to skip the third grade they were supposed to be giving me)
I turned it down, which my parents were happy about (they were concerned it would affect my social development - I, meanwhile, turned it down because my understanding of college at the time was based on media, and it was that I'd have to live on campus with a probably-very-annoying roommate who would constantly be annoying me with having girls over.)
Could I have completed the degree?
Yeah, no problem. I was learning out of a college-level chemistry book a couple of years later without issue.
Regarding the abstract vs concrete reasoning and so on.
Smartphones are now, AIUI, pretty much ubiquitous in Africa. If you can use a smartphone to check the market prices in distant location or transfer money to vendors etc., which is again from my understanding, common for sub-Saharan Africans then presumably you are exposed to abstract thinking skills and probably from a relatively early age now. So I wonder whether more recent IQ surveys will show an increase in apparent IQ for these countries.
Right, rural electrification is changing Africa pretty fast these days.
In the earlier comments, I linked to a study that found an 11 point rise in kids' IQ scores in a rural area in Kenya in just the 14 years of 1984-1998 in which a lot of good things had happened to bring it into the modern world.
I've taught students ranging from about 70 IQ to future Ivy League grad students in the hardest of hard sciences and math, and I can affirm that in casual conversation, you cannot necessarily differentiate people with IQ's from 70-85 from those with IQ's well above that. Sometimes people with extremely high capacities seem almost disabled. One young woman (currently completing a Ph.D. in math from an Ivy League school) I'll call Dee seemed rather slow to me at first. She rarely spoke and seemed to have no particular reaction to literature even though she was "supposedly" reading it.
Then you try to teach them something . . . and everything becomes clear. A student of 13, whom I'll call Mel, who has an IQ of 70, struggles with second-grade level material that is easy for a bright seven-year-old. Future grad student Dee turns out to learn everything without being taught it at all, and adds some things I did not even think of myself, and does all of that while learning classical violin and mastering two foreign languages.
I gave group-administered IQ tests (the CogAT) and can tell you that students with high IQ's can overcome specific skill deficits with dazzling speed, but those with sub-normal IQ's plod along making at best a year's progress if they have heroically determined and inspiring teachers using the best possible materials.
I found it all rather bad news.
The other thing I found was that we could raise CogAT scores a standard deviation or so (about 12-16 points) with kindergartners, but after that, five points was a really good gain. And even that flattened out, eventually.
I wonder how much of the difference between "Mel" and "Dee" is due to an inability of Mel to compute the result of calculations, and how much is due to Dee having far better memory. I note for myself that I can understand mathematics perfectly well, but it just doesn't stick in my memory. I have to go and look up formulas every time I need to use them.
Are the sub-normal IQs failing to understand concepts, or are they failing to memorize them? If it's generally the first, that's a harder problem to do with the general architecture of their brain, but if it's the second we can perhaps conceive in the future of some kind of memory aiding medicine that strengthens new synaptic connections.
Of course not. But if you have to constantly retread the same ground, you end up not making a lot of progress. You certainly aren't going to be discovering anything substantially novel.
Being good at anything intellectual requires some kind of reliable short term and long term memory, and math absolutely requires this. Remembering formulae is just one example, and it's not a small one.
They fail to understand concepts and principles. They also fail to retain them, I suppose, but it is very difficult to remember what you do not understand.
"I note for myself that I can understand mathematics perfectly well, but it just doesn't stick in my memory. I have to go and look up formulas every time I need to use them."
I have a PhD in algebra. While having good memory is certainly advantageous, even professional mathematicians consult formula tables quite often. Bartsch's collection of mathematical formulas, a book I used almost daily during my studies, has 830 pages!
Memory is only an auxiliary tool, though. The ability to grasp ("grok") abstract concepts is what matters. If you don't understand what a polynomial is, you will struggle with every non-trivial task regarding a polynomial, even if you had almost perfect memory.
But if you understand just fine what a polynomial is, and then have to relearn it almost from scratch every time it comes up, you will make very slow progress on more advanced math.
I haven't met anyone whose memory would be SO bad...
I met a lot of people who "shallowly forget" things, including math structures, but as long as the time between work sessions that use those structures isn't terribly long, they refresh their knowledge much faster. I.e. the second, third ... etc. instance isn't really "learning from scratch".
If someone suffers from such a deep memory defect, they are basically disabled.
Well yes, I was using an extreme example to illustrate. But I think a big part of learning any difficult topic is how many repetitions it takes, and how close together they have to be, before you just *know* a concept off the top of your head and can easily use it to help you understand more advanced ideas. I don't think IQ tests can measure this - they measure short term memory, which is different. But maybe they don't have to because it correlates well enough with other aspects of intelligence.
"I’ve noticed that I cannot tell, from casual conversation, whether someone is intelligent in the IQ sense.
I’ve interviewed job applicants, and perceived them all as “bright and impressive”, but found that the vast majority of them could not solve a simple math problem. The ones who could solve the problem didn’t appear any “brighter” in conversation than the ones who couldn’t.
I’ve taught public school teachers, who were incredibly bad at formal mathematical reasoning (I know, because I graded their tests), to the point that I had not realized humans could be that bad at math — but it had no effect on how they came across in friendly conversation after hours. They didn’t seem “dopey” or “slow”, they were witty and engaging and warm.
I’ve read the personal blogs of intellectually disabled people — people who, by definition, score poorly on IQ tests — and they don’t read as any less funny or creative or relatable than anyone else.
Whatever ability IQ tests and math tests measure, I believe that lacking that ability doesn’t have any effect on one’s ability to make a good social impression or even to “seem smart” in conversation."
I also engage with teachers most of my working life and you can definitely tell which ones are the bright ones. Just like everywhere else in society. I'm not sure what she's on about here, it doesn't pass the smell test.
An observation I made when I was 8(?), is that lots of people get by with mimicry. They just run social-scripts. And most social situations don't really require much more than that. (When the script is thrown out the window, things suddenly feel "awkward"). The inability for some to think for themselves, when called upon, is something I've found frustrating for as long as I can remember.
And this plays into why I'm not really surprised that the current ML paradigm is able to pass a Turing Test by just aping what it sees online. Also, something something Heinrich.
I have a coworker who is a mathematician by training. He is very slow and deliberative, and you could easily mistake him for not being very smart because of that. This would be pretty badly wrong, but not every smart person is fast on their feet in a conversation.
That seems consistent with what I've read of the literature, yes. You can get impressive gains in really young kids, but heritability increases with age and there doesn't seem to be any way to make gains from intervention permanent. In developed countries, at least.
Was going to note that my aforementioned foster-kid who was a 63 was pretty normal (well, normal for an 11-year old, anyway) in conversation.
It was when he was doing his homework and he was struggling through the kind of homework my 5 year-old blows through fine at pre-school that the light clicked on.
FWIW, his sister (also a placement, not tested as far as I know), was notably not much better at reading than him; I'd guess she might have been somewhat better but not in any meaningful way.
"The part of Lyman’s comment that gives me the most pause is his observation that, if the mean IQ is 60, a decent fraction of people must be 45, and a non-negligible portion 30. At this point, even third-grader comparisons don’t save us. I guess this is where I bring in the claim that IQ breaks down as a guide to practical living skills below some point."
Or, what could break down is IQ being normally distributed. If IQ fails to follow a normal distribution below a certain level, we wouldn't expect the usual proportions of people to be 1 or 2 SD below the mean.
This is certainly true. However, non-normality could be instructive. If it is true that symbolic learning has a huge impact on IQs in Africa, then one might expect that rather than shifting a mean (and obviously with it sigma), it would effectively just make a significant number of questions unanswerable to most takers. This ought to show up in skewness with a much larger sigma for takers above the mean than below.
If IQ is bounded by zero on the left, it is not a pure normal distribution. Truncated normal distributions are much harder to work with mathematically.
That the average IQ is 100 and that the standard deviation is 15 are just arbitrary conventions, so there is nothing special about an IQ of 0. (Setting the average, which is the only special value in the distribution, to 0 would have made a lot more sense to me, but there are historical reasons that it is the way it is.) If IQ were perfectly normally distributed, approximately one person in 76 429 666 480 would have an IQ <0 (ditto for IQ > 200). In reality there are many more people with very low IQs than one would expect from a normal distribution, because there are single mutations that have a large negative effect on intelligence (whereas a normal distribution would be the result of a very large number of genes each with a very small effect on IQ).
> In reality there are many more people with very low IQs than one would expect from a normal distribution, because there are single mutations that have a large negative effect on intelligence
That only makes sense if you have some absolute way of measuring IQ. Given two smart people of slightly different intelligence and two dumb people of slightly different intelligence, how can you compare the differences? You could just give everyone a test and see how many questions they got right, but the distribution will depend heavily on the test.
I agree that the difference in intelligence between the lowest 0.0001% of the population and average is more than the difference between average and the highest 0.0001%, since the lowest 0.0001% is someone in a coma and the highest is still just comparing two people, but I don't see how you could define differences in a precise way that can reflect that.
Yeah, I can't argue with you there, actually. I do wonder what it means when people say that the IQ distribution has a thick left tail – like where does that aberration from normality come from when IQ is supposed to be constructed to be normal.
The idea that changes in IQ aren't correlated with changes in development makes total sense, because the changes in IQ in question are generally not reliable even though the mean values are. I also doubt his analysis was done correctly here, since he portrayed incorrect national IQ estimates. It's a confused point followed up by points that aren't even clear, like "By the way— a pure hereditarian would insist that African IQ is around 80-90. Why? Because in the highest-quality data on African-ancestry people, namely, tests in the U.S., African-ancestry people tend to score around 80-90 on average."
Does this refer to African Americans, who are 15-25% European by admixture or to the selected immigrants who go to the U.S.? If it's the former, then a hereditarian would surely insist that the African IQ is the AA IQ minus the delta from European admixture. Depending on where estimates sit in his range, that could mean going below 80 or sitting in 80-90, so it's just unclear what he meant. He follows this unclear point up with "If you’re really a strong hereditarian, the U.S. environment should not boost IQ very much, which implies that the national IQ data is incorrect" which, as far as I can tell, is just describing a case no one believes in, including Lynn, who explicitly rejected it on multiple occasions, including his debate with Wicherts (see: https://www.sciencedirect.com/science/article/abs/pii/S1041608010000348).
A lot of other people also made weird claims, so I'll write something on this.
Lyman is worth reading, as he grapples more with data than the typical pundit, but I wouldn't always say I find his conclusions reliable. For example, he seemed to be ruling out urbanization as an explanation for falling fertility when it was actually consistent with his data, and he just didn't have enough to be sure in the sense of statistical significance (which as Andrew Gelman never ceases of reminding us, is not the same as actually being correct) https://x.com/TeaGeeGeePea/status/1876509720707751980
Yes, I think this conclusion about the value of changes is particularly unreliable, since it doesn't deal with all the data and for the additional reason that it doesn't deal with the psychometric meaning of changes in scores.
As an example of the latter, I used the China Family Panel Studies to look at changes in Chinese IQs by age cohort and found that, adjusting for psychometric bias in the measurements, score trends over time were largely eliminated. Lyman's interpretation of this would be based on the raw scores, but the underlying ability level barely changed over time, so why should we expect those raw score changes to have any validity?
Maybe this is just a specialization thing? I can alphabetize like a professional but then I spent years in school learning the alphabet among other Rs.
But if you plopped me onto a subsistence farm in the middle of nowhere, I reckon I'd be pretty crummy at surviving!
I would dispute that SSA's subsistence farmers have especially optimised strategies, by the way. I occasionally get newsletters from a charity called Forest Gardens International, which routinely teaches farmers in west africa to triple their agricultural yields, even without any additional use of artificial fertilisers or pesticides.
People forget that pre-haber-process improvements in britain's agriculture resulting from enclosure and crop rotation more than doubled farmers' yields, and britons had been living off subsistence farming for thousands of years beforehand.
(EDIT: For the life of me I cannot google for Forest Gardens International any more? I could have sworn it existed, but maybe it was just Trees For The Future?)
That is not dead which can for four years lie, and with a new election even death may die.
Seriously, I just figured he was making more money off Substack than his day job. Or the NYT had already gone after him, so what else does he have to lose?
Strategically speaking, I can see one problem. He and a lot of other people around here are really worried about AI X-risk. (I am agnostic as I really don't understand the issues that well, but their concerns seem at least somewhat credible.) If you really think superintelligent AI is going to destroy the world, it would not make sense to associate yourself with one side of a controversial issue. You still have people on the left worried about it. If worries about AI X-risk become fully right-coded then that would make developing any kind of a consensus to limit or align AI properly much more difficult, and likely impossible.
If someone on the other side has to disavow anything associated with Scott Alexander, it's going to make it harder for her to argue for concern about anything other than AIs being racist in ways the left doesn't like. And then, if this AI X-risk thing is real, we're all dead.
"That is not dead which can for four years lie, and with a new election even death may die."
But it is hard to rekindle a fire from the ashes.
The fervor of 2021 isn't going to repeat itself anytime soon. People are too burnt out internally. In history, it is almost unheard of for a time of high ideological fervor to repeat itself soon, it usually takes several generations, if the ideology survives at all. Most ideologies die sooner that that. Some are very long-lived (traditional religions), but even there you see religious revivals every century or two or so. Not every five years.
Certainly the DEI bureaucracy etc. will live on and struggle to keep some grip on power, much like late-stage Communism worked (the parties left all the Stalinist passion behind and kept power for power's sake). But that is not the same as feverish witchhunts, and the overall climate is going to be a lot more tolerant of dissent.
The bureaucracy is still in place. The young (at least the women) and the elites in the Ivies and other citadels of privilege that staff the ruling class and their immediate underlings believe this stuff. PC was temporarily turned back in the 1990s and everyone on the right and many of the traditional liberals thought it was over and they came back with a vengeance in 2014 or so and gained a lot more ground.
> The bureaucracy is still in place. The young (at least the women) and the elites in the Ivies and other citadels of privilege that staff the ruling class and their immediate underlings believe this stuff.
Why do you think those are problems that can't be solved? Past governments in other countries have been able to force cultural and structural reforms with some effort. And a lot of bloodshed.
Well, DEI is going away because of a fear of judicial backlash. Once acting on woke (by introducing de-facto quotas etc.) runs the risk of being subject to a major lawsuit, even the academia will think twice. They have huge coffers and these can be raided for damages.
And without acting on woke, the system tends to hollow out.
Universities are woke reservoirs in the same way that monkeys are reservoirs for ebola. We may have fought off this epidemic but a new one is sure to rise eventually. It might not be for 20 years - sometimes it takes a generation for antibodies to fade.
It is unrealistic for woke to go entirely away on a short-to-medium timescale, but it is now in a much more vulnerable position than in, say, 2021.
Even in the academic sphere, there will be budgetary struggles soon, because the demographic development caught up with the universities and there will be fewer young people entering a vast system of higher education = less money overall.
I think we are looking at a future where woke exists, but has to compete against a whole team of other ideologies. The dominant position it held for a short time is not coming back.
Yes, I agree universities are in for a tough time and their position of cultural respect is definitely gone. It remains to be seen what the new equilibrium will look like but yes you do point out reasons to be optimistic. If I had Trump's ear I'd tell him to cut federal funding for any university that had either anything that looks like DEI or any grievance-studies dept (gender studies, chicano studies, etc). That would truly drive a stake through the heart of woke.
There will probably be some entirely new and exciting bit of numbskullery that takes over all the high-profile institutions for a few years that looks nothing like wokism/Kendism/etc.
Almost certainly so, fooling themselves is a popular pastime of intellectuals.
The best variant would be, if multiple competing creeds of this type emerged at the same time, with none achieving dominance and plunging the fanatics into an endless ideological civil war within academia.
<If you really think superintelligent AI is going to destroy the world, it would not make sense to associate yourself with one side of a controversial issue. You still have people on the left worried about it. If worries about AI X-risk become fully right-coded then that would make developing any kind of a consensus to limit or align AI properly much more difficult, and likely impossible.
That may be true, but I am pretty confident Scott does not think that way. He has an exceptionally strong drive to figure things out, and that accounts for most of the variance in what he says and how he says it. He probably does do some minor things mostly out of a desire to keep his audience happy, but they would be things on the order of putting up a humorous post now and then. He does not do the kind of political calculating you are talking about, and I don't think he would turn out to have much talent for it if he tried.
He's committed to saying and doing things (like this) that will make him persona non grata among the wokest parts of the left. That is not even close to being the same thing as "committed to the right".
In our stupid binary political system, that makes him a right-coded figure. The guy's a Democrat in San Francisco in a polyamorous relationship but the only guys who quote him are conservatives.
If we were multiparty he'd be in a 'liberal' (in the European sense) party like the FDP (Germany), Lib Dems (UK), Liberals (Australia), or Fine Gael (Ireland).
I should add that I believe he is telling the truth and I agree with him.
OK. but "right-coded" and "committed to the right" are not the same thing. If for some reason someone starts a smear campaign tarring you as a proponent and practitioner of traditional Aztec religious practices, and gets most people in their ingroup to believe that, that doesn't make you "committed to human sacrifice"
I don’t see why this commits him to the right now. He is clearly not engaging in motivated reasoning, but trying to think straight about the question of whether differences found in national IQ are valid. If the left rejects him for arriving at some right-compatible conclusions, maybe the left isn’t worth identifying with. Do they only want people who engage in left-motivated reasoning? Then they are mind-controlling jackasses. (And if the right embraces him they are in for some rude shocks when they learn his views about some other issues. I think they will soon spit him out.)
There are certain things a leftist isn't allowed to say, and he's said them. I absolutely agree they only want people who engage in left-motivated reasoning; that's why a lot of people are here, after all. Intellectual honesty is rarer and rare. And I agree about the right; he'll be one of these guys claimed by neither side. Which is better for a writer, after all.
Still, I've been in a lot of settings where it is clear you are supposed to think a certain way, but if you say something out of line with that, while staying friendly, and maybe saying something like "I know you guys don't think about it that way," it kind of wakes some people up out of their trance. Or maybe the effect is just from you providing a demonstration that not following the party line doesn't immediately get you punched in the mouth.
Even if Scott has made himself less appealing to the left, he may also have given some committed lefties permission to speak more freely.
To me, a critical thing absent in these discussions is that in the cutting edge of intelligence science, whether in AI or biology, these IQ things play no role. The entire area is apparently a dead field where decades of the study of "intelligence" has yielded basically zero actionable information relevant to anyone doing actual science of either making artificial intelligence, working on neuroscience, or, afaik, educational efforts aimed at human intelligence.
Think of stuff like PPP, global workspace, spaced repetition, even really dubious stuff like fmri studies on everything have helped out like 100x the rate of this field
Whether that's because they're all dumb-as-rocks racists who can't science their way out of a paper bag is I suppose mildly interesting but overall? The less time spent on basically anything related to this dead end of enquiry the better. As a tremendously useful thinker, Scott, I think maybe you've been nerd sniped by the controversial ness or something. It'd be more fruitful to explain why this dead field shows any promise whatsoever to uou of shedding the tiniest bit of useful information relevant to the understanding of intelligence at all, after decades of nothing at all.
Angry people wanting to stamp something out means it can't also be a trivial waste of time? Obviously it depends where you're measuring importance and it seems like you're wanting to conflate two different measurements.
Imagine this: Two overlapping circles. One contains "not obviously massively important" pursuits. The other contains "pursuits worth stamping out". Some pursuits are deemed to be in one or the other circles, some pursuits are deemed to be in the overlap.
If this seems paradoxical that's because, again, the notion of "importance" is being measured in two different ways. Building model train sets might be not very important. If building model train sets has become an addiction for someone such that they are neglecting taking care of themselves and their dependents, it could be worth stamping out. The model train addict could say "you can't tell me this hobby is not important because here you are putting so much energy into trying to stop me from doing it!" The confusion is because an activity's "importance" is being reduced to a single measurement, as if anything can only be important in a singular or general way. It's a simple semantic confusion.
Because it provides cover for and thus at least marginally increases the rate of people saying unkind things about black people, and they are very much in the habit of shouting down people who say unkind things about black people.
And because they think this line of inquiry has *literally no other value* than facilitating unkindness towards black people. I'm pretty sure they're wrong about that, but most takes on the matter that aren't done with Scott Alexander levels of rigor and charity do a poor job of arguing that point.
So, if the subject comes up, they're going to try to shout it down because they have one possibly small reason to do that and no reason to take any other position, and because while they might prefer to be discussing something they feel is more important, someone else raised this subject so that's what is being talked about right now.
And nobody is buying the babe in the woods routine.
Despite any valid inquiries or insights into improved public policy that could arise from understanding the limits of environmental interventions, the primary reason many people interested in this subject engage with it and seek to mainstream it is that it offers what appears to be a rational basis for racist discrimination and stereotyping.
People excessively focused on group differences in trait distributions are not especially dedicated to combatting stereotypes or prejudice or treating individuals based on their personal merits; they want more justifications for their existing biases. They're engaging in motivated reasoning.
This is another attempt in a long line stretching back hundreds of years to rationalize oppressive social hierarchies delineated by race as being natural and rational. Before genes, it was calipers, etc.
Of course, others are driven by different interests or simply by curiosity about topics restricted by social taboos.
But please, let's not pretend that there isn't a significant number of racists out there eager to promote seemingly objective arguments to justify their racism and degrade the taboo against it.
Do you mind explaining what exactly was so crazy about the 'calipers' routine? Modern MRI studies have repeatedly turned up racial differences in brain structure, and Samuel Morton's cranial measurements were re-examined a few years back and found to be essentially accurate. Heck, police forensic teams have been using skull measurements to identify the racial background of murder victims for decades, it's about an 80-90% accurate technique.
What do you think the antidote to racial stereotyping is supposed to be? Is our only contact with other races supposed to be through the medium of marvel movies, and not through glancing at FBI crime statistics or, I don't know, having your eyes open when you walk around California?
The attempt to eliminate racial discrimination is like any other socialist shakedown racket- you can't actually achieve equality of opportunity without achieving equality of outcomes, and trying to achieve the latter is both impossible and insane, so the grift never ends.
Do you think afro-american communities have not been sufficiently compensated for whatever historical wrongs they might or might not have suffered? A civil war, decades of affirmative action, and trillions of dollars sunk into welfare subsidies, business grants and housing projects aren't enough to get whites out of the doghouse?
If not, what is your target here? Afro-americans have had the highest material living standards of any black population on earth for literally centuries, so... what does 'good enough' look like, exactly?
If you make the totally unrealistic assumption of a world with zero racism and zero historical injustice, and still grant IQ diffs and HBD, what do you think their income levels, family structure and homicide rates would be, precisely?
Many studies, such as those of Gregory Clark (referenced in the following link) find that status is remarkably consistent and reverts quickly after exogenous shocks. Consistent with that, the causal impact of slavery on descendants of slaves disappeared by 1940 (https://www.cremieux.xyz/p/black-economic-progress-after-slavery).
For that reason, it's unsurprising that when controlling for IQ, the Black-White gap in socioeconomic mobility is usually eliminated or reduced to a pittance (https://www.cremieux.xyz/p/whats-in-a-black-name).
To be clear, I'm not suggesting that the only discrimination against African Americans was slavery, but rather, that just as the impacts of the most profound form of discrimination faded out within a few generations, the impacts of later lesser forms of discrimination faded out as well, likely even more quickly, as is typical for many groups all around the world, as discussed in the first link.
Each different bad outcome will have a different mix of causes. The performance gap in education probably has a very different set of causes than the black/white difference in rate of unwed birth, or the wealth gap, or the difference in voting patterns. There's not going to be some one-parameter model that explains everything, whether you label your one parameter "racism" or "IQ."
This could be more imaginative. <Bad thing X> might just be something we can't easily fix by direct intervention. For example, we're currently not very good at predicting earthquakes, but that doesn't mean someone has to be blamed for the earthquake problem. Similarly, maybe we're not good at fixing race disparities, and there's just nobody alive worth blaming. Sandra Day O'Connor thought things would equalize in ~25 years; predictions like this are based off of some very optimistic assumptions, and it might be less wasteful to devote time to that rather than IQ.
We blame earthquakes on plate tectonics, which is in fact an accurate description of how they occur, so I don't see how this analogy is useful.
Genetics/IQ is the actual explanation for most racial outcome disparities, and every other explanation is going to lead to faulty predictions and empty promises. I don't see the benefit to this aside from the comfort of cowardice.
In this analogy, the problem is our inability to predict earthquakes, not just earthquakes itself.
The benefit to this aside is that you can earn easier victories against leftist overreach with good arguments that don't alienate people. It's so facially absurd that the IQ obsessives think race realism has any persuasive weight in a world where (a) most people are viscerally revolted by it and (b) most race realists would, frankly, have been persuaded by much less evidence. With the way IQ frequently gets shoehorned into discussions where it's dubiously relevant - often I suspect to soothe the self-esteem of less impressive whites - it's clear that the scoreboard isn't looking good, and the project to normalize race science has been a pitiful failure. So, to circle back to my original aside: just be more imaginative.
My dude this crew has contributed zero to cognitive testing. They use these tests but as I said have no theoretical perspective on cognition and so make no contributions to them.
Also, I am not saying they should be stamped out, I'm saying they should be ignored as a dead field.
It's notable that the Sailer's, Hanannia's, and Murray's tend to be trained as political scientists, journalists, and even marketers rather than, say, statistical geneticists. I stumbled on a post where Hanania agreed with Murray about not being quite smart enough to understand linear algebra and was dumbfounded that ppl like these confidently pontificate about IQ, which is essentially a statistical and correlational measure. They also (scientific claims to the contrary) tend to be aggressive social activists with crude eugenic agendas.
No wonder they keep confidently repeating dumb things. Seriously: will you clowns stop assuming a person or group's "genetic IQ" can be determined from a linear combination of the IQs of their ancestors or groups?
There are a number of HBD advocates that clearly are well trained in statistical analysis (Kirkegaard and Cremieux, for example. Even David Reich is at least mildly red-pilled on the topic, even if he doesn't bang on about IQ diffs much in interviews.)
"Well trained" by what standard? Hanania believes he's well trained because he took some "methods" classes for a nonsense UChicago PHD but by reading his writings (including his own PHD thesis) his understanding of statistics is below the level of a good undergrad. Ditto Kierkegaard and Crimeux (his real identity is known). No idea who Reich is.
In short: a bunch of people too stupid to make real contributions to science, making completely obvious mistakes (JFC these idiots STILL confound hereditability and genetic influence), claiming that it is a mere conspiracy the overwhelming majority of experts disagree with them, and having Dunning Krueger geniuses like the OP for backup.
You can dig up youtube talks where he mentions 'interracial genetic differences are only about 1/6th as large as the average differences between individuals, but 1/6th is not nothing' and I believe he co-published a recent paper documenting a ~2 SD increase in EA-PGS among europeans over the past 10,000 years. He knows what's going on.
I'm pretty sure Kierkegaard and Crimeux are above 'undergrad' level, or I think Scott or Charles Murray would have pointed out their elementary errors by now.
Emil is an idiot about a broad range of issues. I don't know Crimeux so perhaps. Murray is completely unqualified to have any opinions on this topic. So far Reich is the only credible person you've mentioned, and as you said yourself it's not clear he is part of the HBD sect.
Your standards of scholarship are basically no standards at all. No different than any other group of conspiracists whose worldview depends on elevating the few pseudo scienitific cranks that agree with them and ignoring the overwhelming evidence against.
I'm hearing a lot of ad hominem, and your credentialism is essentially a kind of circular argument. You can't trust the 'racists' because they're not credentialed, and they don't get credentialed because they're 'racists'. You know this perfectly well.
I mean, if this isn't good enough for you, John Rawls was pro-eugenics and both James Watson and Francis Crick are basically on-record agreeing that the B/W IQ gap is mostly genetic. It's hard to get better credentials on the topic of heredity than those two.
> as you said yourself it's not clear he is part of the HBD sect
No, he's definitely said that the egalitarian position is untenable, he's just being very... cautious about the magnitude of expectable racial differences in socially-consequential traits. (Pygmies and maasai differing in height by far more than 1/6th of a standard deviation does not bother him, apparently, but this is obvious to anyone who thinks about the topic for more than five minutes.)
Entering a discussion - None. Making scientific claims - The competence to at least understand the science. Trying to drive social outcomes - extremely high!
All of these "race scientists" make social and scientific claims about topics that are (despite their claims to the contrary) extensively and even exhaustively studied. You will note that Murray made predictions in "the bell curve" about the impact of dysgenics on American society. 30 years later he was wrong about basically everything.
"We believe X based on these 40 year old studies that constitute 1% of the field and think all of the experts in the field are simply wrong" is certainly a view you are entitled to have. You are not entitled to have people take your opinions seriously.
And yes - these people are trying to drive social outcomes. Scum like Hanania have argued for race based immigration policy (including mass deportation not based on legal status) and outright eugenics. I think I am correct for having a high burden of proof to take their ideas seriously
Eugenics is completely inevitable regardless of what you think about HBD, by the way. There is no genetically neutral human society and we can't ban deleterious mutations.
Natural selection is inevitable. Formal programs to try to improve the genepool are not inevitable--we don't have them, most societies haven't, and the ones that have existed mostly didn't have much effect.
Which ideas? You should be skeptical of race-based immigration and coercive eugenics policies because those are bad policies, but I don't think their badness turns on empirical questions about IQ distributions across racial groups. When Sailer, Hanania, or Murray make empirical claims, I think we should evaluate those on whether or not they are accurate, not on whether or not we think those guys are smart enough or have the right credentials to discuss the matter.
Tons of people talk about IQ, race, the performance gap, the causes of differences in outcomes between blacks, whites, hispanics, and Asians, and similar topics. Hardly any of them are experts in statistics, psychometrics, and genetics.
Do you also decry the ones who aren't experts in these areas, but who express confident opinions that the black/white gap in school performance is due to structural racism, prejudice among educators, or worse childhood environments for blacks? Because if you don't, it seems to me that you are engaging in an isolated demand for rigor--holding your side to a much lower standard than the other side.
>Which ideas? You should be skeptical of race-based immigration and coercive eugenics policies because those are bad policies, but I don't think their badness turns on empirical questions about IQ distributions across racial groups.
The people that support these policies use arguments about IQ distributions to justify them, so they certainly think their "goodness" rests on these questions. FWIW these ppl are not disinterested scientists either - they tend to be deeply committed bigots. I blocked Hanania for just being arbitrarily nasty to black ppl before I realized he had any scientific theory behind it.
>When Sailer, Hanania, or Murray make empirical claims, I think we should evaluate those on whether or not they are accurate, not on whether or not we think those guys are smart enough or have the right credentials
Their empirical claims are based on 50 year old science (twin studies) whose validity is effectively debunked. They claim that the debunking is one big PC conspiracy by the entire scientific establishment against them, the real truth knowers. It is therefore important to point out that their grasp of statistics is weak, since the scientific debate is essentially one about causal inference (statistics!). If you're interested Turkheimer's "Understanding the Nature-Nurture Debate" is a solid recent book.
> Do you also decry the ones who aren't experts in these areas, but who express confident opinions that the black/white gap in school performance is due to structural racism, prejudice among educators, or worse childhood environments for blacks
It depends on how confident their opinions are, what evidence they bring to bear, and the costs of their proposed remedies. Some of the claims are hand wavy, but we know that the U.S. government for centuries conspired to impoverish its black population, and that deprivation leads to worse educational outcomes. You can of course counter about the relative performance of blacks vs. whites/asians in the same SES. I suspect a lot of this is culture and can be teased out by within group comparisons. In any case, it is very likely multi-factorial and we can study and propose interventions.
My objection to the race scientists is that they claim massively overconfidently that the primary cause is genes and then propose solutions that happen to align with their personal racial hatred: keep blacks or w/e race I don't like out of this country, out of key professions, etc. They are also intensely arrogant - think of the God tier superiority complex it takes to propose eugenic solutions that treat other humans as mere rodents to be tinkered with. This is why I mock Hanania and Sailer - the arrogance is bizarre from people that aren't even top tier intelligences.
The SJW structural racism view at least involves solutions that are likely net positive even if you dispute the ROI: more support for early childhood education, nutrition, etc. When they propose solutions that are harmful (increase segregation, stop advanced classes for students that can test into them, etc) I object. The only solution the race science chuds ever propose is "do racism".
Also I don't know if asking you not to make third grader assumptions (A person's traits are a linear combination of traits from these poorly defined genetic groupings) is much of a demand for rigour.
Poor countries only rarely take the big global tests like PISA and TIMSS, but there are regional international tests in Sub-Saharan Africa and Latin America. Scores on those tests can be converted to PISA-like scores using the example of when countries that take the regional tests took the big global ones. It's much like Sean Reardon's Stanford database of school achievement test scores from every school district in America that are harmonized versus how well their states did on the national NAEP.
The one sub-Saharan country that did really well in the World Bank's database was Kenya at 455 on a scale where 500 is average for a First World country.( Gabon also did well, but that's an oil country, so I don't know who sat the test - Exxon engineers' kids from Texas?).
Trustworthy? Kenya's scores bounced up and down a lot. I looked at the report on the highest scoring year which was like a northern European country. The report wasn't obviously faked. If it was fraudulent, they put a lot of effort into it. So, I dunno ... I don't think I believe the score, but I also think Kenya is one of the better countries in Africa.
> Maybe I should have had a stronger opinion on whether Lynn’s exact studies were correct?
Your title was "How To Stop Worrying And Learn Love Lynn's National IQ Estimates" which even with the degree of irony implied by the reference reads as saying they are substantively correct. Rather than your actual position which seems to be that they are unreliable, but it doesn't matter because of other data.
Replication is crucial in science. If Lynn's data didn't replicate, it would be worthless. But, despite Lynn's occasional sloppiness, it largely has been replicated multiple times since 2002, although, admittedly, Wicherts et al argue that African IQ scores are about 1.3 standard deviations below Europe rather than Lynn's estimate of 2.0.
Africa, obviously, is the hardest continent to get reliable data out of, so I'm not going to go out on a limb and argue for either Wicherts' or Lynn's African estimate. I'd lean toward Lynn's because I think nurture can matter a lot in IQ, whereas Wicherts' estimate that Africans score almost as well as African-Americans (whose nutrition and health care is good enough for them to be likely the world's best athletes -- Lamar Jackson, for example, isn't a stunted individual) suggests that variance in IQ is almost all genetic, a position that I, being a congenital moderate, am averse to admit.
Also, the low-hanging fruit in making your people smarter are probably nutrition and disease, and those are places where sub-Saharan African countries have a lot of available improvements. Raising the average IQ of the next generation of Togoans by 10 points is probably *way* more feasible than doing the same for the next generation of Germans.
It's really interesting to see how some Americans see the world......
In Africa we like staying alive and in good health whatever your IQ as measured by whatever test. Obviously improving health outcomes is important - including improving nutrition. It's especially important for infants and children. Fortunately the literature and data on how stunted growth impacts overall health and education outcomes is solid and has been researched by credible scientists. So far there has been steady but slow progress. You'll be pleased to know that iodine supplementation was a subject of public health campaigns in countries like Zimbabwe more than three decades ago. You cannot sell salt which is not iodised. Despite high levels of extreme poverty, salt is affordable to everyone. I believe Malawi passed similar legistation more than 25 years ago.
Secondly, we want children to have access to schools and to stay in school and finish their secondary education. (There's another discussion to be had about improving the quality of education in those schools.) Again, there has been slow but steady progress on education-related metrics.
There is no government trying to improved 'average IQ', but they have been working to improve health and education outcomes which might incidentally improve Mission IQ to your delight. Thankfully we believe our children are not inherently inferior to children in other parts of the world. Given the right interventions, they can achieve their best outcomes. Not every one of them will make it to Med School or Eng/CS in our modest universities, but they can go for vocational training. If you find any groundbreaking, useful insights backed by decent data in IQ-Land, we'll consider it.
My impression is that there's a package of stuff that we know how to do to improve how well everyone does, in terms of physical health and brain development and all that. Those are interventions that do seem to raise average IQ in places that do them, and also lower infant mortality, raise birth weight, raise life expectancy, etc.
That's basically the stuff you're talking about--as your society gets richer, you can make sure everyone gets enough to eat and doesn't have any deficiencies due to (say) low iodine in the local soil, eradicate most of the endemic diseases so you don't have kids supporting a heavy parasite load while developing, get everyone their shots, make sure everyone has access to some kind of education, etc. This is the environmental version of "catch-up growth," where poorer countries can very quickly improve things because they can benefit from other countries' experiences.
In the early 1900s, there was still endemic hookworm in the US South, and Army recruits from those areas had lower average IQ scores than other regions. The US only really got on top of this about a century ago.
"And in fact, economists routinely find that African small-plot subsistence agriculture is actually highly optimized; farmers make very precise choices about where to plant which seeds, which fertilizer to use, etc."
This was totally untrue where I worked in Ghana. Economists struggled to work out why African farmers don't use improved seed varieties, techniques or chemical fertiliser when it would improve their outcomes, and why they used labour inefficiently. The only question is whether this is for mainly conventional economic reasons - individual farmers don't get to keep enough of their earnings, their husband spends the earnings on booze etc., or more "behavioural economics" reasons - farmers fall prey to biases, short time preference etc.
Intelligence didn't enter the official conversation, of course, but, as with every profession, it's obvious that smarter farmers do better.
My vague impression is that Midwest American farmers these day face cognitive challenges similar to, say, corporate vice-presidents. To be a good corn farmer these days, it helps to have a STEM bachelor's degree and an MBA-like degree in farm management. Farming has had gigantic productivity improvements, so American farms are now typically managed by people who are the best farmer of their siblings who are descended from the best farmer among their grandparents' generation. So, Midwestern owner-operators tend to be really good at what they do.
African subsistence farmers haven't been through that selection process to the same extent.
To attempt to offer a synthesis of Lyman's thesis that African farmers seem smart about their business and your antithesis that they don't learn well, I'd suggest that most people have intelligent things to say about their current jobs, but that IQ plays a role in whether they can easily learn new ideas. The American agriculture industry has come up with a huge number of new techniques. It would be great if African farmers could comprehend these breakthroughs well enough to apply them in their own fields, but that's asking a lot.
Speaking of African farmers learning new techniques:
"At the beginning of the seventeenth century, the Portuguese transported manioc from South America to West Africa for the first time. They did not, however, transport the age-old indigenous processing protocols or the underlying commitment to using those techniques. Because it is easy to plant and provides high yields in infertile or drought-prone areas, manioc spread rapidly across Africa and became a staple food for many populations. The processing techniques, however, were not readily or consistently regenerated. Even after hundreds of years, chronic cyanide poisoning remains a serious health problem in Africa. Detailed studies of local preparation techniques show that high levels of cyanide often remain and that many individuals carry low levels of cyanide in their blood or urine, which haven’t yet manifested in symptoms. In some places, there’s no processing at all, or sometimes the processing actually increases the cyanogenic content. On the positive side, some African groups have in fact culturally evolved effective processing techniques, but these techniques are spreading only slowly."
> American farms are now typically managed by people who are the best farmer of their siblings who are descended from the best farmer among their grandparents' generation
Also, farms have massively consolidated, which largely involved the worse farmers being bought out by their better-at-farming neighbours
"It's hard for people from literate societies with tons of exposure to text & graphical representations to see the extent of the gap."
I think the growth in our exposure to text and graphical representations due to the ongoing Information Revolution going back to Gutenberg accounts for some of the Flynn Effect. Early IQ test designers in effect anticipated which way daily life was going to move, which is why raw IQ test scores went up all over the world over much of the 20th Century:
Question for those who have looked at the meta-analyses more closely:
Do we get a sense of how much control there is across testing: facilitation, environments, etc?
Also, have there been any studies looking at how cultural values and, more interestingly, personal attitudes, might correlate with test scores -- i.e. a pre-test survey along the lines of "how do you feel about scoring well on this test? how much effort will you put into answering questions correctly?" etc.
> Africa is extremely genetically diverse, but I think most of the countries measured in the paper, including Malawi, are some variety of Niger-Congo speakers, who I don’t think are that much more diverse than white people or anyone else. The really interesting African ethnicities, like the Khoi-San, don’t show up as much at a national level.
Languages don't follow genes. Both Indians and Brits speak PIE-derived languages, and they're not super genetically related; inversely, Turks only have a 21.7% East Asian (incl. Central Asian) contribution, because most of their ancestors are Anatolian Greeks with the Mediterranean genes. In Africa, you can even see this visually - the Twa are a pygmy people who speak a Bantu language, so obviously their short genes didn't get swamped out by Bantu immigrants. The spread of Niger-Congo languages came with the spread of Niger-Congo culture, technology, etc, but that doesn't mean they overwhelmed of the native populations' genes. IIRC, in Cabinda they did, but mostly African populations' genes are substantially "ancestral," i.e. predating the Bantu expansion.
In the last comments thread it was explained why that is incorrect. Most people in SSA are descended from the Niger-Congo population who replaced (killed, out-bred, genocided) hunter-gatherers across most of west, east and south Africa. There is not much phenotypic variance within SSA populations or between them and their (17%) admixed American descendants.
No, they interbred with the hunter-gatherer populations as they went. There is significant admixture with locals among basically all Bantu-speaking populations, which can and has been identified. They did not engage in some kind of proto-anti-miscegenation law in the year 2000 BCE.
I’m not informed about the genetic and linguistic makeup of the Niger-Congo people, but in prehistory it’s relatively common for an invading people to almost completely genetically replace the preexisting people.
Off the top of my head there’s the Dorset culture (which has literally no genetic presence in the modern Inuit), the Mesolithic European population replaced by the Anatolian farming peoples (which has very little genetic presence) and the Jomon of Japan (who only represent ~10% of the modern Japanese genome depending on the area).
I don’t know about the Niger-Congo expansion, and how much interbreeding actually happened (I’m sure it varied dramatically), but I wouldn’t be at all surprised if there was quite a bit of genetic similarity between those people, as you typically don’t spread a language without spreading at least some of your genes in ancient history.
"GDP only depends on the real IQ", hmmm perhaps reread Henrich:) as not all key traits leading to successful societies like obedience, conformism, cooperation, etc, are correlated to intelligence (some even against it:). Dr Jan van de Beek did interesting research into which immigrants in the Netherlands had a net positive contribution.to its society. The cultural indicator proved to be defining. >
1. I wonder whether there might be a "spare capacity effect"? If you're using all your wits to survive your brain has rewired to do that and you don't have much spare capacity to handle new problems.
2. in farming (subsistence or otherwise) you're used to thinking long term, you have a small number of very important decisions with plenty of time to think about them. That doesn't lead to a brain that will do well in IQ tests.
3. What psychologists used to call rigidity (maybe they still do). If your culture is very conservative you're probably programmed to resist new ideas and have never learnt to solve problems creatively.
> but that sometimes you can get IQ through other means (like an extreme pro-education push by a poor country), and this doesn’t itself cause development
But why wouldn’t it? If levels of education improve cognitive ability is that what matters, even if the general IQ stays the same. Ability at intellectual tasks is the ultimate prize, regardless if the cause is good IQ hampered by bad education, or bad IQ improved by good education.
Keep seeing mentions of an article(s) in "Aporia".
For those of you unfortunate enough not to be European entomologists, I should point out that Aporia is the scientific name of the butterfly, Black-veined White.
I'm fascinated by the global iodine map, I thought that surely the results for Ireland had to be wrong - we're an island! Fish on Fridays was the norm in the hey-day of Catholicism! Seaweed was used as fertiliser!
"Dr Mullan said this is important for “our women and our babies” as iodine deficiency is a contributor to low IQ. A 2013 analysis of results from the Avon Longitudinal Study of Parents and Children showed a link between the low IQ of children and low iodine levels in their mothers’ first trimester, she said in her lecture.
High levels of iodine are often linked to a good intake of milk, yoghurt, and white fish. However, changes in farming practices, the dissolution of the milk marketing board in 2002, and a rise in “alternative” plant-based milk add to the decrease in iodine levels, according to Dr Mullan.
Cow’s milk intake has declined in women and girls, and “young women don’t drink any more than a quarter of a pint per day”. Plant-based milk alternatives are much lower in iodine than conventional and organic cow milk. There is also low availability of iodised salt in supermarkets on the island of Ireland, according to a 2019 study from the European Journal of Clinical Nutrition, which Dr Mullan was involved in."
And this really intrigues me: breathing in air from seaweed increases your iodine intake. Clearly, I was smarter when I lived right beside the sea as a child 😀
"Results for UI from both schoolchildren and adults in the present study confirm that living in a coastal environment does not automatically ensure adequate iodine intake, as the apparent iodine deficient status of the Irish population might not have been predicted on an island where few live more than 200Km from the sea. Iodine deficiency has been reported in many coastal areas while recent studies have shown that coastal communities on the Atlantic islands of Madeira and the Azores have median UI values lower than on the Portuguese mainland. These islands support very little seaweed growth (Limbert, Private Communication). Thus the conventional wisdom that iodine intake in coastal regions exceeds that in inland areas needs to be re-evaluated. What appears to make a significant contribution to the coastal iodine environment from this study is the presence of abundant seaweed growth as evidenced by the higher median UI values observed in both schoolchildren and adults living in coastal areas close to abundant seaweed growth (hotspots). The most significant manifestation of this was the % of individual UI values >150µg/L which was 45.6% in schoolchildren and 43% in adults living near a seaweed hotspot compared with lower prevalences (2.3-16%) in subjects living in inland or low seaweed abundant coastal areas. While factors other than living near seaweed hotspots may explain the differences in UI observed, the findings suggest or at least do not exclude, the possibility of seaweed derived gaseous iodine inspiration through respiration making a contribution. Consumption of iodine rich foodstuffs might have applied when food consumed was almost exclusively locally produced but is less likely in the era of supermarkets with widely sourced products. Thus the findings in this study do not support the thesis that living in close proximity to the sea is in itself a guarantee of adequate iodine intake. This may require a seaweed rich environment, allowing for intake by respiration of seaweed derived gaseous iodine. Possible confounders to the results include the failure to obtain detailed dietary information and data on possible exposure to seaweed derived gaseous iodine on all subjects and findings are therefore dependent on the geographical location of sample collection."
So bring your kids down to the beach where there is abundant seaweed and let them play amongst it for the benefits of gaseous iodine!
While I do accept that there is a thing we call "intelligence", which we can measure, and stick fancy labels on (whatever 'g' turns out to be), I am coming more and more to the opinion that IQ tests are measuring a plethora of things, which get confused.
Yes, one thing they're measuring is something we can roughly call intelligence (though I suspect it's more 'mathematical capability' and we can have a whole other fight about "does 'being smart' mean 'good at maths' only? and what does that mean for times in history when we didn't have or need the mathematical talents for, say, computer programming which are what is now deemed a measure of being smart, successful, and rich? What about the coming of AI, which will take over those mathematical tasks for us and do them even better than any human could, will humans then be deemed 'stupid' even with the same IQ results?")
But I think another thing it is measuring is "how good are you at taking tests? how practiced are you?"
One comment on the original post talked about the Raven's Matrices style tests and that the solution to one question was " Like I recall there being a problem that ended up being a geometric version of a square root."
Now, my reaction to that was "You can *do* that?" (and hard on the heels of that, "*Why* would you do that?") but mainly, you could sit me down in front of that question from now to Doomsday and (speaking as a white chimpanzee, thank you Charles Kingsley!) it would never, ever occur to me that the collection of lines and black and white circles was representing a square root. And I've *heard* of square roots.
What if you're living in, yes, Malawi and you never heard of a square root in your life? Maybe the really smart can figure out the pattern is "if this blob is X, and that blob is Y, then the next blob should be Z", but if you're having to invent the entire concept of square roots from scratch, I don't think you'll do well on a timed test.
Conversely, if you've trained on these tests and can recognise the shortcut that "oh yeah this one is about square roots", you're going to score better.
I wonder if most of us would score well on a test asking us, for example, "which of these pictures of cereal crops is sorghum?" but a Malawian subsistence farmer would ace a test about crops and planting times. Who then would be the fools and who the smart?
"The main economic products of Malawi are tobacco, tea, cotton, groundnuts, sugar and coffee. These have been among the main cash crops for the last century, but tobacco has become increasingly predominant in the last quarter-century, with a production in 2011 of 175,000 tonnes. Over the last century, tea and groundnuts have increased in relative importance while cotton has decreased. The main food crops are maize, cassava, sweet potatoes, sorghum, bananas, rice, and Irish potatoes and cattle, sheep and goats are raised. The main industries deal with agricultural processing of tobacco, tea and sugar and timber products. The industrial production growth rate is estimated at 10%."
It would seem that Malawi also lacks mineral resources, the soils are depleted, and rainfall is unreliable, with a period of good rainfall (and thus adequate and even abundant crops) later replaced by droughts (and hence shortages and famines). I think all of this has to have an effect on the population, and we need to clear away the effects of starvation, malnutrition, poverty, lack of schooling, etc. before we can even get to testing "are Malawians really all dumb?"
" "Does 'being smart' mean 'good at maths' only? "
The answer would probably be that math abilities correlate reasonably well with language and reading skills, and even musical aptitude (the latter bothers me most). However, the correlation comes apart in the ends as it should, so when you look at university professors, the linguists are not incredibly good at math or vice versa.
Although, for some reason, kids with better social skills had lower math skills. This contradicts the otherwise nicely correlating set of all cognitive abilities. https://www.nature.com/articles/srep23011
I am also sceptical about GDP, given that Ireland bumped up its GDP so surely that means we're all now smart and successful? Yeah, well, there's the GDP as recorded and the real GDP, which is the one where you strip out all the multinationals and get what we're actually producing.
An easy way to boost GDP is to become a tax haven for American FAANGs. That has an amazing effect on the perception that everyone in the country is now rich 😁
"The last few years have highlighted the unique structure of the Irish economy. Ireland has a standard domestic retail/services economy which is augmented by a highly productive and fast-growing multinational sector. The two-speed nature of the Irish economy means that volatility can be expected in the usual measures of economic activity.
Following a growth rate of 8.6% in 2022, GDP fell by 5.5% in 2023. In Budget 2025, the Department of Finance forecast GDP to fall by -0.2% in 2024.
However, given that Irish GDP is distorted by the presence of multinationals, Modified Domestic Demand (MDD) is a more accurate measure of underlying domestic growth. Final MDD grew by 2.6% in 2023 and is predicted to grow by 2.5% in 2024. In Q4 of 2024, MDD increased by 4.1% compared to the same period last year. This growth in the domestic economy is largely driven by positive trends in the labour market and in personal consumption."
>"Yeah, well, there's the GDP as recorded and the real GDP, which is the one where you strip out all the multinationals and get what we're actually producing."
The latter figure is GNP (gross national product), what's produced by the Irish; GDP (gross domestic product) is what's produced in Ireland. There was a switch over time in which one economists focused on because it's much easier to measure GDP than GNP.
You're spot on that FAANG causes the two to diverge far more in Ireland than in most (all?) other countries.
Real GDP is what you get by subtracting inflation from nominal GDP (only NGDP is actually measurable, and there are multiple ways to construct an estimate of inflation). Real vs. nominal GNP would have a similar relationship, but harder to calculate due to multiple currencies and their distinct inflation rates.
What if IQ has three components: a genetic component, a nutrition/disease/parasites component, and an education component which is less "real" than the other two? Then you couldn't boost GDP with education if there's still a high disease burden making everything worse.
"Key point is basically: it really isn't true that an IQ 60 person can run a farm functionally."
I would be hesitant to put much stock in this, because I don't know exactly how many smart people you need to run a farm but I'm sure it's not a lot. As an analogy, consider a car: you need reasonably smart people designing the car, but assembling the car can be done with people who are much less smart. Likewise, you need some (maybe just one per farm) smart people to organize the farming, but much of the work can be done without much intelligence.
Scott's responses to many of the comments about point 2 ("but they don't appear disabled...") seem self-contradictory to me. On the one hand, he wants to say that we're badly calibrated to what a low IQ without other disabilities should "appear" like. On this view, Malawians "really" have a low IQ, and the estimate of say, 60, is valid. On the other hand, he also says a few times that the notion of IQ itself falls apart in these cultural contexts. On that view, Malawians don't "really" have a measurable IQ at all, and estimates of 60 are not valid. Both of these can't be right.
I have lived and worked for decades in Africa including Malawi. Malawian farmers are clearly and obviously not retarded or stupid. They are on the whole, lovely, engaging, kind, religious and hard-working people and manage well in a subsistence agriculture world. At the same time they are much less efficient and effective farmers than south-east Asian or mid-west farmers.
And they are not very good at abstract thinking, pattern recognition, strategy, foresight, mathematics. They are not winning Field's prizes. They generally have a great sense of humour and can be very warm and kind. They can also be superstitious, quick to anger, and easily fooled by bad actors.
They are diverse and the population includes some brilliant, nerdy, introverted, eccentric people. But averages mean something and this does not describe the average Malawian.
I support agricultural, medical and educational initiatives in Malawi and elsewhere and encourage others to do so too. At the same time, I don't think Africa or the world will really benefit from millions or billions more Malawians who are a net cost to every high income country they migrate to. And inside Malawi there are very strong patriarchal and social norms which keep society peaceable and friendly (although inequitable and elitist) and when Malawians leave these norms break down which can be bad for them and bad for their hosts.
"At the same time they are much less efficient and effective farmers than south-east Asian or mid-west farmers."
I have commented before about how, in dealing with my 135+IQ wife vs ~115ish me, one of the fundamental tension is her attempt to force various systems of hers on me to produce "efficiency" for me (which are not efficient due to baseline differences in intellectual capacity). Indeed, I think I've read that love of efficiency and desire to promulgate that around may be the hallmark of high-IQ people but not sure where that comes from.
So I could actually believe in an average IQ of low 60s manifesting itself not in people being "obviously stupid or retarded" but in them just being simply "much less efficient and effective" in a similar task.
60 is an outlier, an estimate for one country in one study.
It would be more helpful to think about 70 as an average. There are neighborhoods in the United States where the local average isn't all that much above 70.
http://lagriffedulion.f2s.com/city.htm (from 25 years ago) finds an IQ of Blacks in (majority Black) Baltimore City of 76. Given variation in neighborhoods, probably some are lower than that and some are higher.
Wow! I guess this is caused by migration in/out, right? Basically a mix of "boiling off" (people with alternatives leave, and smarter people have more alternatives) and "condensation" (people with few alternatives end up there and stay)?
That's right, the article discusses it. In the case Baltimore, with its shrinking population, it's probably much more a matter of selective emigration than selective immigration.
What are the implications if IQ does vary and it is mainly genetic and it correlates strongly with ethnicity?
Discussions about this topic always end up arguing about whether this is true and why. Here, we find an argument that IQ differences by country are true. The racists want to say that it's mainly genes, the liberals want to say that it's mainly environment. Even if it's 50/50 that's a big role for genes. What if it's true?
I think this is the most important problem that society needs to come to terms with. Even though the 'racist' view seems to be mostly advocated by conservatives, it's not actually clear to me that the implications lead to solutions conservatives like.
If people have low-IQ mainly due to bad genes, and you can't educate them out of that, doesn't it tend to support some kind of Rawlsian leftist view of society? We need some kind of a welfare state system for them, because those people will never be able to be competitive in society. They need some alternative more lenient legal system, because they are incapable of fully appreciating the consequences of their actions to the same extent high-IQ people are. We could end up with a lot of Woke policies.
Actually, I think this is just the first stage of a broader problem. Society has gotten so complex, technologically and socially, that the low-end of the IQ distribution is incapable of navigating it. We're seeing the consequences most clearly in the groups most affected. It's a lot easier to see that as a racial issue than as an IQ issue, but it's more correlation than causation.
I feel like you've just completely glossed over the other solution, which is eugenics. Though, it might not actually be a good idea to select for high intelligence across the entire population. Optimally you want a caste system where the laborer castes are selected for moderate intelligence and high obedience/low agency, and the learned castes are selected for high intelligence.
This seems to stretch the word 'optimally' beyond all possible reason for most people, for whom 'optimally' and 'caste system' are basically not compatible.
You will also note that claims to simply "be curious" to the contrary, the "race scientists" are social activists. As I noted in another thread this is a reason to demand rigour.
You're not just spouting off about which sports stars are best or which entertainers are the most attractive here.
There's a really important thing to remember when you're talking about how much of IQ differences are genetic vs environmental: it's not constant across changes in environment. In environments where substantial numbers of people are malnourished or have hookworm or something, the heritability of IQ will be lower than in environments where basically nobody is malnourished or parasite-ridden.
dIQ/dEnv is a function of environment, not a constant!
I'm not an expert in any of this. But presumably "intelligence" (or whatever IQ measures) boils down to doing a better job of taking in sensory inputs, processing them, and producing outputs, where better means doing things faster, with more complex inputs, processing, and outputs. This implies that the sense organs and brains of people with high IQ differ from those of people with lower IQ. Has this ever been observed experimentally? If so, how do those sense organs/brains differ? More neurons? More connections between neurons? Faster neurons/synapses? Any/all of the above concentrated in certain functional parts of the brain? Something else entirely?
The obvious one is larger brains. There is a clear association between brain size and intelligence. The current best guess for the correlation between whole brain size and intelligence (g) in adults is about r = .30
"Regarding the size of differences, the Black-White, Hispanic-White, and Other-White whole brain volume differences are, respectively, 0.99d, 0.54d, and 0.36d. "
This might be a clue why our average IQs are presumably low in Africa. We are not politically correct as kids, so we just nicknamed people with big heads 'Big Head'. I suspect most kids made sure their heads remained small.
"The best fit between brain traits and degrees of intelligence among mammals is reached by a combination of the number of cortical neurons, neuron packing density, interneuronal distance and axonal conduction velocity—factors that determine general information processing capacity (IPC), as reflected by general intelligence. The highest IPC is found in humans, followed by the great apes, Old World and New World monkeys. The IPC of cetaceans and elephants is much lower because of a thin cortex, low neuron packing density and low axonal conduction velocity. By contrast, corvid and psittacid birds have very small and densely packed pallial neurons and relatively many neurons, which, despite very small brain volumes, might explain their high intelligence."
Biology is not my area of expertise, I'd just remembered hearing about the research at the time.
That said, I see no a priori reason it *wouldn't* hold intraspecies, although the ranges of neurological parameter values & individual intelligences would be much narrower than interspecies, making the relationship noisier (id est, sensitive to measurement errors and/or species-specific factors).
The evidence-based explanation for why petrostates don't convert GDP-per-capita to IQ at the same rate as their GDP peers is that their kleptocratic governments don't convert GDP per capita to years of schooling at the same rate, not that their people are genetically less capable of converting years of schooling into IQ. Qatar, Kuwait, UAE, Bahrain, Brunei form a very obvious cluster below the curve here: https://ourworldindata.org/grapher/average-years-of-schooling-vs-gdp-per-capita
The selective collapse of intellectual standards on this is really a sight to behold.
And yet educational attainment scores in petrostates show that kids in the exact same classes in the exact same schools get different scores according to their ethnic background. By some strange, and never properly explained coincidence, east Asian kids score highest, followed by caucasians, with arabs far behind.
In the last commentsn thread it was explained why that is incorrect. Most people in SSA are descended from the Niger-Congo population who replaced (killed, out-bred, genocided) hunter-gatherers across most of west, east and south Africa. There is not much phenotypic variance within SSA populations or between them and their (17%) admixed American descendants.
Obviously, there's much left to unpack about what IQ tests measure and how they do it.
Two thoughts -
1) The comment about "IQ of a 3rd grader" statement is in complete ignorance of how IQ is measured. '3rd graders on average in comparison to adults', gets closer, but here too the context is missing - US 3rd graders? compared to US adults? There are proportionally equivalent 3rd graders with IQs on an equal spectrum as there are adults IQs on a spectrum. IQ isn't a measure of absolute, it's a measure of relative. (another aside - don't assess the 3rd graders you personally know as being representative).
2) the question of IQ and development level... Perhaps IQ is more related to the complexity of society as a whole, and development level then is a consequence of the complexity of society. Thus, 5-year-old IQ assessments still remain valid despite 5-year-olds having perhaps less exposure to development, but with substantial exposure to complexity.
"One possible answer is that the causal pathway is high GDP → lots of education → lots of practice with abstract reasoning → high abstract/symbolic IQ. I don’t think this can be the whole story, because some countries that “cheated” to get high GDP (eg oil sheikhdoms) can’t translate it into IQ points at the same rate as everyone else."
It's probably not that they can't, but that they won't. If your high GDP is largely a matter of resource extraction, then you have very little incentive to create a robust educational system to prepare the citizenry for manipulation of complex symbols and performance of abstract reasoning. You just need them to know what a mine cart is. (Unless, like Qatar, you prefer to handle that part via "guest workers" who can never become permanent residents.) The map and table here https://data.worldbank.org/indicator/NY.GDP.TOTL.RT.ZS?view=map shows what countries have most of their GDP via resource extraction.
There's this video I've always found really interesting called "How I kept a job at McDonald's with an IQ of 70" where the guy talks about his life experience with surprising reflectiveness. I'd always assumed it must be fake or his IQ test was bungled or something, but now I'm wondering if it really is legitimate. Even after the examples given of how people with 70-ish IQ can seem normal though, some of the things he says in the video seem a bit too abstract for me to buy it.
Scott: You still seem to be avoiding the issue that Lynn used shitty IQ data for many of his countries, and he "estimated" the data for other countries. I don't understand why you take Lynn's national IQ dataset seriously.
This exposes a more general problem with Bayesian analysis. Unless the researcher performing the Bayesian meta-analysis takes a careful look at the original datasets (and frequently, they're unavailable), it could be garbage in equals garbage out.
Did you read Cremieux very good post dealing with this? He quite clearly shows that Lynn's data was not perfect but good enough. Remember it went through several iterations and while the earlier did depend on extrapolations and interpolations, later versions had more and more good studies to back them up.
And also, all the IQ data is aligned with other data such as the educational scores, the attainment scores, the number of scientific papers per capita, the patents, innovations, prize winners, historically great figures etc etc.
It is all completely consistent. Ashkenazi and East Asians best, followed by Caucasians, a long way behind Indians and Arabs, far behind Africans, at the absolute last Australian aborigines, pygmies, negritos. We are not making this up. This has been shown over and over again by hundreds and thousands of studies. Ignoring this really does show some amazing talents in self-deception.
Cremieux has his axes to grind. He loves to quote studies that agree with his biases and ignores those that don't.
Does anyone know what his academic background is? I suspect he's an example of one of those hobbyist people that Sasha Gusev described in his Abhorrent Science article — "who, with no credible academic affiliations, typically publish in journals that they themselves have founded and review. The underlying studies, to the extent they are conducted, are low quality and highly repetitive because the primary goal is to generate memes for political blogs and forums."
I suspect he is someone who scores very high on openness, very high on conscientiousness, and very high on disagreeability. It is extremely hard to break group consensus and think against the herd. He is obviously intelligent and organised.
Probably a hobbyist. There are more than enough degrees of freedom to come to preordained conclusions on a topic like this. Causal inference is 99% a shitshow and no fancy structural model will save you. Professional x-metricians know this and generally don’t make confident assertions like he does.
The data isn't perfect but we work with what we have. The answers are still informative even if they have large implied error bars. What axe do YOU have to grind that you're upset that anyone is even talking about it? If you have a problem with his methods then contradict him with evidence.
One random thing that stuck out to me, mostly unrelated to the broader point.. the iodine deficiencies in Cuba that they seem to have not covered up for some reason is harrowing.. it makes me wonder just how bad all their other metrics actually are. I’ve been there and it was just awful.
I’d just like to see to see some of this research repeated by somebody who wasn’t a scientific racist OR one of his scientific racist colleagues. We wouldn’t spend so much time talking about an anti-vaxxer’s report on vaccine-related harms.
What do you mean by racist? Do you mean someone who thinks that populations who were separated for thousands and tens of thousands of years and were subject to different evolutionary pressures evolved different characteristics and that evolution did not stop at the neck but also influenced the brain where 25% of human genes are expressed?
Or do you mean someone who has an irrational hate of those who come from a different ethnic background and does not want to treat them fairly or decently no matter how intelligent, kind, friendly, useful and willing to integrate they are?
Or do you mean someone who notices patterns?
If you help me out with your understanding of racist I can try and explain where I am coming from. It is not a place of irrational hate.
He’s probably referring to this stuff (from the article):
He wasn't someone who had "personal racist opinions" that he kept as "personal racist opinions". He was the editor-in-chief of Mankind Quarterly, a white supremacist journal that was founded by people like:
Henry Garrett an American psychologist who testified in favor of segregated schools during Brown versus Board of Education, Corrado Gini who was president of the Italian genetics and eugenics Society in fascist Italy, and Otmar Freiherr von Verschuer who was director of the Kaiser Wilhelm Institute of anthropology human heredity and eugenics in Nazi Germany. He was a member of the Nazi Party and the mentor of Josef Mengele, the physician at the Auschwitz concentration camp infamous for performing human experimentation on the prisoners during World War 2. Mengele provided for Verschuer with human remains from Auschwitz to use in his research into eugenics.
It's funded by the pioneer fund, an organization he was a board member of and that has been classified as a white supremacist hate group, with one of its first projects being to fund the distribution in US churches and schools of "Erbkrank", a Nazi propaganda film about eugenics.
He's not just called racist, he *is* racist, he even describes *himself* as a racist.
Yes, it really depends on whether he meant racist in the terms of "I would like to round up gypsies and jews and work them to the point of death and then gas and incinerate them" or if he meant racist in terms of "I believe in evolution". The term racist could apply to either. My understanding is Richard Lynn was a racist of the second type.
The paragraphs I quoted are about Richard Lynn though. Personally I agree with Scott that Lynn being a total racist doesn’t mean we can’t still use his estimates, but yes he was very racist.. in the first sense of the word you mentioned.
I have listened to him and read some but not all of his work and I disagree. I think he would be totally horrified at the round them up and kill them racism, but would probably have bought into the HG Wells, Galton, progressive racism that was the default for the progressive liberals in the 1930s and a bit before and after.
I think this is a good lesson. If someone is making insanely false statistical statements about the world, it shouldn’t be *that* hard to correct them simply by gathering the correct statistics yourself.
To be really high decoupling here, I don't understand why people say "how can a society with an average IQ of 70 even survive?" when we see troops of chimpanzees and bonobos surviving quite well. Of course they aren't farming but they do use tools and maintain territory and form coalitions and have politics and take care of their young and have friendships and do lots of other quite complex things. Inventing farming is hard, but farming itself in a subsistence world is not a lot harder than maintaining status in a troop of primates.
Though, their ability to farm was coded into them through natural selection over an extremely long period of time. As far as I can tell, human farming is a purely adaptive behavior. They don't have any hardcoded methods to work with.
Humans are very good at imitation. Human subsistence farming is mostly just imitation of previous generations. Not much abstract thought or math involved. The imitation drive is very deep in the homo lineage.
Epistemic status: I'm an interested amateur, not an expert. I've thought about this a bit but could definitely be wrong.
Compare IQ scores with implicit bias scores. In both cases, there are very smart people with a nice coherent underlying theory doing experiments and measures and coming up with test scores that are pretty consistent across time for the same person. The only difference is that IQ scores meaningfully predict outcomes at school, at work, and in life, whereas implicit bias scores don't seem to meaningfully predict anything.
We know IQ scores predict things because of observations on people in first-world countries, overwhelmingly raised under first-world conditions. Just about nobody in the populations used for those studies were ever malnourished, they basically all got their shots and went to school, they overwhelmingly grew up in 20th century first-world conditions and so had stuff like TVs and music and books and such. The studies that have found the correlation between IQ and outcomes we care about are the only reason why we should care more about IQ scores than implicit bias scores.
As far as I know, we don't have studies telling us about relationships between IQ score and school, work, or life outcomes in Mali or Liberia or Kenya or wherever else. My guess (and it's just a guess) is that we would probably not see the same relationship between IQ score and those outcomes for the populations we're talking about. At a guess, that is probably a mix of modern industrial society requiring more abstract thinking, and also encouraging more abstract thinking by presenting you with situations that need it every day of your life.
If that is true, it's consistent with what we see. When you give someone in a first world industrial democracy an IQ test, you get something that's useful for predicting outcomes, and so seeing someone with an IQ of 70 implies some very bad things for that person's future. You can look at that and say that, assuming the score is correct, he's probably going to have a very hard time graduating high school, and he's not going to be able to do things as mentally demanding as, say, reading a five-page set of instructions for how to mix the cleaning fluid properly to get the floor clean without breaking the floor waxing machine. Perhaps an IQ score of 70 in some kid who grew up in a mud hut in Tonga and never saw the inside of a schoolhouse just doesn't mean the same thing.
I have lived and worked in Africa and other developing regions for decades. Many of the commentators seem to have no understanding of contemporary Africa. It is not a country of people in leopard-skin loin cloths with bones through their noses. Most of modern Africa has gone to primary school and is connected to radio, tv, smartphone and/or internet and has been for a quarter of a century.
Only a tiny minority of Africans have never seen a schoolhouse. An African with an IQ of 70 means someone who is probably very pleasant, talkative, interested in you, speaks a couple of languages (wants to sleep with you if you are a girl). But they can't do hard math, they haven't read good literature, they have no interest in abstract discussions, they are annoyed and bored over intellectual pursuits, they have a deep distrust of everyone, think people into shrimp welfare are so crazy as not even to be human.
Many Africans have an IQ of 70. They are often really nice, kind, hard-working, religious people. It really seems to be genetic and no evidence it isn't. We can be quite sure it is not famine or cultural deprivation because of the Dutch hunger, adoption studies etc.
Trust is a funny thing. Some societies which seem a bit alien to me like Swedes and Germans and mid-west have a lot of trust. I suspect they are like cheetahs and went through some sort of horrible genetic bottleneck which means they can all donate organs to eachother.
I know that in low IQ societies there is very little trust and for very good reasons. That was my "lived experience" but I am happy to see if you google studies, people with more poverty and more diversity have less trust.
This is so obvious to me I wonder why you don't grok it? Do you think with people being nasty and brutish and short it would be fun to hang out with them? Take my word, it is not fun.
But you just characterized them as often "really nice, kind, hard-working, religious people". And if they live in smaller communities, why would they have a lot of diversity? Poverty, okay. But now you've switched from "they're really nice" to "they're brutal, impoverished scum."
There are a bunch of measures of trust (how people play in various experimental economics games, for example) that differ across societies. I think you should expect that culture/upbringing/experiences in your society will drive this, independent of IQ. ISTR that Russians come out very low on trust, for example. Russia isn't a low-IQ population, but there's a culture and history and current environment that would plausibly push you toward skepticism and suspicion.
Of late one of the biggest problems is young people who are educated but cannot find jobs because the economy is not big enough to accomodate them. Grade inflation has risen dramatically with access to e-learning and practice tests. So it's now competitive to get into the colleges. It's fair to say 'high IQ' is no longer translating to the good outcomes. Thinking that 'high IQ' would translates to higher national GDP is ridiculous and far-fetched.
All the evidence suggests that IQ predicts good things in the modern world/ west/ global north/ high income countries (e.g. educational attainment, marital stability, health, income, longevity, non-criminality etc) just as well for Africans and Asians living in modern Africa and Asia as IQ clearly predicts good outcomes for people born in high income countries. IQ is the single best single predictor of all of these nice outcomes. Yes, better than parental income or SES.
There is some evidence that people with "smarts" are also better hunters, gatherers, "survivors on a desert island" than people who are obviously more dumb. But it also seems there are some skills useful in a hunter-gatherer lifestyle which are not highly correlated with IQ.
Having spent many decades in developing countries I have a lot of first hand experience working with populations around an average IQ of 70. Yes there are many great, hard-working, religious, serious people I am happy to count among my friends. No they would never have got to an industrial revolution.
Something funny and special happened in NW Europe and especially England. Similar things were in the air around parts of Asia. Greg Clark has interesting things on how England ratcheted up IQ through primogeniture. HBDchick had interesting things on how within the Hajnal line marriage patterns led to high IQ.
Good fun reading this and so glad it has escaped the Overton window.
I was in a small village in west Africa, doing blood tests for an endemic disease, and I left my bag with all the money for the whole study like 500 USD on a post in a field. And the African villagers just took the money back to me like you idiot you forgot it. They were making less than 50 USD a year and they just hurried back to me saying you forgot your wallet.
The topic of racial or national differences in IQ has always tied me in knots. I’m pretty sure my knots are not unique so I thought it might be helpful to just name them here. There are 3 knots:
The first is that I grew up in the south, and absorbed what you might call the soft version of southern racism. My family thought it had been a mistake to integrate southern schools, because “nigras are just different from white people, honey.” They would certainly not have been in one of the groups of enraged whites protesting one little black girl entering the white school. They weren’t rageful people, and would have had empathy for the little girl. They just thought integration was a mistake because nigras were different. And what they meant by different was dumber, more prone to violence, and more sexual. On the other hand, my father had been parented mostly by a black nanny named Jenny. His own mother was a brittle socialite. He loved Jenny, and in fact when she was dying, when my father was in his 40’s, he traveled to DC to be with her and was actually holding her in his arms when she died. So blacks were seen as more primitive and animal, but for also as people who had a lot to give in contexts where earthiness and practicality were important. Anyhow, I did not question what I was taught about black people, and actually did not have much interest in the subject. I am now much better informed about race, and my feelings about the matter are much more nuanced. But a shrunken version of the old, of-course-they’re-inferior point of view is still drifting around in me, and occasionally I’m in touch with it.
The second is that I have always thought of myself as a smart person, and when I was in my teens things like IQ and SAT scores loomed large in my mind. That side of me has shrunk greatly over the years but it is not gone and I am ashamed of sometimes feeling better than other people because I can tell that in that IQ test, SAT kind of way I am superior to them. Also, there was an occasion in my life when the rubber really met the road as regards race and intelligence: I decided to adopt a foreign-born child, and had to decide what countries to consider adopting from. I ruled out Africa, and ended up adopting from China. I told people I was ruling out Africa because I did not want a child who would have to deal with the black/white issues in the US, and while that was true, it was also true that I was afraid an African child would have a lower than average IQ. I feel guilty about that.
The third is that I get occasional glimpses, sometimes on here, of people full of truly venomous hatred of blacks, people who sound like they would like to exterminate them. I feel guilty about participating in discussions that indirectly support the ideas of the venomous, and that might nudge neutral people closer to the venomous attitude.
None of these knots make me wish Scott had not posted his piece. But they interfere with my thinking about the subject. There are so many aspects of it that make me feel guilty. And I don’t feel able to distinguish clearly between being racist and being realistic.
...Don't feel too guilty about any of this. This is just a tiny forum of outcasts, divorced from social realities. History will move on, with or without us.
I really like this comment, it's very honest on a subject where it's hard to be.
I layer on top of your personal guilt, which I share, a dose of suspicion -- why are some rationalists, the people who are most impressed by iq, are most proud of their IQs and are most convinced that bands of high iq engineers are going to bring about a possibly utopian singularity, the same people who are so keen to make group iq differences an area of focus?
I think i'd rather hear about the subject from people for whom iq is not already a major part of their identity.
also seem blind to certain forms of excellence not
captured well by IQ tests. I’ve seen a few tech guys post on here that we don’t need poets any more because AI can produce a poem on any subject with correct rhyme and meter. Yeah, but it’s doggeral! Somebody put up a post saying they didn’t believe anyone really liked Shakespeare— what’s really going on is that people pretend to in order to win culture points. Good grief!
people "with" low IQs, but who seem canny in most regards, might just have not tried very hard on the test and now are subject to... I want to say "contract-lawyering about the results", but there's really no arguing even happening. It's more like, some people see an IQ test result and expect not to be challenged if they treat the person like that's what their IQ is, and the person with the low result either doesn't think it's their place to contest the result or is letting society martyr them because they have nothing better to do
I'm one of those people who thought these "national IQs" of 70 had to be wrong, and what I'm seeing in this discussion isn't changing my mind. Defenders are saying, IQ 70 doesn't mean that whole nations are retarded, in fact the average IQ of the USA itself was 70 in 1900. Is this really consistent with the historic usage of IQ, or are we seeing equivocation between different ways of defining IQ? I already know that at some point there was a shift from "ratio IQ" to "deviation IQ" and that this impacts the meaning of higher IQs; is there a similar rescaling of lower IQs too, that leads to a "softening" of what it means to be IQ 70 or IQ 50?
The gains in IQ from 1900 were mostly not on g or innate intelligence but on IQ subtests that are more related to 'being modern' e.g. getting better at taking tests, Ravens, weird stuff people in the Victorian age weren't used to. Subtests on vocabulary or general knowledge or counting backwards haven't changed much. If you met your great-great-grandfather they would probably have talked to you very intelligently about literature or politics or history but been very bad at reading comics or doing sudoku.
And people with an IQ of 70 are really perfectly fine and nice and interesting and good people. Just they don't do abstract thinking and have no interest in shrimp welfare or the Monty Hall problem. Can't do hard math, don't read poetry for fun, tv people.
It's not clear to me you're properly calibrated here. What you just described are "average" people, whose IQ is supposedly 100 in the U.S. An 85 IQ is the bottom 20% of Americans. In my experience the bottom 50% have no orientation whatsoever towards intellectual matters and can't do even easy math problems. A 70 should be something else entirely.
The people you are describing are not 70s and might very well be 90s.
"it seems like there's some disconnect between do-well-on-tests intelligence and run-a-subsistence-farm intelligence, and the abstract/concrete and novel/cultural distinctions are the best explanation that I can think of"
Exactly- people would be 80% less annoying about IQ if we called it "abstract intelligence"
Having worked with subsistence farmers in Africa and Asia for decades I disagree. The farmers who are better at tests and maths and vocabulary are better at subsistence farming. Plenty of studies show that good predictors of more profitable and productive subsistence farming are younger, better educated, smarter, and male.
"Africa is extremely genetically diverse, but I think most of the countries measured in the paper, including Malawi, are some variety of Niger-Congo speakers, who I don’t think are that much more diverse than white people or anyone else. The really interesting African ethnicities, like the Khoi-San, don’t show up as much at a national level."
there are two issues
- the average genome of an african is more 'diverse' than a non-african. so an african has 15-20% more SNPs (~5 million) in the genome (3 billion base pairs) than a non-african (a bit over 4 million). this is for an individual
- but, the pairwise between *populations* the genetic distance between populations is not always that great; that between various bantu-speaking ethnicities for example is similar to that between european nationalities. so not that high
naively, you could then make the cause that african IQ variance should be higher in the aggregate (since 15-20% more genetic variance), but the between-group differences should be like those in europe (i think the range is 95-105 except for a few outliers like albanians).
The idea that African subsistence agriculture is highly optimized reminds me of an anecdote in this recent Economist article. They discuss a very effective intervention for improving dairy yields: have the farmers stop starving their cows.
And then your population would grow to consume the surplus. Malthus is mostly dead these days, but he can crop up in the subsistence-farming parts of the world.
Flynn himself has suspected that the eponymous effect might be due to more time spent in learning abstract thinking. He also pointed out that earlier IQ tests had more practical life questions and that IQ tests became more abstract and symbolic over time. Finally... It appears that modern populations do worse on these practical questions than the same countries did 100 years ago.
"First, he corrected my misconception about Reich on ancient European cognitive evolution. Reich had said that pre-agriculture Europeans were “2-3 standard deviations” below moderns. I had interpreted that as IQ deviations of 15 points, making them genetic IQ 55-70, which would have been pretty crazy. Stone tells me he actually meant PGS deviations, each of which was about 3-4 IQ points, so he’s claiming that pre-agriculture Europeans had genetic IQ of 90 (they probably also had lower IQ for environmental reasons)."
This is not quite right either. The real answer is that we don't know how to scale the PGS deviations from ancients to moderns. It is unwise to assume the modern correlation and use that. Think about it this way. If the real gap moved 45 IQ points (3 SD), and we use a PGS with validity of 0.33, this would mean the PGS would move 1 SD. There is no real way to interpret these PGS gaps in terms of real IQ points, or heights, for that matter.
Lyman Stone is still wrong about National IQs. His chief mistake, ironically, is trusting each datapoint as accurate. Single studies for countries are not that reliable as indicators of the mean IQ of that country. If one tracks them over time, one can easily get the impression nothing is reliable. This is the wrong conclusion to draw. I think in modern times Ron Unz was the first to do this kind of data digging, so we may aptly name it the Unz fallacy. In any case, Seb Jensen already replied to Lyman Stone. https://substack.com/home/post/p-155059660
As a numbers person, this is obvious about everything. If you want to know the average milk yield of cattle globally you will find a bunch of studies. In high income countries you likely find many, representative, large studies. In low income countries you will find fewer studies often with problems in sample size, selection methods etc.
And you will find that cows in America and Australia give most milk and cows in Africa give very little. The global milk yield of cattle looks strangely similar to global IQ or patents or homicide or anything else nice or nasty.
The few bad studies on milk yield from Malawi and Chad will likely show a lot of variation study to study and year to year. But the overall trend is obvious. It is just the same with IQ.
Harmonized Learning Outcomes use tests of learning outcomes, not psychometric tests. That preprint “updates” Lynn’s numbers by averaging them with the learning outcomes data. Do you see the problem with using a measure of development to estimate national IQ, then turning around and using that as evidence that national IQ causes development?
That’s maybe most true of PISA reading questions (it also has math and science which are pretty unambiguously functional competency) but generally a mischaracterization. TIMSS is math and science, most of the others are math and literacy/reading comprehension. They very obviously test and are designed to test education quality.
There are many ways to talk about poverty in Africa, or in Malawi in particular. Malawi is one of the poorest countries in Africa. So you can choose how to describe the level of poverty. You can use hard economic measures, health, or education as defined by an alphabet soup of international organization. Education also has many different metrics, from basic literacy, years in school, tertiary education to the quality of education when attended. Alternatively, you can come up with statistics about IQ. (Btw I'm neither confirming nor contesting the validity of IQ testing as a measure of intelligence, the validity of those tests in Malawi or the accuracy of data in your map in the event it's actually valid). I'm just sharing a perspective as an African.
>(Sidenote - there is more genetic variation in what you call 'sub-Saharan' Africa than between Asians and Europeans)
The ethnic groups in Malawi overlap into Zambia. Malawians have also been significant migrants to Zimbabwe and South Africa for over a century and chiChewa is now an Official language in Zimbabwe. Ethnic groups in Zimbabwe also overlap into Zambia, Botswana, South Africa and Mozambique. The Batswana people live in Botswana, Zimbabwe and South Africa. I grew up in Zimbabwe learning and playing with kids of Malawian heritage before spending most of my adult life in South Africa. I suspect I'm qualified to claim that the Bantu Peoples are one big family. The genetic evidence is overwhelming and our different languages are obviously related. Maybe just anecdotal, bit there's nothing inherently inferior about Malawians or their intelligence. If the data is actually valid, I suspect they just measured a proxy of poverty - as if we needed other people to say it AGAIN.
African countries have structural macroeconomic challenges which have been studied extensively for decades. The extreme poverty in Malawi is certainly not because of low IQ. Abit of history before I make an ASK:
Firstly, Zimbabwe (as Rhodesia) and South Africa were the last bastions of European settler colonialism. For over a century, Europeans settlers knew Africans were capable. That is why they reserved lucrative skilled jobs for the European settlers. They knew that if they competed equally on merit, Africans will get the jobs despite 'native' education being underfunded by a factor of 10!
Secondly, Africans have been subjected to degrading studies to prove their inferiority for centuries. There're still thousands of human remains in European archives used for these studies. So far, they have all been proven to be nonsense and some of us thought Europeans and others would finally let it go. Africans were also turned into spectacles for the amusement of Europeans and even exhibited in zoos. Each time they protested the maltreatment.
The poverty in Malawi is a very serious matter and not a spectacle for your amusement. So. Will it be asking too much if I ask you to take down your articles? If that is too much, can I ask you do your studies about IQ and whatever other variables using data from other continents in future. Malawians are the most hard working and kindest people I know. Perhaps they behaved that way because they were migrants in Zimbabwe. I wonder where the'd rank in kindness stats???? So let me make a final request - If Africa is irresistible for you, can you use Zimbabwe instead? You have my permission. It is also a poster child of notions of African inferiority in online circles and will likely generate a lot of engagements too.
The natural state of humanity is poverty. Malawi is not poor because of colonial studies but because all humans were poor until an unusual sub-population in Europe put together innovations and institutions that moved them out of the brutal and grinding poverty that characterised the lives of my European ancestors.
I don't want to be snide, and I know it is a very common mis-perception that Africa is only poor because someone (who?) stole something (what?) from them. But economists have been studying this seriously for decades and this is absolutely incorrect. Noah Smith is a good populariser.
People can absolutely be hard-working, kind, very religious, have a low IQ and be poor. It seems that what keeps populations poor is not their propensity for hard-working or kindness but a combination of low IQ and strongly associated factors such as low executive function, high impulsivity, high discount values, little ability to do hard math, can't abstract, doesn't read or write poetry for fun, doesn't care about shrimp welfare, believes strongly in magic and god, good sense of humour, explosive strength in athletics, friendliness etc etc many which have a substantial genetic contribution.
That is the dismal truth. But the world as a whole will have worse outcomes if we refuse to recognise it.
I did NOT suggest that Malawi is poor because someone stole from it. Apologies if I gave you the impression that I’m looking for the reasons or bold insight that shrimp welfare and poetry appreciation are the missing links to poverty alleviation. I do not mean to snide.
My experience and all of the literature suggests Malawians are not just otherwise identical to north-western Europeans except that they happen to be poor. There are many differences between them and some of the differences are highly correlated. Loving Shakespeare does not make you rich, nor does ability to do advanced calculus. But understanding complex language and maths is closely related to high IQ. Countries with low average IQ are going to find it difficult to escape the poverty which is the natural state of humanity.
I myself have Autism, and as a kid I took an IQ test and they scored me at 126, IIRC, and told me I was in the 96th percentile or so - smarter than 95 other people in the average 100 person room. But I really don't feel that smart in day to day activity. So I think IQ just isn't very well correlated with day to day stuff at all.
All of this IQ discussion is making me curious. Is there any way for me to take the test without going into a psychiatrist office? Or, how do I convince my psychiatrist I need to take this test?
"Even a "normal" low-IQ person cannot do the implied math and long-term planning involved in this kind of farming. And in fact, economists routinely find that African small-plot subsistence agriculture is actually highly optimized; farmers make very precise choices about where to plant which seeds, which fertilizer to use, etc. Key point is basically: it really isn't true that an IQ 60 person can run a farm functionally."
Makes sense to me. IQ tests are timed, so they partly measure speed. I'm pretty sure, I could figure out all questions correctly on a normal IQ test, if given enough time with pen and paper. That would not mean, I have the highest possible IQ at all!
Similarly, farming runs on seasonal cycles and you also get every incentive to get it right. If you spend an entire day on removing weeds, being 60 IQ probably means it takes you hours to figure out the optimal wheedwhacking pattern, that someone at 140 IQ spots within seconds and someone with 100 IQ spots within minutes. But with 60 IQ, you probably still spot it after a day, if you've got nothing else to think about. And then remember and tell your low IQ compatriots. They can perhaps not learn to fluidly navigate truckloads of abstract information from text or video, but you don't need to be comfortable with that amount of information that the farming lifestyle provides. Especially, if it's communal farming where knowledge/skills/responsibilities can be distributed, so no single person needs to understand the entire emergent workflow, which may in total end up very sophisticated and complex.
Except African small-plot subsistence farming is not highly optimised. It has extremely low productivity which is directly related to inadequate use of inputs and poor management. This is not just costly inputs like fertiliser and hybrid seeds but studies have found African subsistence farmers would greatly increase yields just by weeding (labour no input) in societies where people are chronically under-employed and not time scarce (except at short intervals around harvest etc.). Likewise if African farmers would give their cows enough water to drink they would greatly improve milk yields. Recent decades have seen African agricultural yields increasing due to cultivating more land (bad for ecosystems) and not through increasing productivity which is woefully low.
But you can use basic common sense. If everyone in Malawi knows that tall parents tend to have tall kids and smart parents tend to have smart kids and we know from Finland that IQ is nearly as heritable as height it takes a lot of perverse reasoning to remain resolutely agnostic as to whether height, or lactose intolerance, or IQ are largely heritable in Finland but largely environmental in Malawi. One of these is more likely than the other.
National IQ estimates tell us something about what is causing under-development, how optimistic we should be about current development interventions, how much we should start looking for other possible development interventions, how immigrants from a country are likely to perform when exported to another country, how many immigrants we should accept from other countries, how much we should address inequality of outcomes between people of one nationality by efforts to correct prejudice or compensating them because of a presumption they have been unfairly treated.
"Scientific racism" is obviously pejorative. There is nothing racist about noting that every two real normal distributions of non-integer values will have different means. If you spend all of your time obsessing about IQ and black vs. white, on the other hand, you just might be racist.
Now I feel stupid for having spent all that time reading the comments.
Why?
Presumably because he could've saved time by just reading the best comments in this highlight post.
I, on the other hand, am going to add 3 points to my personal IQ estimate for wisely having chosen to skip them.
If we know there are 30 point differences between pre and post industrialization within the same societies shouldn't it be pretty self evident we see 30 point gaps between countries?
Are you referring to the pre- and post-agriculture confusion Scott talks about here?
"Reich had said that pre-agriculture Europeans were '2-3 standard deviations' below moderns. I had interpreted that as IQ deviations of 15 points, making them genetic IQ 55-70, which would have been pretty crazy. [Lyman] Stone tells me he actually meant PGS deviations, each of which was about 3-4 IQ points, so he’s claiming that pre-agriculture Europeans had genetic IQ of 90"
I'm skeptical of that summary of the finding. Any selection pressures that impacted the genes used to compose a PGS score probably impacted the frequency of a bunch of other genes for intelligence we haven't identified yet.
(This is why PGS tracks with observed group-level phenotypic IQ across racial/ethnic/religious lines so well, even though it only directly accounts for ~10% of the variance.)
Fair enough, but the "30-point differences" is still based on a misunderstanding of the point that was made (even if that point was wrong). Or does it come from somewhere else?
No I'm referring to the 30 point flynn effect over the past 100 years for European and east Asian countires.
https://www.sciencedirect.com/science/article/abs/pii/S0191886911001437
OK. I wasn't aware that worked out to a flat 30 points.
In fact, I doubt that it does represent a flat 30-point industrialization bonus for all societies, because societies are presumably starting from different baselines on all the factors that may contribute to the Flynn effect (poverty, mal-/nutrition, education, environmental toxins, etc.) So it really gets to the heart of the topic: Yes, pretty much everyone agrees that there's a very significant environmental contribution, but we don't know how big it is, or how it is distributed, among or within societies. It's a bit facile to say it is self-evidently 30 points.
(And as for MA_browsing's skepticism, it is interesting, but still not particularly relevant to my reply when I misunderstood your comment, and even less relevant to your original comment.)
I was generalizing a bit too much, sorry for that. The initial Flynn effect was the re-meaning of IQ scores because they increased by roughly 3 points per decade in the United States. The strength of gains vary according to country, intelligence domains, and the investigated time span. Some areas like sub-Saharan Africa have shown low or no gains while, countries like South Korea and Japan have had periods of 7.7 point per decade gains with 35+ point gains overall.
http://www.iapsych.com/iqmr/fe/LinkedDocuments/nijenhuis2011ip_FE_Korea.pdf
If my reading of the Flynn effect is correct, this suggests pre-agricultural Europeans had an IQ about in line with Americans in the 1980s? Seems to undermine basically any suggestion in my mind that this means anything whatsoever.
This speaks specifically to genetic IQ, whereas my understanding of the Flynn effect is that it is best understood as a raising of the mean largely due to environmental changes that raise the lower part of the distribution in the US (through better education, nutrition, less poverty, etc.) up into the 80s. Also the 3-point pr decade probably can’t easily be generalized to all other periods and places. It would be astounding if genetic IQ increased by something like 3 points (or even 3%) every decade – especially ever since the agricultural revolution.
Does genetic IQ make sense as a concept? Is this not just a renaming of g? I would guess the explanation would be what they'd score if they were brought up in the existing American system, but why is the educational system of the US privileged as the "normal" education system? It seems impossible to determine what the objective measure would be because everything is a mixture between genetic and educational components. Maybe the US has terrible education and very strong genetic advantage? Maybe the opposite.
In principle even pure g-factor can be lowered by environmental factors, like malnutrition or cyanide poisoning or viral infections of the nervous system and so on.
Flynn Effect gains vary a great deal depending on the test-type you look at, and have probably co-occurred with genotypic decline.
https://www.sciencedirect.com/science/article/pii/S0160289622000241
The Flynn effect doesn't represent a true intelligence gain because it isn't measure invariant. AFAIK there's no reliable way to norm test results over time so it's not really meaningful to say that modern Americans are 30 points smarter than 19th century Americans. Informally it seems pretty obvious that that can't be true because if the average IQ a century ago was 70 then we never could have modernized.
That only holds if IQ is actually normally distributed; if in the past there were more prevalent environmental limitations to most people's IQ but the distribution were more heavy-tailed, we'd still have enough smart people to modernize society.
(Yes, I know IQ is normally distributed by definition; just pretend I said some sort of "effective IQ").
This doesn't mean that your overall statement is incorrect, though.
I am inclined to believe the IQ test results are primarily a product of human adaptability at early ages. That is, if you grow up in a literate culture with lots of emphasis on the kinds of thought present in solving pen-and-paper puzzles, you will be much more capable of understanding them, and much less capable of understanding the things that you would understand were you brought up as an illiterate hunter-gatherer. Not explaining myself very well, but a friend told me about a study suggesting that literacy restructures our brain by overwriting a portion of it that would otherwise be used for recognizing faces. Assuming something like this is right, I would imagine it extends to the sorts of concepts involved in any puzzles, and so I probably wouldn't expect an illiterate hunter-gatherer of average intelligence to be capable of understanding anything he was shown on pen and paper, regardless of whether it involved words or not, to the same level as a literate person with a high school diploma.
What explains the variance among literate people with a high school diploma?
While I agree the original statement is too strong, IQ test could measure something closer to intelligence in literate cultures and fail to measure that in illiterate cultures, through the process he said above. Especially considering IQ test already require normalization to work correctly, even within literate societies.
It doesn't seem to work that way, though: even on e.g. purely visual or mechanical tasks, the rank over doesn't reverse.
By the same token even intelligent people from literate societies often appear to hunter gatherers to be childlike in their inability to acquire basic observational and tracking skills so the „overwriting“ theory makes sense to me.
Do you have links that go into more detail on this?
I bet no—or, at least, I would be surprised if there were any substance to this claim; sounds like Did You Know? & Just World pontificating, t'me.
Most stories I've read involving contact between Bushmen and Westerners involves a very different dynamic.
There are.anecdotes.along those ones in Diaonds Guns, Germs.And Steel, and Evertetts book on the Pirahã.
Some of the best hunters in the world are westerners who have just spent more time hunting than the rest of us. It's probably not anything to do about "overwriting." You can learn hunting skills even as a fully developed and literate adult.
I would strongly expect that all of those individuals were also taken hunting at a young age, or at the very least participated in types of play that used adjacent skills.
"Overwriting" might not be quite the right word, since it implies replacing skills that already existed, but it's certainly true that early development devotes more or less "space" for skills based on their usage, sometimes in in areas (evolutionarily) "intended" for other skills. (Of course, neuroplasticity doesn't entirely go away with age either, it just gets less effective)
Why wouldn't you just expect that hunting is a skill like any other?
I expect that hunting is like anything else -- if you devote a lot of effort to learning about it and practicing it, even starting as an adult, then you can become very good at it -- if not among the best in the world, then at least as good as the average tribesman.
Right, but *like any other skill*, the people who are the best in the world usually started practicing when they were young. Like I said, neuroplasticity is just less for adults, and at the top of the bell curve you usually have people with every advantage. And starting will usually give you a greater effect for less effort. (Plus in this specific case, 'dads taking their son hunting' is practically a stock story, so it's not like it's uncommon.)
OK, but now we're less in "overwriting neurons" territory and into "normal skill acquisition" territory.
Best hunters in the world with what tool package and education? Would these hunters really be comparable to the skillset developed by societies with a hunter-gatherer cultural package?
You may be referring to The Weirdest People in the World By Joseph Henrich.
Yes, Henrich mentions that early on in the book. He claimed that literacy re-purposed the ability to process faces.
Unfortunately the claim is based on studies by Dehaene with very small Ns (~10) and interpretations of fMRI images (which are controversial).
There is a paper by Huettig, F., & Mishra, 2014, that concludes:
"Dehaene and colleagues (2010) have suggested that face perception abilities will suffer as individuals become literate because orthographic representations ‘invade’ the ventral visual system and interfere with the enlargement of faces and houses into the surrounding cortex.
There is no solid experimental evidence from behavioral studies to support such claims. Huettig et al. (in preparation) compared complete illiterates and literates from the same village and the same community of Indian society on the recognition of the emotions of the standardized Ekman (Ekman & Friesen, 1975) faces. There was no hint of a difference between the two groups. The Ekman faces are sometimes criticized for being culture-specific (i.e., Western), and thus, perhaps these results have to be interpreted with some caution. Orihuela, Carreiras, and Dunabeita (2013) have also investigated whether illiterates and literates differ in face perception.
These authors found in fact the opposite effect, illiterates made more errors than literates in a face recognition task of varying degrees of difficulty. Thus, as in the case of memory research, literacy may improve face perception rather than being detrimental."
Thanks for that. I bought the book cos I was interested in the repurposing but then was disappointed to find that he just skipped over it.
I offer my own n=1 observation: I am pretty bad with faces (not terrible, but noticeably worse than most other people) and I learned to read very young.
An ordinary 21st century Westerner with a TV can easily recognise thousands of faces, more faces than an average hunter-gatherer will ever encounter in their whole lives. If literacy has invaded the facial-recognition part of our brain then what are we supposed to have lost?
The hunter gatherer will have learnt thousands of plants and animals, more than an ordinary 21st century Westerner will ever encounter in their whole lives.
Evolution gave us that capability and it wasn't so we could recognize Z-list celebs!
> The hunter gatherer will have learnt thousands of plants and animals
This sounds unlikely. Can you back it up?
I was echoing Melvin's comment - the numbers aren't important. But hunter-gathering is memory/information intensive: you need to remember which trees produce edible fruit or nuts, where they grow, what time of year they produce, how to collect and how to prepare them to remove toxins - multiply this by all the other things you need to eat over the different seasons.
What explains the variance between groups on IQ test proxies that don't involve "pen-&-paper reasoning"?
According to Aporia, the claim that Lynn described himself as a “scientific racist” is taken out of context. He said, “If we are talking about people who believe there are genetic differences between the races, then I am definitely a scientific racist.” I can’t speak for the other claims about his beliefs or who he was affiliated with, however.
https://www.aporiamagazine.com/p/the-dissident-psychologist?utm_source=publication-search
I don’t think there’s any other definition of scientific racist.
Disagree. I don’t think the science about race differences can justify racism, which that term implies.
I’m assuming he meant genetic differences *in intelligence* between the races, but I guess it’s unfair to attribute that broader context to this one quote
Even if it was it's still not racist. You're making a category error in conflating empirical claims with normative ones. If races differ in IQ for genetic reasons then they do. That's a scientific question that has an objective answer. It doesn't make anyone racist to have a good-faith opinion about objective reality.
In fact it's very dangerous to tie anti-racist views to arguments like this. What would you do if it became the indisputable scientific consensus that IQ is 90% genetic and definitely differs between races? Would you then be compelled to describe yourself (and everyone else) as a scientific racist?
I’m just describing how the concept is used; not commenting on its normative valence. For 95% of people, if you tell them you think Whites are genetically smarter than Blacks, they will call you racist.
Fair enough, but this idea needs to change.
Another critique I put down below:
“Okay, but you shouldn’t use that definition. Like I would refuse to use the word “Quran” to mean “the true word of Allah” even if societal developments lead it to be commonly defined that way because it’s unhelpful for many reasons. It’s better to use a different word without the assumption of “true” within the definition.”
Has no one shared the link yet? Ok then, I will. Required reading by Scott on the topic. https://slatestarcodex.com/2017/06/21/against-murderism/
I don't know if it's 95%, but you're pretty much right. But probably because they are assuming (incorrectly) that there is no good evidence, and (possibly with some justification depending on their social context) that people who say it don't like other races, otherwise they would respect the taboo and not investigate the topic, and/or just keep their mouths shut.
A non-trivial percent will also say stupid incoherent things like "racism = prejudice + power" and only whites (all whites, everywhere, even poor homeless ones, because they "participate in and benefit from whiteness" or something), so if a Korean, for example, says, "god I hate f***ing n***ers, and I intend to discriminate against them", that's not racist. People have all sorts of stupid and incoherent beliefs.
As Scott argues, people switch between different definitions without realizing it, but the only one that really makes sense, and also the one that people ultimately fall back on in practice is the definition from motives (being motivated by ill-will toward a particular race(s)).
+1 - you can acknowledge differences and still treat people without prejudice and not make categorical judgements about them. The term racism includes discrimination based on race.
How are you defining 'racism', then?
I don’t have a perfect definition of racism. I just know that it can’t be that because scientific differences between races doesn’t justify cruelty, disrespect, lack of rights, etc. towards other races.
The word's meaning has swollen quite a bit beyond that limited definition, though. Almost all the people who fought to abolish slavery in the 19th century would be considered 'scientific racists' by modern standards.
Okay, but you shouldn’t use that definition. Like I would refuse to use the word “Quran” to mean “the true word of Allah” even if societal developments lead it to be commonly defined that way because it’s unhelpful for many reasons. It’s better to use a different word without the assumption of “true” within the definition.
The way I learned it at least is that racism was trying to allocate resources based on racial differences. Just positing a difference between/among races isn't enough.
So if I say "Koreans and Nigerians are different", that's not racist according to the definition I learned in college. Even if I say "Koreans and Nigerians are physiologically and biologically distinguishable", that's not racist. But if I say "Koreans and Nigerians are different, and therefore I won't let one group into my store", that's racism.
Once you tie some kind of resource allocation to being a member of a particular race, that's where you run into trouble.
From what I can see on the Wikipedia article, it mirrors what I learned in college. Wikipedia defines scientific racism as "is the pseudoscientific belief that the human species is divided into biologically distinct taxa called "races", AND that empirical evidence exists to support or justify racial discrimination, racial inferiority, or racial superiority" (emphasis mine). Without that second part, I don't think you have racism.
I think that many people (especially today) drop the second clause when attributing racism, and think that any highlighting of (especially biological) differences counts as racism.
Why? In general I think that there *had* been a decoupling of these beliefs, at least in terms of people willing to discus so publicly. If you imagine the quadrant, nearly all public discussion got sorted into the groups of (Differences + Discrimination) or (No Differences + No discrimination). This state of affairs lasted for roughly the long second half of the 20th Century. Highlighting major genetic or biological differences between groups of people that may gesture towards "race" being real was increasingly the domain of people who happened to also be racist or at least vaguely gestured in that direction.
It wasn't purely coincidental that this original sorting from 4 groups to 2 happened. The descriptive claim isn't purely independent of the normative claim. Consider racial justice or feminist campaigns of recent years. The ~roughly mainstream scientific (and Scott's?) view "racial differences are real but we can treat people equally" view is now more popular and acceptable, mostly due to the sheer amount of genetic evidence coming out every day. But it also saw the decoupling on the other side, people who thought (innate) racial differences weren't real BUT you should treat people differently.
You might correctly identify differences between groups, deny the biological explanation, and then attribute said differences to solely environmental factors. If you then think "and that means we should act differently to if all these groups had similar outcomes" (e.g. DEI) then it means your descriptive claim has influenced your normative claim. But if you instead think that "wait, no, a lot of these differences are actually due to *biology* or *genetics* (and those cluster in groups etc. etc.)" then you can see how the maintenance of "differences in people should entail differential treatment" (the same macro view as before!) becomes racist.
If the four views then are "90s liberal race blind", "technocentrist enlightened genetics liberal", "DEI advocate" and "dictionary definition racist", which of those today could get you labelled as a racist?
For the last 25 or 30 years, I've been publicly articulating the moderate, reasonable, empirically-backed position that the racial differences due to nature and nurture that we see when watching sports on TV are also seen in other aspects of society. (I realize that most of Scott's readers don't watch sportsball, but a whole lot of people really care about it.)
I've pointed out that the obviousness of racial differences in sports performance (e.g., a Kenyan is about a zillion times more likely to win the 3000m Steeplechase gold medal at the 2028 Los Angeles Olympics than a Bangladeshi) has not led to the genocide of the (literally) weaker or stronger races.
Instead, people really like sports, which leads them to having heroes of different races. White football fans really admire Lamar Jackson (a black QB) and black football fans really admire Josh Allen (a white QB). They might be modestly prejudiced in favor of the football hero of their own race, but ... that's okay.
When people ask me when I became aware of racial differences, I point to an October 13, 1968 college football game I attended at the Los Angeles Coliseum featuring two All-American running backs. #13 Oregon State's star was the huge Steve Enyart, who used his upper body strength to devastate #1 USC's defensive line in the first half. But in the second half, USC's superstar tailback used his lower body speed to overwhelm Oregon State. His name was O.J. Simpson.
But these obvious racial differences in sports didn't cause American elites to conclude that, well, sure, nature and/or nurture lead to racial differences. Instead, they concluded that anybody daring to mention the obvious was an evil enemy of society.
I've oversimplified, obviously there are "acceptable" ways of treating people differently due to their differences that don't get you labelled as a racist, high performance sports being one of them. Nozick chose Wilt Chamberlain for his famous thought experiment after all. If you interrogate this too deeply or take it toward logical conclusions (reward elite talent, but don't discriminate at the bottom end..?) it falls apart of course. So you could self deceive that you sit in one quadrant and not in the other.
Anyway this is more for public consumption, I don't need to explain this to you Steve. I do note that your own views (and self!) have become more mainstream as of late and this is largely due to that 2 quadrant equilibrium breaking down and the millions of sequenced human genomes we now have.
If you truly think that widespread acceptance of the idea that black people are less intelligent than whites would not lead to a massively more negative perception against black people from other races, you're deceiving yourself. Most people assume that white people are not that good at basketball, which doesn't do white basketball players any favors, but doesn't move the needle that much because basketball is a tiny part of life. If people assumed that black people had the skill level for jobs requiring intelligence as white people have at basketball, it would lead to a significant increase in bias against black people.
Professional athletes are outliers. Already I am seeing (unprompted by anything I choose to engage with) pretty racist memes on Instagram Reels and (especially) twitter based on the idea that black people are less intelligent. Maybe that theory is correct, but if it becomes widely accepted things will be much, much worse for a lot of people, and I don't believe you are being honest when you suggest otherwise.
That suggests a useful academic study: how much are white athletes hurt by stereotype threat and the like? Over a 20 year stretch that recently ended, no white players were regular starters at cornerback. How come? Well, blacks tend to be better cornerbacks? But 100% black with no whites or Samoans? Really? Probably something else like the old corporate saying: no executive ever got fired for buying IMB. Similarly, no coach ever got fired for starting a black cornerback, but many could imagine getting fired for starting a white cornerback and he gives up a touchdown.
So, here's the question that could be studied: how much different would be a world where nobody perceived race at all? How many more whites would have started at cornerback?
My guess is: a few. Not really all that many. Blacks really are much better at cornerback on average than nonblacks. But I could be wrong. Perhaps racial stereotyping kept dozens of whites and Samoans from playing cornerback.
Doesn't anybody want to know this?
Especially because it seems obviously useful to estimating the size of the harm you dread as people like Scott make it more socially respectable to be unignorant about race and IQ.
An interesting aspect of the widespread ignorance of how huge is the intelligence gap between Asians and blacks is that college admission affirmative action is highly unpopular with the public, even though affirmative action is the only way to have more than negligible black representation at our most elite academic institutions. I've often joked that when you become president of Harvard, they give you the key to unlock the Harvard President's Book of Secrets, which turns out to be a dog-eared copy of "The Bell Curve."
The point isn’t that what you’re saying isn’t true, that point is that for most people who have heard of Steve Sailer, Steve Sailer is the textbook example of a scientific racist. So I think we have a good working definition of scientific racism, which is just, whatever Steve Sailer says. I say this unironically with no disrespect; I also consider myself a scientific racist.
Some people consider me a textbook example of a public-spirited realist.
Kenyans, Ethiopians and Ugandans who do well in middle and long distance running grow up and train at altitude where the air is thin. The body adjusts to the conditions and becomes more efficient at handling oxygen. Uganda recently build a training centre at altitude open to athletes from all over the world. The distance running genes haven't been discovered (yet).
Kenyans and Ethiopians also do pretty well in distance running when they move to the US and compete at low altitude, or when their parents do.
For example, in the all-time great 1500 meter men's final in 2024 Olympics, the bronze metal was won by Yared Nuguse, an Ethiopian-American who was born in Arlington, Virginia and went to high school in Louisville.
https://en.wikipedia.org/wiki/Yared_Nuguse
I looked up the top 160 high school cross country runners in America about 15 or 20 years ago, and 9% of them were of East African ancestry.
Setting aside the obvious problems (that AA programs are textbook-racist under the above definition), I think it's going to be very difficult, in practice, to acknowledge the 'biologically distinct taxa' element without leading to the 'discrimination/hierarchy' element.
"We should always judge individuals on their individual merits" all sounds very fine and noble, but you can't evaluate individual merit unless you have perfect information about individuals, which is never *perfectly* true and often largely missing, especially from people who arrive on rubber dinghys or chuck out their passports on arrival. And in any case even carefully-screened individuals are bound to their background groups through loyalty to kinship networks and regression-to-the-mean effects, which we can't reliably predict at an individual level. It's not irrational or unjustified to fill in the blanks here with probabilistic information derived from an individual's background group, which is definitionally 'racist'.
The case for this can be made with or without reference to genetic factors, of course, but the case certainly gets stronger when genetics enters the picture.
This doesn't make, e.g, assimilation of some non-zero number of foreign migrants necessarily *impossible*, but to make it work you're talking about a multi-generational process with necessarily higher standards of scrutiny than would be required of native-born citizens to compensate for any background risk-factors, which... again, is definitionally 'racist'.
>obvious problems (that AA programs are textbook-racist under the above definition)
Why is that an obvious problem? Can it not be the case that they are textbook-racist and people still find them worthwhile?
Okay, but then there's such a thing as 'good racism'.
What changed during the Great Awokening versus the previous several decades is that expressions of racial animus against whites in the mainstream media became not just acceptable but fashionable.
I favor doing as much as is practical to get better information about individuals: e.g., colleges should go back to requiring the SAT or ACT.
But I am against adjusting scores for reasons of racial regression toward the mean. For example, sure, a black applicant who gets a perfect 1600 on the SAT the one time he takes it would be likely to average a little lower than an Asian applicant who got a 1600 the one time he takes it if they both took it nine more times. But let's not penalize him for that statistical pattern. He got a 1600. Good for him.
On the other hand, say your mother's car breaks down in the middle of a dark block and her phone isn't charged, so she has to decide in which direction to walk. Hanging out on one corner are a half dozen Asian-American teens and hanging out on the other corner are a half dozen African-American teens. I think she can be forgiven in this case for doing some racial profiling and walking toward the Asian-Americans.
I was referring to regression-to-the-mean (maybe a misnomer, I think Galton used the term 'regression-towards-mediocrity') as a phenomenon seen among the children of exceptional parents- e.g, the children of very short people tend to be taller, on average, and the same can be seen at both ends of the bell curve for any complex human trait, due to a combination of non-additive genetic and non-systematic environmental factors.
Notably, as you probably know, different racial groups regress toward different means for IQ (up until the advent of GWAS, this was one of the more powerful pieces of evidence for hereditarianism.) This is relevant to long-term migration policy, but also in the broader sense to any attempt at integrating exceptional individuals into meritocratic societies or institutions.
For example, the children of wealthy black parents do about as well on the SAT as white kids who grow up in trailer parks. And... I have to imagine this could strongly motivate parents of colour to push for the... shall we say, relaxation of meritocratic standards, when it comes to their child? Potentially true for any parent, of course, but more likely to be true the larger the regression effect. And this has to be a factor when considering whether to include such parents in the social milieu that decide standards of merit to begin with.
Looking back, I think that the definition I was using was from Omi and Winant. They added an "essentialism" component.
So that rescues most AA programs.
By way of example, let's say that the Nazis put all the Jews in camps and stole their property. You could, without being racist, say that the Jews should be compensated - not because of some essentialist characteristic of being Jewish, but because of their historical experience suffering in Nazi concentration camps.
As for the rest of it, you don't have perfect information to treat individuals the same. I don't run a background check on the people I meet, though I'm sure it would reveal lots of information about whom to trust and whom to shun. Instead, I just treat people like they're all normal, unless and until I get individualized data saying that someone is, say, a convicted felon, or something.
People make decisions based on imperfect information every day. Unfortunately, race is something that's really hard to not see!
I'm not sure I buy that AA programs can be justified on any non-essentialist / contingent grounds: the bar to clear for receiving benefits is "be black", right, not any feature equivalent to "experienced XYZ event or type of event"—or, in other words, if it were that being black was any sort of proxy, there'd be a better way to get at the thing we /really/ care about without need for a proxy.
(e.g., a middle-class black woman from a recent-immigrant family applies for some benefit under AA: well, she & her ancestors experienced no slavery, or none the U.S. was complicit in; she isn't poorer than other applicants; she isn't disadvantaged in any way relative to other applicants; etc... she's just black, and that's the sole criterion.)
Unless I'm wrong and AA in practice requires that one prove U.S. slave ancestry or other white American oppression against you & yours in particular, or something...? I think most proponents are against anything like this, though, even something like "just give help to poor people as a whole"—some have come up with various ways to try to justify making it race-based, but (if I may do some armchair psychologizing) I'd bet you $5 that nah, they see it in essentialist terms in their hearts of hearts. So to speak.
I understand your point, but the problem with allocating AA on the basis of, e.g, poverty is that poverty just doesn't predict social mobility across racial backgrounds, so it wouldn't have the required 'compensatory' effect.
So a couple points. First, I don't think your characterization of the facts is accurate. In California, where I live, the reparations task force said that reparations should be "based on lineage, determined by an individual being a Black descendant of a chattel enslaved person or a descendant of a free Black person living in the United States prior to the end of the 19th Century".
Second, with that fact in mind, I don't think the facts matter all that much for this discussion. I know literally zero people who would support (or oppose) a lineage-based reparation program but who would change their mind for just a race-based reparation program.
So I'll agree that most people who support AA/reparations are racist, albeit in a direction that runs contrary to the bulk of American history. I think that's why the people in charge of, say, California's reparations task force took pains to separate race from lineage, to avoid charges of racism.
Of course, that got them zero converts, because normal people aren't living based on what Omi and Winant were writing about. They're not trying to conform their lives to some ideology - they have vibes and make decisions based on those vibes.
But people who care about the ideology do, in fact, have answers to these questions. I just don't think they represent too many real people.
I agree that people injured through either malice or incompetence deserve compensation, but I also don't think you can just measure per-capita GDP disparities and use this to infer injury, so I want to know what level of compensation is meant to be sufficient. And I've written elsewhere why I don't think affirmative-action programs are an appropriate form of compensation under any circumstances.
https://www.astralcodexten.com/p/how-to-stop-worrying-and-learn-to/comment/86414230
https://www.astralcodexten.com/p/highlights-from-the-comments-on-lynn/comment/86697337
If you mean 'treat the same' in the sense of 'receive the same nominal courtesies during casual conversation', I mean... sure, politeness doesn't cost much. But the risk/reward evaluation can become quite different if you're deciding who to rent an apartment to, or approve for a bank loan, or hire for a job. Once upon a time, you could, perhaps, rely on credentialling institutions or background checks or just asking face-to-face questions to fill in the blanks, but the left has increasingly made it *illegal* to seek out such individuating information, and has actively corrupted every credentialing institution to reduce the exposure of embarrassing facts.
Racism is like democracy or love--it has dozens of different meanings used in different contexts, and often even the speaker isn't aware when they switch from one meaning to another.
To be useful, we really need to distinguish the moral / social stuff from the factual claims. You can't learn anything about the truth or falsehood of a factual claim by its moral or political or social implications. It would be bad if blacks' lower average IQ in the US were due to genetic factors that couldn't be changed, just as it would be bad if burning the cheapest available fuels messed up the climate. But if you let "it would be bad if this were true" influence your evaluation of "is this claim true," you are sabotaging your brain.
Why would it be "bad"? It's simply information that can be used to one's advantage. Information is power, it has no moral weight to it.
It would imply bad things about the future of the country, it would be unfortunate, we would prefer it to be otherwise.
I would suggest accepting reality as it is, not what you want it to be. Especially when you aren't capable of changing it.
Truth has moral weight.
...Yes, that is your opinion. I personally think it's very silly to ascribe moral character to arbitrary pieces of information, but you do you.
I can't seem to find the option to 'like' posts, but +1.
I agree with that, although in some circumstances, it makes sense to circle back and check with your moral / social stuff before publishing the information. Privacy is an obvious example - if my social security number gets released, I'm mad about people obtaining truthful information about me.
Regardless, the bar to clear for "This is true, but for moral / social reasons it must be suppressed" is really, really, really high. I can understand people who want to suppress the study of racial intelligence, but I'm pretty skeptical of the whole endeavor (leaving aside the question of whether it does more harm than good, or whatever).
In that case it still would be bad to publish a false social security number, because that can impact a random person with that number.
A problem is that once you start lying about the truth, you have to expand that lie to deal with hard questions. And suddenly all kinds of useful science becomes off limits.
And you need other explanations for certain things, and in practice this often seems to boil down to racist conspiracy theories that are remarkable similar to those wielded against Jews with rather nasty consequences. So is it really preferable to have people believe those conspiracy theories?
Well said. Funny how no one ever makes this kind of argument to, say, deny the Holocaust or something. "But if you believe than that, then you must think humans are terrible and there's no hope for the future!" Well...maybe? Not necessarily, but regardless... "Therefore it didn't happen" Non sequitur! "and anyone who says it did is an evil person who needs to be shunned and banished from polite company."
Oh, so you claim there's "evidence" that the Holocaust was real? Why are you interested in that? Do you LIKE the idea of millions of Jews being gassed? Gross! What a vile and disgusting person you are!
What about something like "people of X race are stupid, but we shouldn't give people more or less resources just because they're stupid"?
And a somewhat related question: if you think people of X race are stupid, and that they should get disability benefits as a result, are you racist against X or for X?
You kinda have to pick your battles. By way of example, I'm a Dodgers fan. I think that Giants fans are awful and stupid. All right-thinking people agree. But the kind of discrimination that Giants fans have historically endured is pretty small potatoes compared to racial, religious, or gender-based discrimination. If we could stop allocating resources based on racial, religious or gender lines, we'd be doing really, really well. I'd be extremely happy to live in a country where race was just as likely to predict your future financial success as what baseball team you root for.
To your second question, I was remembering the definition of racism used by Omi and Winant and I think they snuck in some requirement of essentialism. So if you believe that people of X race are essentially stupid, and therefore they should get more resources, yeah, that's racist.
Of course, I don't see many historical examples where that ever happened, so I don't think there's any cause for alarm. If in the future racist people start being far too kind and friendly and caring to other races, I'm sure we can figure out a solution.
The question is, of course, /why/ race predicts future financial status. If, hypothetically, the answer were to be "because some races are better at deferring gratification and have more smarts, on average", ought we still engage in redistribution along purely racial lines?
I think we pretty much always end up better not using race as a proxy, regardless—you have to engage in various unintuitive contortions to rescue such a policy over "just help whoever appears to need it most" without worrying about the racial balance of the recipients—but, considering how many people are only too glad to engage in such contortions, and how academia looks upon those who find for the wrong side on questions like "is it mostly differential intellectual ability, or mostly evil whitie being all oppressive and shit?", I'm not hopeful for the future here.
I don't know that we do engage in redistribution along purely racial lines. But if we did - and we just let the chips fall where they may - I'm pretty sure we'd stop caring about the people in Ohio and let unlimited numbers of Chinese people immigrate here.
Strangely, the people who care the most about race and intelligence never seem content to let the people of West Virginia sink into deserved oblivion as the more intelligent Japanese and Chinese people move in next door and buy up everything.
Quite the contrary, they want to ban Chinese investment, like TikTok, rather than just acquiescing in the face of the Asians' superior brain power.
Would you be okay with getting rid of all race/nationality/citizenship distinctions and letting individual merit be the only factor considered for, say, housing, employment, voting rights, etc.?
If you think IQ differences are leading to differences in outcome that should be compensated for by welfare or something, how does race even come into that? You would want to do that by individuals, not by races--a black doctor or engineer doesn't need any compensation for having a low IQ, whereas a white or Asian janitor may need that compensation.
You can have *some* level of social safety net, but, but how do you propose to allocate equal resources to all people regardless of their own productivity? This was tried a couple times and didn't work out well.
Even the 'safety net' approach still requires a reasonable number of net tax payers, by the way, which in turn depends on the typical genetic endowments of your country's citizens.
Most redistribution works out great - human beings depend on redistribution to survive to their teenage years.
Literally no newborn human can support themselves. Not a single one. They depend on net contributions from someone - usually their parents. This strategy differs from almost every other animal or plant ever encountered. Yet human beings are the ones who've taken over the world.
I think it's totally fine to take care of our fellow human beings - after all, every single one of us was taken care of at some point in our lives.
There are many examples of social species that exhibit reciprocal aid and essentially all mammals and birds provide parental care, so this is hardly a unique human innovation.
Do you think that social animals are immune to Darwinism?
Nobody sensible is saying people of race X are stupid. If you can't distinguish "the mean IQ of people of race X is lower than the mean IQ of people of race Y" from "people of race X are stupid," you're basically doomed to talk nonsense in this whole conversation. Seriously, a vast amount of nonsense on this topic comes from people who can't keep the difference between the mean and the distribution straight in their heads.
Who doesn't think there are genetic differences between the races?
The counter claim is usually that "race" is an arbitrary way of grouping people that doesn't map well onto real populations. You see the conceptual confusion Scott has in these articles in his treating of "black" as a natural kind.
It maps well enough to self-identified race in the US to have over 99% accuracy:
https://med.stanford.edu/news/all-news/2005/01/racial-groupings-match-genetic-profiles-stanford-study-finds.html
I think you may be reading too much into it. That study looked at four broad racial categories and discovered that they could tell which of the participants was in what category by their genes— the participants being people who were pre-selected to be among those four categories with none of the participants "mixed-race".
African-Americans are a mixed race population, as are mestizo Hispanics. Discrete notions of race do break down if you don't have concepts like "mestizo" or "mulatto" (much less "quadroon"), which are more popular in Spanish than English-speaking settler colonies, but it's still a fact that there are genetic differences that multiple services can now use to portion out your ancestry.
I don't think that obviates the point, which is that the population breaks down into separate races as well genetically as it does socially; hence, it isn't arbitrary: these are real genetic categories, and the expectation would have been (and, previously, a common claim was that) "conventional" race /wouldn't/ map to genetics in this way. Compare the prediction for how well an actually arbitrary grouping would be sliced along genetic joints (e.g., "people who like lettuce" vs. "people who like spinach", or something, barring some SNP that turns out to mediate this by itself or whatever—you could probably eventually get okay at slotting participants into the correct category, but 99% would be surprising).
Cluster analysis supports this as well—the only arbitrary portion being how many clusters you choose, which many commonly misunderstand as somehow a relevant consideration—and it's also just what you'd expect when eyeballing the geographical population barriers (& hence structures) of the world (if something , like a mountain range or ocean, is in between two populations in different environments that thus don't do much mixing for X millennia, one would expect to end up with two genetically-distinct groups).
Is black, not the race Black, but the color black, a natural kind? If yes, why not the race? If not, how is Scott confused is he treats Black and black as scientific concepts and not natural kinds?
Black the color is the absence of reflected light, and basically every culture has words to distinguish black from white. As you go up in the number of colors, the percentage of cultures with distinct words for them goes down.
In terms of ancestral race, African vs non-African is one of the biggest divides, since all non-Africans are descended from a relatively small group of (presumably) east Africans who then appeared to have mixed with Neandertals in the Middle East before splitting again into west & east branches (with the east branch mixing with Denisovans). But supposedly even within Africa there are splits that go back even further between the majority of Africans vs Khoi-San or pygmies.
A complication is that certain Oceanians (such as the aptly named "Melanesians" as well as Australian Aborigenes) have been called "black", because they have dark skin. They aren't any more closely related to Africans than any other non-African, in fact their Denisovan ancestry makes them LESS related to Africans than Europeans are.
The color black was a bad example lol.
The point is that red, as distinguished from orange, will have many edge cases, and the dividing line between them will appear arbitrary. From this, it appears that either there is no Platonic ideal/natural kind that is “redness”, or there are infinitely many; either way as humans we rely on our concept of redness, somewhat arbitrarily defined. Yet there is a fact of the matter that apples are red, and bananas are not.
Similarly, the division of humans into races will be somewhat arbitrarily defined, but there is a fact of the matter that LeBron is Black and Messi is not, even if Black is not a natural kind.
You hear that all the time, but what are the implications?
For example, on the 2010 Census, Barack Obama counted as black and only black because that's what he chose to self-identify as (according to a White House press release). Of course, he had a white mother.
If you don't count as Barack Obama as black, would that make the white-black IQ gap smaller?
No.
We've got the DNA technology now to measure racial admixture to 3 decimal places. That's now been done with two databases of around ten thousand young people each for which we also have their IQ scores: the national ABCD database and the Philadelphia Neurodevelopment database. It turns out that the more white ancestry a self-identified African American has, the higher his IQ on average. The effect is not huge but it's definitely there.
Is the 'One Drop Rule' no longer used? What's the percentage/threshold required to be black? (3 decimal places).
In the US, the rule for being black for affirmative action purposes seems to be that you must have at least one parent who self-identified as a member of the black community and was socially recognized as a member of the black community.
So, you can't claim to deserve affirmative action just because one of your ancestors 1024 ancestors ten generations ago was black.
That actually works surprisingly well. We have strikingly few disputes over who is black for affirmative action purposes.
One issue would be if your parent was part black but decided to pass as white. Can you reclaim blackness? But, ever since the introduction of affirmative action in 1969, not that many people who could justify a claim to being black have chosen to self-identify as white. State and society now provide benefits to identifying as black, so most people who can reasonably claim to belong to the legally privileged race.
The rules for claiming American Indian ancestry are more specific for individual tribes and more vague for the whole race.
For being Hispanic, having one Hispanic grandparent seems to work, but several judges have felt that being 1/8th Hispanic was ridiculous.
I don't think there are genetic differences between races in any *meaningful* or *persistent* sense, only in a technical and temporary sense. Gene alleles aren't locked to races, but to individuals, and through combinations of intermarriage and geographical isolation over time, alleles that are common in individuals in one group can transfer into and become prevalent in another group, and over long periods of time racial groupings at one time can dissolve and resort into completely different racial groupings at another time (and in fact, by investigating ancient human DNA we have proof that this has in fact happened).
I think those alleles have persisted for a LONG time. You're right that intermarriage will result in recombination, but we're still far from panmixia (Brazil is probably the place where socially constructed race has diverged most form genetic ancestry, but most of the world isn't Brazil).
I'm not just talking about what will happen, but what has already happened. I highly recommend the book *Who We Are and How We Got Here* by David Reich. To quote the summary from wikipedia, "almost all human populations are mixtures resulting from multiple population migrations and gene flow. "
Yes, modern day populations are the results of even more ancient populations mixing. If you go back far enough it wouldn't make sense to talk of people in terms we use for modern populations. This doesn't change the fact that modern populations are distinct from each other.
So, uh, how can 23andMe work?
People do, in fact, vary significantly.
Yeah, but you're talking about a process that takes centuries at minimum, if not thousands of years, and many of these large-scale migration+admixture events were heavily co-extensive with genocide (the Bantu expansion and Indo-European invasions, for example.)
I don't see how that really resolves the anxiety of anyone concerned with present-day race relations.
Why do people of African descent have dark skin? I always assumed it was a genetic thing, but if there aren't meaningful genetic differences between races, it must be something else.
No two real distributions of non-integer values will have the same mean value. IQ is a root summed square of a nearly astronomical number of genetic components and thus non-integer. Accordingly, every race has different mean IQ. So do men and women. Clearly the definition of "scientific racist" is not, "actually understands statistics." I think you need a new definition.
The formally correct response to your criticism "yes there might be differences but they are about 10^-8 order of magnitude and do not even align with observed phenotype differences"
Hard to tell whether 10^-8 "order of magnitude" is more arbitrary or more redundant.
I hope we don't ignore 10^-8-level differences when we decide whether we need to screen early for hypertension, skin cancer, or Taye-Sachs.
Couple of nitpicks: if the genetic components were actually "nearly astronomical", with no systematic difference between the groups, the quantity of resulting from their averaged impact would actually have negligible variance (1/sqrt(N)). To get appreciable differences, you need either a small number of random unbiased factors (such that 1/sqrt(N) is not too small), or a bias. But it would be a miracle if selection pressures for intelligence and for the other effects of the involved genes had been identical for all population groups.
The question is, are the differences large enough to be reliably detectable, and relevant for practical purposes. That is an empirical question that cannot be answered from theoretical arguments.
I am not sure that systematic is the correct word there. I also believe that individual genes aren't exactly identical in their contribution (there is no reason to believe they would be so not 1/sqrt(N)). On the other hand, obviously the collection of genes present in various populations are neither identical nor randomly distributed and thus the variation in populations without heavy crossbreeding are not merely noise.
I was making the more narrow point that nobody should be surprised that there are variances between distributions (it's not Racist to observe fact). The presence or absence of racism shows up more in an individual's obsession and need to shoehorn it into every conversation.
Scientific racism is the idea that there are genetic differences between the races, and that these differences result in certain races being smarter or more law-abiding and that sort of thing. Basically, that it makes sense to judge people on their race. As opposed to things like the genes that control skin color being different so you can easily identify their race, but it's totally useless to figure out which of two people you should hire to do a certain job.
So, I understand the definition you're using, but isn't it kind-of odd that scientific racists like Sailer and Murray advocate for hiring / admitting people to school entirely on merit, whereas antiracists advocate for hiring/admitting to school based on race?
One last comment on changes: I could not find anywhere that any proponents of Lynn’s data had ever shown what it shows about changes. I had to download the record-level data (not the country aggregates they estimate) and calculate it myself. My sense is that the researchers who favor this data will probably just respond by saying: all the studies are noisy proxies for stable underlying traits, the apparent changes are fake, etc. And they will point out that we observe the least change in the best measured countries. My response to that would be: if you think these are all noisy proxies for a latent trait you should be calculated national IQs with a latent class model not what you are actually doing, and also, isn’t this a huge own-goal on the credibility of the data that you don’t even think your own selected estimates within the same country are properly comparable? If same country estimates can’t be compared, how can we trust the cross country estimates? But I will be interested to see what responses you get!
I published an article about 15 or 20 years ago for which I graphed all the changes over time of national IQ scores in Lynn's database to see if the Flynn Effect was closing gaps. You could seem some growth in scores in northeast Asia. But it mostly looked like the gaps were about the same.
I couldn't tell you about in the 21st Century.
But you might remember when the PISA test was fashionable in the press around 2013. But there isn't much change in results every three years, so the media has gotten bored with it.
Since you're here, are you really confident that the PGS score differences between ancient and modern europeans only imply a ~10-IQ point difference? Any selection pressures that impacted the genes used to compose a PGS score probably impacted the frequency of a bunch of other genes for intelligence we haven't identified yet.
I like the dichotomy of abstract vs. concrete thinking skills a lot.
Is a third grader IQ really as low as 60? I thought once someone can read and understand pictures, their measured IQ should more or less lock in, in this case to the average 100?
I think there's a normalization thing going on here. My kids took IQ tests at age four, and got scores much higher than 60, but they weren't the same IQ tests that you would give an adult.
So maybe a third grader scores 60 on an adult IQ test, but 100 on the version of the test calibrated for third graders.
I thought about that but I don't think third grader would be that different to adult (certainly compared to 4 years old). Is there any non-normalized data somewhere?
Not what you're asking for but maybe intuitively helpful is the original way of calculating IQ scores, which worked something like this:
Suppose IQ doesn't increase past age 16. Someone 16 years old or older that scored like a typical 9 year old on an IQ test would have an IQ of 9/16 = (0.)56
So typical 9 year old would be equivalent to an adult with an IQ of 56 in terms of test scores.
I think this sort of calculation was used to justify Marilyn Vos Savant's 220+ IQ for inclusion in the Guinness Book of Records before they removed that category. She supposedly tested at the mental age of a 22 year old when she was 10. This is all from vague memory though so take it with a grain of salt.
Yeah, but if IQ doesn't increase much after 16 that's obviously silly. Like if a 10-year old is as smart as the typical (half-senile) 90-year old, that would give him an IQ of 900.
I used an item response model to generate IQ scores from raw response data in the Louisville Twin Study. Subject to many limitations and assumptions, it looked like 15 year olds were roughly as smart as adults, and 7 year olds were ~2 SDs worse on the most age-sensitive tasks and ~1 SD worse on the least age-sensitive. So I'd expect 3rd graders to be more like 1.5 SDs worse overall (or 78 IQ on the adult scale).
There is no way a third grader could finish any college degree. Ten years later they can. If capable.
I think a lot of that is emotional control, accumulated knowledge, and other "maturity" aspects. My understanding is in accordance with yours that a third grader simply does not have the abstract reasoning capabilities to major in, say, theoretical physics or something. But there are plenty of majors that I think a fairly (non-child genius) third grader could get through if they had sufficient maturity and self-control.
I was offered placement in college when I was in second grade grade (after the school "accidentally" gave me an IQ test instead of the test to skip the third grade they were supposed to be giving me)
I turned it down, which my parents were happy about (they were concerned it would affect my social development - I, meanwhile, turned it down because my understanding of college at the time was based on media, and it was that I'd have to live on campus with a probably-very-annoying roommate who would constantly be annoying me with having girls over.)
Could I have completed the degree?
Yeah, no problem. I was learning out of a college-level chemistry book a couple of years later without issue.
Regarding the abstract vs concrete reasoning and so on.
Smartphones are now, AIUI, pretty much ubiquitous in Africa. If you can use a smartphone to check the market prices in distant location or transfer money to vendors etc., which is again from my understanding, common for sub-Saharan Africans then presumably you are exposed to abstract thinking skills and probably from a relatively early age now. So I wonder whether more recent IQ surveys will show an increase in apparent IQ for these countries.
Just in time for smartphones to make humans dumber by doing all the hard thinking for us!
Right, rural electrification is changing Africa pretty fast these days.
In the earlier comments, I linked to a study that found an 11 point rise in kids' IQ scores in a rural area in Kenya in just the 14 years of 1984-1998 in which a lot of good things had happened to bring it into the modern world.
Cool.
I forget the details, but life in this village in Kenya was much more mentally stimulating in 1998 than it had been in 1984.
Is this the record for shortest gap between "Post" and "Highlights From the Comments on Post"?
I've taught students ranging from about 70 IQ to future Ivy League grad students in the hardest of hard sciences and math, and I can affirm that in casual conversation, you cannot necessarily differentiate people with IQ's from 70-85 from those with IQ's well above that. Sometimes people with extremely high capacities seem almost disabled. One young woman (currently completing a Ph.D. in math from an Ivy League school) I'll call Dee seemed rather slow to me at first. She rarely spoke and seemed to have no particular reaction to literature even though she was "supposedly" reading it.
Then you try to teach them something . . . and everything becomes clear. A student of 13, whom I'll call Mel, who has an IQ of 70, struggles with second-grade level material that is easy for a bright seven-year-old. Future grad student Dee turns out to learn everything without being taught it at all, and adds some things I did not even think of myself, and does all of that while learning classical violin and mastering two foreign languages.
I gave group-administered IQ tests (the CogAT) and can tell you that students with high IQ's can overcome specific skill deficits with dazzling speed, but those with sub-normal IQ's plod along making at best a year's progress if they have heroically determined and inspiring teachers using the best possible materials.
I found it all rather bad news.
The other thing I found was that we could raise CogAT scores a standard deviation or so (about 12-16 points) with kindergartners, but after that, five points was a really good gain. And even that flattened out, eventually.
I wonder how much of the difference between "Mel" and "Dee" is due to an inability of Mel to compute the result of calculations, and how much is due to Dee having far better memory. I note for myself that I can understand mathematics perfectly well, but it just doesn't stick in my memory. I have to go and look up formulas every time I need to use them.
Are the sub-normal IQs failing to understand concepts, or are they failing to memorize them? If it's generally the first, that's a harder problem to do with the general architecture of their brain, but if it's the second we can perhaps conceive in the future of some kind of memory aiding medicine that strengthens new synaptic connections.
Being good at mathematics isn't the same as "remembering formulas".
Yeah, in particular if you're good at math you can usually just rederive a lot of the formulas if you've forgotten them
Of course not. But if you have to constantly retread the same ground, you end up not making a lot of progress. You certainly aren't going to be discovering anything substantially novel.
Being good at anything intellectual requires some kind of reliable short term and long term memory, and math absolutely requires this. Remembering formulae is just one example, and it's not a small one.
They fail to understand concepts and principles. They also fail to retain them, I suppose, but it is very difficult to remember what you do not understand.
"I note for myself that I can understand mathematics perfectly well, but it just doesn't stick in my memory. I have to go and look up formulas every time I need to use them."
I have a PhD in algebra. While having good memory is certainly advantageous, even professional mathematicians consult formula tables quite often. Bartsch's collection of mathematical formulas, a book I used almost daily during my studies, has 830 pages!
Memory is only an auxiliary tool, though. The ability to grasp ("grok") abstract concepts is what matters. If you don't understand what a polynomial is, you will struggle with every non-trivial task regarding a polynomial, even if you had almost perfect memory.
But if you understand just fine what a polynomial is, and then have to relearn it almost from scratch every time it comes up, you will make very slow progress on more advanced math.
I haven't met anyone whose memory would be SO bad...
I met a lot of people who "shallowly forget" things, including math structures, but as long as the time between work sessions that use those structures isn't terribly long, they refresh their knowledge much faster. I.e. the second, third ... etc. instance isn't really "learning from scratch".
If someone suffers from such a deep memory defect, they are basically disabled.
Well yes, I was using an extreme example to illustrate. But I think a big part of learning any difficult topic is how many repetitions it takes, and how close together they have to be, before you just *know* a concept off the top of your head and can easily use it to help you understand more advanced ideas. I don't think IQ tests can measure this - they measure short term memory, which is different. But maybe they don't have to because it correlates well enough with other aspects of intelligence.
Your first remark reminded me of Sarah Constantin's https://srconstantin.wordpress.com/2019/02/25/humans-who-are-not-concentrating-are-not-general-intelligences/ Long quote:
"I’ve noticed that I cannot tell, from casual conversation, whether someone is intelligent in the IQ sense.
I’ve interviewed job applicants, and perceived them all as “bright and impressive”, but found that the vast majority of them could not solve a simple math problem. The ones who could solve the problem didn’t appear any “brighter” in conversation than the ones who couldn’t.
I’ve taught public school teachers, who were incredibly bad at formal mathematical reasoning (I know, because I graded their tests), to the point that I had not realized humans could be that bad at math — but it had no effect on how they came across in friendly conversation after hours. They didn’t seem “dopey” or “slow”, they were witty and engaging and warm.
I’ve read the personal blogs of intellectually disabled people — people who, by definition, score poorly on IQ tests — and they don’t read as any less funny or creative or relatable than anyone else.
Whatever ability IQ tests and math tests measure, I believe that lacking that ability doesn’t have any effect on one’s ability to make a good social impression or even to “seem smart” in conversation."
Ehhh ... I can tell from a few 280-character maximum tweets about how smart somebody is.
Scott, for example, spends a lot of time clearing his throat to bore the chuds into leaving, but he's obviously immensely intelligent.
I also engage with teachers most of my working life and you can definitely tell which ones are the bright ones. Just like everywhere else in society. I'm not sure what she's on about here, it doesn't pass the smell test.
An observation I made when I was 8(?), is that lots of people get by with mimicry. They just run social-scripts. And most social situations don't really require much more than that. (When the script is thrown out the window, things suddenly feel "awkward"). The inability for some to think for themselves, when called upon, is something I've found frustrating for as long as I can remember.
And this plays into why I'm not really surprised that the current ML paradigm is able to pass a Turing Test by just aping what it sees online. Also, something something Heinrich.
I have a coworker who is a mathematician by training. He is very slow and deliberative, and you could easily mistake him for not being very smart because of that. This would be pretty badly wrong, but not every smart person is fast on their feet in a conversation.
That seems consistent with what I've read of the literature, yes. You can get impressive gains in really young kids, but heritability increases with age and there doesn't seem to be any way to make gains from intervention permanent. In developed countries, at least.
Was going to note that my aforementioned foster-kid who was a 63 was pretty normal (well, normal for an 11-year old, anyway) in conversation.
It was when he was doing his homework and he was struggling through the kind of homework my 5 year-old blows through fine at pre-school that the light clicked on.
FWIW, his sister (also a placement, not tested as far as I know), was notably not much better at reading than him; I'd guess she might have been somewhat better but not in any meaningful way.
"The part of Lyman’s comment that gives me the most pause is his observation that, if the mean IQ is 60, a decent fraction of people must be 45, and a non-negligible portion 30. At this point, even third-grader comparisons don’t save us. I guess this is where I bring in the claim that IQ breaks down as a guide to practical living skills below some point."
Or, what could break down is IQ being normally distributed. If IQ fails to follow a normal distribution below a certain level, we wouldn't expect the usual proportions of people to be 1 or 2 SD below the mean.
This is certainly true. However, non-normality could be instructive. If it is true that symbolic learning has a huge impact on IQs in Africa, then one might expect that rather than shifting a mean (and obviously with it sigma), it would effectively just make a significant number of questions unanswerable to most takers. This ought to show up in skewness with a much larger sigma for takers above the mean than below.
If IQ is bounded by zero on the left, it is not a pure normal distribution. Truncated normal distributions are much harder to work with mathematically.
Why would it be bounded by 0?
What would a negative IQ even mean? I'm probably too dumb to conceptualize such a thing :(
The opposite of IQ 200+. Minus 7 standard deviations. Only 30 people on literally the entire planet are dumber than that.
EDIT: Seems I got my math in the last sentence wrong. It is more like "literally the dumbest human ever".
"...and they all work at the New York Times"
That the average IQ is 100 and that the standard deviation is 15 are just arbitrary conventions, so there is nothing special about an IQ of 0. (Setting the average, which is the only special value in the distribution, to 0 would have made a lot more sense to me, but there are historical reasons that it is the way it is.) If IQ were perfectly normally distributed, approximately one person in 76 429 666 480 would have an IQ <0 (ditto for IQ > 200). In reality there are many more people with very low IQs than one would expect from a normal distribution, because there are single mutations that have a large negative effect on intelligence (whereas a normal distribution would be the result of a very large number of genes each with a very small effect on IQ).
> In reality there are many more people with very low IQs than one would expect from a normal distribution, because there are single mutations that have a large negative effect on intelligence
That only makes sense if you have some absolute way of measuring IQ. Given two smart people of slightly different intelligence and two dumb people of slightly different intelligence, how can you compare the differences? You could just give everyone a test and see how many questions they got right, but the distribution will depend heavily on the test.
I agree that the difference in intelligence between the lowest 0.0001% of the population and average is more than the difference between average and the highest 0.0001%, since the lowest 0.0001% is someone in a coma and the highest is still just comparing two people, but I don't see how you could define differences in a precise way that can reflect that.
Yeah, I can't argue with you there, actually. I do wonder what it means when people say that the IQ distribution has a thick left tail – like where does that aberration from normality come from when IQ is supposed to be constructed to be normal.
Wouldn't it just turn into a Poisson distribution?
Mean 100, sigma 15, so the section of the curve below 0 is negligible.
Lyman's post seems to be riddled with errors. There are theoretical issues, like his total misunderstanding of intellectual disability, but even on the most basic grounds, the most recent estimates for countries like Haiti and Cambodia are simply not what he portrays them as: https://www.sebjenseb.net/p/average-iq-of-haiti, https://www.sebjenseb.net/p/changes-in-relative-cognitive-performance.
The idea that changes in IQ aren't correlated with changes in development makes total sense, because the changes in IQ in question are generally not reliable even though the mean values are. I also doubt his analysis was done correctly here, since he portrayed incorrect national IQ estimates. It's a confused point followed up by points that aren't even clear, like "By the way— a pure hereditarian would insist that African IQ is around 80-90. Why? Because in the highest-quality data on African-ancestry people, namely, tests in the U.S., African-ancestry people tend to score around 80-90 on average."
Does this refer to African Americans, who are 15-25% European by admixture or to the selected immigrants who go to the U.S.? If it's the former, then a hereditarian would surely insist that the African IQ is the AA IQ minus the delta from European admixture. Depending on where estimates sit in his range, that could mean going below 80 or sitting in 80-90, so it's just unclear what he meant. He follows this unclear point up with "If you’re really a strong hereditarian, the U.S. environment should not boost IQ very much, which implies that the national IQ data is incorrect" which, as far as I can tell, is just describing a case no one believes in, including Lynn, who explicitly rejected it on multiple occasions, including his debate with Wicherts (see: https://www.sciencedirect.com/science/article/abs/pii/S1041608010000348).
A lot of other people also made weird claims, so I'll write something on this.
Thanks. It would be interesting to see your take.
This might be that take https://www.cremieux.xyz/p/national-iqs-are-valid
Yeah, I just saw it, thanks
Lyman is worth reading, as he grapples more with data than the typical pundit, but I wouldn't always say I find his conclusions reliable. For example, he seemed to be ruling out urbanization as an explanation for falling fertility when it was actually consistent with his data, and he just didn't have enough to be sure in the sense of statistical significance (which as Andrew Gelman never ceases of reminding us, is not the same as actually being correct) https://x.com/TeaGeeGeePea/status/1876509720707751980
Yes, I think this conclusion about the value of changes is particularly unreliable, since it doesn't deal with all the data and for the additional reason that it doesn't deal with the psychometric meaning of changes in scores.
As an example of the latter, I used the China Family Panel Studies to look at changes in Chinese IQs by age cohort and found that, adjusting for psychometric bias in the measurements, score trends over time were largely eliminated. Lyman's interpretation of this would be based on the raw scores, but the underlying ability level barely changed over time, so why should we expect those raw score changes to have any validity?
https://www.cremieux.xyz/i/146562912/terrible-times-make-for-powerful-studies
> couldn't alphabetize files
> successfuly subsistence farmers
Maybe this is just a specialization thing? I can alphabetize like a professional but then I spent years in school learning the alphabet among other Rs.
But if you plopped me onto a subsistence farm in the middle of nowhere, I reckon I'd be pretty crummy at surviving!
Professional Haitian bureaucrats (who were literate) couldn't alphabetize, not average Haitian farmer.
Why the assumption that the mentioned Haitians were "professional bureaucrats"?
It's impression that I got from reading Scott's experience from Haiti. Maybe this applies to lower end bureaucrats, i.e. clerks.
I would dispute that SSA's subsistence farmers have especially optimised strategies, by the way. I occasionally get newsletters from a charity called Forest Gardens International, which routinely teaches farmers in west africa to triple their agricultural yields, even without any additional use of artificial fertilisers or pesticides.
People forget that pre-haber-process improvements in britain's agriculture resulting from enclosure and crop rotation more than doubled farmers' yields, and britons had been living off subsistence farming for thousands of years beforehand.
https://en.wikipedia.org/wiki/British_Agricultural_Revolution
(EDIT: For the life of me I cannot google for Forest Gardens International any more? I could have sworn it existed, but maybe it was just Trees For The Future?)
Scott feels safe to post his actual opinions on race and IQ - woke really must be dead
That is not dead which can for four years lie, and with a new election even death may die.
Seriously, I just figured he was making more money off Substack than his day job. Or the NYT had already gone after him, so what else does he have to lose?
Strategically speaking, I can see one problem. He and a lot of other people around here are really worried about AI X-risk. (I am agnostic as I really don't understand the issues that well, but their concerns seem at least somewhat credible.) If you really think superintelligent AI is going to destroy the world, it would not make sense to associate yourself with one side of a controversial issue. You still have people on the left worried about it. If worries about AI X-risk become fully right-coded then that would make developing any kind of a consensus to limit or align AI properly much more difficult, and likely impossible.
If someone on the other side has to disavow anything associated with Scott Alexander, it's going to make it harder for her to argue for concern about anything other than AIs being racist in ways the left doesn't like. And then, if this AI X-risk thing is real, we're all dead.
"That is not dead which can for four years lie, and with a new election even death may die."
But it is hard to rekindle a fire from the ashes.
The fervor of 2021 isn't going to repeat itself anytime soon. People are too burnt out internally. In history, it is almost unheard of for a time of high ideological fervor to repeat itself soon, it usually takes several generations, if the ideology survives at all. Most ideologies die sooner that that. Some are very long-lived (traditional religions), but even there you see religious revivals every century or two or so. Not every five years.
Certainly the DEI bureaucracy etc. will live on and struggle to keep some grip on power, much like late-stage Communism worked (the parties left all the Stalinist passion behind and kept power for power's sake). But that is not the same as feverish witchhunts, and the overall climate is going to be a lot more tolerant of dissent.
I hope you are right and I am wrong.
The bureaucracy is still in place. The young (at least the women) and the elites in the Ivies and other citadels of privilege that staff the ruling class and their immediate underlings believe this stuff. PC was temporarily turned back in the 1990s and everyone on the right and many of the traditional liberals thought it was over and they came back with a vengeance in 2014 or so and gained a lot more ground.
> The bureaucracy is still in place. The young (at least the women) and the elites in the Ivies and other citadels of privilege that staff the ruling class and their immediate underlings believe this stuff.
Why do you think those are problems that can't be solved? Past governments in other countries have been able to force cultural and structural reforms with some effort. And a lot of bloodshed.
Well, DEI is going away because of a fear of judicial backlash. Once acting on woke (by introducing de-facto quotas etc.) runs the risk of being subject to a major lawsuit, even the academia will think twice. They have huge coffers and these can be raided for damages.
And without acting on woke, the system tends to hollow out.
It will be back in some other form with some other set of justifications and some other set of words.
Universities are woke reservoirs in the same way that monkeys are reservoirs for ebola. We may have fought off this epidemic but a new one is sure to rise eventually. It might not be for 20 years - sometimes it takes a generation for antibodies to fade.
It is unrealistic for woke to go entirely away on a short-to-medium timescale, but it is now in a much more vulnerable position than in, say, 2021.
Even in the academic sphere, there will be budgetary struggles soon, because the demographic development caught up with the universities and there will be fewer young people entering a vast system of higher education = less money overall.
I think we are looking at a future where woke exists, but has to compete against a whole team of other ideologies. The dominant position it held for a short time is not coming back.
Yes, I agree universities are in for a tough time and their position of cultural respect is definitely gone. It remains to be seen what the new equilibrium will look like but yes you do point out reasons to be optimistic. If I had Trump's ear I'd tell him to cut federal funding for any university that had either anything that looks like DEI or any grievance-studies dept (gender studies, chicano studies, etc). That would truly drive a stake through the heart of woke.
There will probably be some entirely new and exciting bit of numbskullery that takes over all the high-profile institutions for a few years that looks nothing like wokism/Kendism/etc.
Almost certainly so, fooling themselves is a popular pastime of intellectuals.
The best variant would be, if multiple competing creeds of this type emerged at the same time, with none achieving dominance and plunging the fanatics into an endless ideological civil war within academia.
<If you really think superintelligent AI is going to destroy the world, it would not make sense to associate yourself with one side of a controversial issue. You still have people on the left worried about it. If worries about AI X-risk become fully right-coded then that would make developing any kind of a consensus to limit or align AI properly much more difficult, and likely impossible.
That may be true, but I am pretty confident Scott does not think that way. He has an exceptionally strong drive to figure things out, and that accounts for most of the variance in what he says and how he says it. He probably does do some minor things mostly out of a desire to keep his audience happy, but they would be things on the order of putting up a humorous post now and then. He does not do the kind of political calculating you are talking about, and I don't think he would turn out to have much talent for it if he tried.
I don't know. He's committed to the right now. Whether that's good or bad remains to be seen.
He's committed to saying and doing things (like this) that will make him persona non grata among the wokest parts of the left. That is not even close to being the same thing as "committed to the right".
In our stupid binary political system, that makes him a right-coded figure. The guy's a Democrat in San Francisco in a polyamorous relationship but the only guys who quote him are conservatives.
If we were multiparty he'd be in a 'liberal' (in the European sense) party like the FDP (Germany), Lib Dems (UK), Liberals (Australia), or Fine Gael (Ireland).
I should add that I believe he is telling the truth and I agree with him.
OK. but "right-coded" and "committed to the right" are not the same thing. If for some reason someone starts a smear campaign tarring you as a proponent and practitioner of traditional Aztec religious practices, and gets most people in their ingroup to believe that, that doesn't make you "committed to human sacrifice"
I've seen Matt Yglesias and Freddie DeBoer quote Scott, so I'm not sure he's only quoted by the right....
I don’t see why this commits him to the right now. He is clearly not engaging in motivated reasoning, but trying to think straight about the question of whether differences found in national IQ are valid. If the left rejects him for arriving at some right-compatible conclusions, maybe the left isn’t worth identifying with. Do they only want people who engage in left-motivated reasoning? Then they are mind-controlling jackasses. (And if the right embraces him they are in for some rude shocks when they learn his views about some other issues. I think they will soon spit him out.)
There are certain things a leftist isn't allowed to say, and he's said them. I absolutely agree they only want people who engage in left-motivated reasoning; that's why a lot of people are here, after all. Intellectual honesty is rarer and rare. And I agree about the right; he'll be one of these guys claimed by neither side. Which is better for a writer, after all.
Still, I've been in a lot of settings where it is clear you are supposed to think a certain way, but if you say something out of line with that, while staying friendly, and maybe saying something like "I know you guys don't think about it that way," it kind of wakes some people up out of their trance. Or maybe the effect is just from you providing a demonstration that not following the party line doesn't immediately get you punched in the mouth.
Even if Scott has made himself less appealing to the left, he may also have given some committed lefties permission to speak more freely.
The most rational thing to do in the face of ASI is birth as many high IQ individuals as possible.
How do you get from A to Z? I'm not against having smart kids, but I don't see the connection.
Also, it's funny I never see this argument from women. I'm not a feminist but I can guess why.
To me, a critical thing absent in these discussions is that in the cutting edge of intelligence science, whether in AI or biology, these IQ things play no role. The entire area is apparently a dead field where decades of the study of "intelligence" has yielded basically zero actionable information relevant to anyone doing actual science of either making artificial intelligence, working on neuroscience, or, afaik, educational efforts aimed at human intelligence.
Think of stuff like PPP, global workspace, spaced repetition, even really dubious stuff like fmri studies on everything have helped out like 100x the rate of this field
Whether that's because they're all dumb-as-rocks racists who can't science their way out of a paper bag is I suppose mildly interesting but overall? The less time spent on basically anything related to this dead end of enquiry the better. As a tremendously useful thinker, Scott, I think maybe you've been nerd sniped by the controversial ness or something. It'd be more fruitful to explain why this dead field shows any promise whatsoever to uou of shedding the tiniest bit of useful information relevant to the understanding of intelligence at all, after decades of nothing at all.
Huh?
Cognitive testing is obviously massively important in the modern world.
Investigating the collective average IQ level of different demographic groups is not obviously massively important.
For being a trivial waste of time, a whole lot of angry people sure seem to want to stamp it out.
How come?
Angry people wanting to stamp something out means it can't also be a trivial waste of time? Obviously it depends where you're measuring importance and it seems like you're wanting to conflate two different measurements.
There are lots of trivial waste of time hobbies. Do angry people devote a lot of effort to try to get them banned?
Instead, it appears that angry people tend to be extremely angry over the idea that people of different racial ancestries often have different IQs.
Imagine this: Two overlapping circles. One contains "not obviously massively important" pursuits. The other contains "pursuits worth stamping out". Some pursuits are deemed to be in one or the other circles, some pursuits are deemed to be in the overlap.
If this seems paradoxical that's because, again, the notion of "importance" is being measured in two different ways. Building model train sets might be not very important. If building model train sets has become an addiction for someone such that they are neglecting taking care of themselves and their dependents, it could be worth stamping out. The model train addict could say "you can't tell me this hobby is not important because here you are putting so much energy into trying to stop me from doing it!" The confusion is because an activity's "importance" is being reduced to a single measurement, as if anything can only be important in a singular or general way. It's a simple semantic confusion.
Because it provides cover for and thus at least marginally increases the rate of people saying unkind things about black people, and they are very much in the habit of shouting down people who say unkind things about black people.
And because they think this line of inquiry has *literally no other value* than facilitating unkindness towards black people. I'm pretty sure they're wrong about that, but most takes on the matter that aren't done with Scott Alexander levels of rigor and charity do a poor job of arguing that point.
So, if the subject comes up, they're going to try to shout it down because they have one possibly small reason to do that and no reason to take any other position, and because while they might prefer to be discussing something they feel is more important, someone else raised this subject so that's what is being talked about right now.
You know full well why.
And nobody is buying the babe in the woods routine.
Despite any valid inquiries or insights into improved public policy that could arise from understanding the limits of environmental interventions, the primary reason many people interested in this subject engage with it and seek to mainstream it is that it offers what appears to be a rational basis for racist discrimination and stereotyping.
People excessively focused on group differences in trait distributions are not especially dedicated to combatting stereotypes or prejudice or treating individuals based on their personal merits; they want more justifications for their existing biases. They're engaging in motivated reasoning.
This is another attempt in a long line stretching back hundreds of years to rationalize oppressive social hierarchies delineated by race as being natural and rational. Before genes, it was calipers, etc.
Of course, others are driven by different interests or simply by curiosity about topics restricted by social taboos.
But please, let's not pretend that there isn't a significant number of racists out there eager to promote seemingly objective arguments to justify their racism and degrade the taboo against it.
Do you mind explaining what exactly was so crazy about the 'calipers' routine? Modern MRI studies have repeatedly turned up racial differences in brain structure, and Samuel Morton's cranial measurements were re-examined a few years back and found to be essentially accurate. Heck, police forensic teams have been using skull measurements to identify the racial background of murder victims for decades, it's about an 80-90% accurate technique.
What do you think the antidote to racial stereotyping is supposed to be? Is our only contact with other races supposed to be through the medium of marvel movies, and not through glancing at FBI crime statistics or, I don't know, having your eyes open when you walk around California?
The attempt to eliminate racial discrimination is like any other socialist shakedown racket- you can't actually achieve equality of opportunity without achieving equality of outcomes, and trying to achieve the latter is both impossible and insane, so the grift never ends.
If IQ explains <bad thing X>, and no-one talks about IQ, I guarantee some other people/things/institutions will be blamed for <bad thing X>.
If you *really* apply yourself, I think you can figure out who might get blamed for, e.g, poor outcomes in afro-american communities.
Do you think afro-american communities have not been sufficiently compensated for whatever historical wrongs they might or might not have suffered? A civil war, decades of affirmative action, and trillions of dollars sunk into welfare subsidies, business grants and housing projects aren't enough to get whites out of the doghouse?
If not, what is your target here? Afro-americans have had the highest material living standards of any black population on earth for literally centuries, so... what does 'good enough' look like, exactly?
If you make the totally unrealistic assumption of a world with zero racism and zero historical injustice, and still grant IQ diffs and HBD, what do you think their income levels, family structure and homicide rates would be, precisely?
Many studies, such as those of Gregory Clark (referenced in the following link) find that status is remarkably consistent and reverts quickly after exogenous shocks. Consistent with that, the causal impact of slavery on descendants of slaves disappeared by 1940 (https://www.cremieux.xyz/p/black-economic-progress-after-slavery).
For that reason, it's unsurprising that when controlling for IQ, the Black-White gap in socioeconomic mobility is usually eliminated or reduced to a pittance (https://www.cremieux.xyz/p/whats-in-a-black-name).
To be clear, I'm not suggesting that the only discrimination against African Americans was slavery, but rather, that just as the impacts of the most profound form of discrimination faded out within a few generations, the impacts of later lesser forms of discrimination faded out as well, likely even more quickly, as is typical for many groups all around the world, as discussed in the first link.
Each different bad outcome will have a different mix of causes. The performance gap in education probably has a very different set of causes than the black/white difference in rate of unwed birth, or the wealth gap, or the difference in voting patterns. There's not going to be some one-parameter model that explains everything, whether you label your one parameter "racism" or "IQ."
This could be more imaginative. <Bad thing X> might just be something we can't easily fix by direct intervention. For example, we're currently not very good at predicting earthquakes, but that doesn't mean someone has to be blamed for the earthquake problem. Similarly, maybe we're not good at fixing race disparities, and there's just nobody alive worth blaming. Sandra Day O'Connor thought things would equalize in ~25 years; predictions like this are based off of some very optimistic assumptions, and it might be less wasteful to devote time to that rather than IQ.
We blame earthquakes on plate tectonics, which is in fact an accurate description of how they occur, so I don't see how this analogy is useful.
Genetics/IQ is the actual explanation for most racial outcome disparities, and every other explanation is going to lead to faulty predictions and empty promises. I don't see the benefit to this aside from the comfort of cowardice.
In this analogy, the problem is our inability to predict earthquakes, not just earthquakes itself.
The benefit to this aside is that you can earn easier victories against leftist overreach with good arguments that don't alienate people. It's so facially absurd that the IQ obsessives think race realism has any persuasive weight in a world where (a) most people are viscerally revolted by it and (b) most race realists would, frankly, have been persuaded by much less evidence. With the way IQ frequently gets shoehorned into discussions where it's dubiously relevant - often I suspect to soothe the self-esteem of less impressive whites - it's clear that the scoreboard isn't looking good, and the project to normalize race science has been a pitiful failure. So, to circle back to my original aside: just be more imaginative.
My dude this crew has contributed zero to cognitive testing. They use these tests but as I said have no theoretical perspective on cognition and so make no contributions to them.
Also, I am not saying they should be stamped out, I'm saying they should be ignored as a dead field.
It's notable that the Sailer's, Hanannia's, and Murray's tend to be trained as political scientists, journalists, and even marketers rather than, say, statistical geneticists. I stumbled on a post where Hanania agreed with Murray about not being quite smart enough to understand linear algebra and was dumbfounded that ppl like these confidently pontificate about IQ, which is essentially a statistical and correlational measure. They also (scientific claims to the contrary) tend to be aggressive social activists with crude eugenic agendas.
No wonder they keep confidently repeating dumb things. Seriously: will you clowns stop assuming a person or group's "genetic IQ" can be determined from a linear combination of the IQs of their ancestors or groups?
There are a number of HBD advocates that clearly are well trained in statistical analysis (Kirkegaard and Cremieux, for example. Even David Reich is at least mildly red-pilled on the topic, even if he doesn't bang on about IQ diffs much in interviews.)
"Well trained" by what standard? Hanania believes he's well trained because he took some "methods" classes for a nonsense UChicago PHD but by reading his writings (including his own PHD thesis) his understanding of statistics is below the level of a good undergrad. Ditto Kierkegaard and Crimeux (his real identity is known). No idea who Reich is.
In short: a bunch of people too stupid to make real contributions to science, making completely obvious mistakes (JFC these idiots STILL confound hereditability and genetic influence), claiming that it is a mere conspiracy the overwhelming majority of experts disagree with them, and having Dunning Krueger geniuses like the OP for backup.
David Reich is the world's leading archaeogeneticist:
https://en.wikipedia.org/wiki/David_Reich_(geneticist)
You can dig up youtube talks where he mentions 'interracial genetic differences are only about 1/6th as large as the average differences between individuals, but 1/6th is not nothing' and I believe he co-published a recent paper documenting a ~2 SD increase in EA-PGS among europeans over the past 10,000 years. He knows what's going on.
I'm pretty sure Kierkegaard and Crimeux are above 'undergrad' level, or I think Scott or Charles Murray would have pointed out their elementary errors by now.
Emil is an idiot about a broad range of issues. I don't know Crimeux so perhaps. Murray is completely unqualified to have any opinions on this topic. So far Reich is the only credible person you've mentioned, and as you said yourself it's not clear he is part of the HBD sect.
Your standards of scholarship are basically no standards at all. No different than any other group of conspiracists whose worldview depends on elevating the few pseudo scienitific cranks that agree with them and ignoring the overwhelming evidence against.
I'm hearing a lot of ad hominem, and your credentialism is essentially a kind of circular argument. You can't trust the 'racists' because they're not credentialed, and they don't get credentialed because they're 'racists'. You know this perfectly well.
I mean, if this isn't good enough for you, John Rawls was pro-eugenics and both James Watson and Francis Crick are basically on-record agreeing that the B/W IQ gap is mostly genetic. It's hard to get better credentials on the topic of heredity than those two.
> as you said yourself it's not clear he is part of the HBD sect
No, he's definitely said that the egalitarian position is untenable, he's just being very... cautious about the magnitude of expectable racial differences in socially-consequential traits. (Pygmies and maasai differing in height by far more than 1/6th of a standard deviation does not bother him, apparently, but this is obvious to anyone who thinks about the topic for more than five minutes.)
What do you think are the prerequisites for entering a discussion on a topic of broad public interest like this?
Also, are you familiar with the term "isolated demand for rigor?"
Entering a discussion - None. Making scientific claims - The competence to at least understand the science. Trying to drive social outcomes - extremely high!
All of these "race scientists" make social and scientific claims about topics that are (despite their claims to the contrary) extensively and even exhaustively studied. You will note that Murray made predictions in "the bell curve" about the impact of dysgenics on American society. 30 years later he was wrong about basically everything.
"We believe X based on these 40 year old studies that constitute 1% of the field and think all of the experts in the field are simply wrong" is certainly a view you are entitled to have. You are not entitled to have people take your opinions seriously.
And yes - these people are trying to drive social outcomes. Scum like Hanania have argued for race based immigration policy (including mass deportation not based on legal status) and outright eugenics. I think I am correct for having a high burden of proof to take their ideas seriously
Eugenics is completely inevitable regardless of what you think about HBD, by the way. There is no genetically neutral human society and we can't ban deleterious mutations.
Natural selection is inevitable. Formal programs to try to improve the genepool are not inevitable--we don't have them, most societies haven't, and the ones that have existed mostly didn't have much effect.
Which ideas? You should be skeptical of race-based immigration and coercive eugenics policies because those are bad policies, but I don't think their badness turns on empirical questions about IQ distributions across racial groups. When Sailer, Hanania, or Murray make empirical claims, I think we should evaluate those on whether or not they are accurate, not on whether or not we think those guys are smart enough or have the right credentials to discuss the matter.
Tons of people talk about IQ, race, the performance gap, the causes of differences in outcomes between blacks, whites, hispanics, and Asians, and similar topics. Hardly any of them are experts in statistics, psychometrics, and genetics.
Do you also decry the ones who aren't experts in these areas, but who express confident opinions that the black/white gap in school performance is due to structural racism, prejudice among educators, or worse childhood environments for blacks? Because if you don't, it seems to me that you are engaging in an isolated demand for rigor--holding your side to a much lower standard than the other side.
>Which ideas? You should be skeptical of race-based immigration and coercive eugenics policies because those are bad policies, but I don't think their badness turns on empirical questions about IQ distributions across racial groups.
The people that support these policies use arguments about IQ distributions to justify them, so they certainly think their "goodness" rests on these questions. FWIW these ppl are not disinterested scientists either - they tend to be deeply committed bigots. I blocked Hanania for just being arbitrarily nasty to black ppl before I realized he had any scientific theory behind it.
>When Sailer, Hanania, or Murray make empirical claims, I think we should evaluate those on whether or not they are accurate, not on whether or not we think those guys are smart enough or have the right credentials
Their empirical claims are based on 50 year old science (twin studies) whose validity is effectively debunked. They claim that the debunking is one big PC conspiracy by the entire scientific establishment against them, the real truth knowers. It is therefore important to point out that their grasp of statistics is weak, since the scientific debate is essentially one about causal inference (statistics!). If you're interested Turkheimer's "Understanding the Nature-Nurture Debate" is a solid recent book.
> Do you also decry the ones who aren't experts in these areas, but who express confident opinions that the black/white gap in school performance is due to structural racism, prejudice among educators, or worse childhood environments for blacks
It depends on how confident their opinions are, what evidence they bring to bear, and the costs of their proposed remedies. Some of the claims are hand wavy, but we know that the U.S. government for centuries conspired to impoverish its black population, and that deprivation leads to worse educational outcomes. You can of course counter about the relative performance of blacks vs. whites/asians in the same SES. I suspect a lot of this is culture and can be teased out by within group comparisons. In any case, it is very likely multi-factorial and we can study and propose interventions.
My objection to the race scientists is that they claim massively overconfidently that the primary cause is genes and then propose solutions that happen to align with their personal racial hatred: keep blacks or w/e race I don't like out of this country, out of key professions, etc. They are also intensely arrogant - think of the God tier superiority complex it takes to propose eugenic solutions that treat other humans as mere rodents to be tinkered with. This is why I mock Hanania and Sailer - the arrogance is bizarre from people that aren't even top tier intelligences.
The SJW structural racism view at least involves solutions that are likely net positive even if you dispute the ROI: more support for early childhood education, nutrition, etc. When they propose solutions that are harmful (increase segregation, stop advanced classes for students that can test into them, etc) I object. The only solution the race science chuds ever propose is "do racism".
Also I don't know if asking you not to make third grader assumptions (A person's traits are a linear combination of traits from these poorly defined genetic groupings) is much of a demand for rigour.
It's a model. Like all models, it's a lie, but it can be useful for making good predictions.
Commentary: https://www.cremieux.xyz/p/national-iqs-are-valid
Very superb article. Much better and more comprehensive than the OG article, if it's actually a reply. Do you actually write all of this that fast?
Cremieux is fast.
Everyone interested in this discussion should immediately read the above link.
Enjoyable read!
Since you solicited input on typos and grammar:
"Compare Lynn’s estimated national IQs to if independent collections of national IQ estimates and see"
Has an extraneous "if."
"and despite that still generally not that impressive "
Could use a comma after "despite that."
Thanks, fixed.
Are there any high scoring sub-Saharan countries?
The World Bank's Harmonized Learning Outcomes database links test scores from around the world.
https://www.takimag.com/article/a-little-learning/
Poor countries only rarely take the big global tests like PISA and TIMSS, but there are regional international tests in Sub-Saharan Africa and Latin America. Scores on those tests can be converted to PISA-like scores using the example of when countries that take the regional tests took the big global ones. It's much like Sean Reardon's Stanford database of school achievement test scores from every school district in America that are harmonized versus how well their states did on the national NAEP.
The one sub-Saharan country that did really well in the World Bank's database was Kenya at 455 on a scale where 500 is average for a First World country.( Gabon also did well, but that's an oil country, so I don't know who sat the test - Exxon engineers' kids from Texas?).
Trustworthy? Kenya's scores bounced up and down a lot. I looked at the report on the highest scoring year which was like a northern European country. The report wasn't obviously faked. If it was fraudulent, they put a lot of effort into it. So, I dunno ... I don't think I believe the score, but I also think Kenya is one of the better countries in Africa.
> Maybe I should have had a stronger opinion on whether Lynn’s exact studies were correct?
Your title was "How To Stop Worrying And Learn Love Lynn's National IQ Estimates" which even with the degree of irony implied by the reference reads as saying they are substantively correct. Rather than your actual position which seems to be that they are unreliable, but it doesn't matter because of other data.
Replication is crucial in science. If Lynn's data didn't replicate, it would be worthless. But, despite Lynn's occasional sloppiness, it largely has been replicated multiple times since 2002, although, admittedly, Wicherts et al argue that African IQ scores are about 1.3 standard deviations below Europe rather than Lynn's estimate of 2.0.
Africa, obviously, is the hardest continent to get reliable data out of, so I'm not going to go out on a limb and argue for either Wicherts' or Lynn's African estimate. I'd lean toward Lynn's because I think nurture can matter a lot in IQ, whereas Wicherts' estimate that Africans score almost as well as African-Americans (whose nutrition and health care is good enough for them to be likely the world's best athletes -- Lamar Jackson, for example, isn't a stunted individual) suggests that variance in IQ is almost all genetic, a position that I, being a congenital moderate, am averse to admit.
Also, the low-hanging fruit in making your people smarter are probably nutrition and disease, and those are places where sub-Saharan African countries have a lot of available improvements. Raising the average IQ of the next generation of Togoans by 10 points is probably *way* more feasible than doing the same for the next generation of Germans.
Right. For instance, iodine supplementation of salt is a known way.
It's really interesting to see how some Americans see the world......
In Africa we like staying alive and in good health whatever your IQ as measured by whatever test. Obviously improving health outcomes is important - including improving nutrition. It's especially important for infants and children. Fortunately the literature and data on how stunted growth impacts overall health and education outcomes is solid and has been researched by credible scientists. So far there has been steady but slow progress. You'll be pleased to know that iodine supplementation was a subject of public health campaigns in countries like Zimbabwe more than three decades ago. You cannot sell salt which is not iodised. Despite high levels of extreme poverty, salt is affordable to everyone. I believe Malawi passed similar legistation more than 25 years ago.
Secondly, we want children to have access to schools and to stay in school and finish their secondary education. (There's another discussion to be had about improving the quality of education in those schools.) Again, there has been slow but steady progress on education-related metrics.
There is no government trying to improved 'average IQ', but they have been working to improve health and education outcomes which might incidentally improve Mission IQ to your delight. Thankfully we believe our children are not inherently inferior to children in other parts of the world. Given the right interventions, they can achieve their best outcomes. Not every one of them will make it to Med School or Eng/CS in our modest universities, but they can go for vocational training. If you find any groundbreaking, useful insights backed by decent data in IQ-Land, we'll consider it.
My impression is that there's a package of stuff that we know how to do to improve how well everyone does, in terms of physical health and brain development and all that. Those are interventions that do seem to raise average IQ in places that do them, and also lower infant mortality, raise birth weight, raise life expectancy, etc.
That's basically the stuff you're talking about--as your society gets richer, you can make sure everyone gets enough to eat and doesn't have any deficiencies due to (say) low iodine in the local soil, eradicate most of the endemic diseases so you don't have kids supporting a heavy parasite load while developing, get everyone their shots, make sure everyone has access to some kind of education, etc. This is the environmental version of "catch-up growth," where poorer countries can very quickly improve things because they can benefit from other countries' experiences.
In the early 1900s, there was still endemic hookworm in the US South, and Army recruits from those areas had lower average IQ scores than other regions. The US only really got on top of this about a century ago.
"And in fact, economists routinely find that African small-plot subsistence agriculture is actually highly optimized; farmers make very precise choices about where to plant which seeds, which fertilizer to use, etc."
This was totally untrue where I worked in Ghana. Economists struggled to work out why African farmers don't use improved seed varieties, techniques or chemical fertiliser when it would improve their outcomes, and why they used labour inefficiently. The only question is whether this is for mainly conventional economic reasons - individual farmers don't get to keep enough of their earnings, their husband spends the earnings on booze etc., or more "behavioural economics" reasons - farmers fall prey to biases, short time preference etc.
Intelligence didn't enter the official conversation, of course, but, as with every profession, it's obvious that smarter farmers do better.
My vague impression is that Midwest American farmers these day face cognitive challenges similar to, say, corporate vice-presidents. To be a good corn farmer these days, it helps to have a STEM bachelor's degree and an MBA-like degree in farm management. Farming has had gigantic productivity improvements, so American farms are now typically managed by people who are the best farmer of their siblings who are descended from the best farmer among their grandparents' generation. So, Midwestern owner-operators tend to be really good at what they do.
African subsistence farmers haven't been through that selection process to the same extent.
To attempt to offer a synthesis of Lyman's thesis that African farmers seem smart about their business and your antithesis that they don't learn well, I'd suggest that most people have intelligent things to say about their current jobs, but that IQ plays a role in whether they can easily learn new ideas. The American agriculture industry has come up with a huge number of new techniques. It would be great if African farmers could comprehend these breakthroughs well enough to apply them in their own fields, but that's asking a lot.
Speaking of African farmers learning new techniques:
"At the beginning of the seventeenth century, the Portuguese transported manioc from South America to West Africa for the first time. They did not, however, transport the age-old indigenous processing protocols or the underlying commitment to using those techniques. Because it is easy to plant and provides high yields in infertile or drought-prone areas, manioc spread rapidly across Africa and became a staple food for many populations. The processing techniques, however, were not readily or consistently regenerated. Even after hundreds of years, chronic cyanide poisoning remains a serious health problem in Africa. Detailed studies of local preparation techniques show that high levels of cyanide often remain and that many individuals carry low levels of cyanide in their blood or urine, which haven’t yet manifested in symptoms. In some places, there’s no processing at all, or sometimes the processing actually increases the cyanogenic content. On the positive side, some African groups have in fact culturally evolved effective processing techniques, but these techniques are spreading only slowly."
https://slatestarcodex.com/2019/06/04/book-review-the-secret-of-our-success/
> American farms are now typically managed by people who are the best farmer of their siblings who are descended from the best farmer among their grandparents' generation
Also, farms have massively consolidated, which largely involved the worse farmers being bought out by their better-at-farming neighbours
MM writes:
"It's hard for people from literate societies with tons of exposure to text & graphical representations to see the extent of the gap."
I think the growth in our exposure to text and graphical representations due to the ongoing Information Revolution going back to Gutenberg accounts for some of the Flynn Effect. Early IQ test designers in effect anticipated which way daily life was going to move, which is why raw IQ test scores went up all over the world over much of the 20th Century:
https://www.unz.com/isteve/the-flynn-effect-across-time-and-space/
Question for those who have looked at the meta-analyses more closely:
Do we get a sense of how much control there is across testing: facilitation, environments, etc?
Also, have there been any studies looking at how cultural values and, more interestingly, personal attitudes, might correlate with test scores -- i.e. a pre-test survey along the lines of "how do you feel about scoring well on this test? how much effort will you put into answering questions correctly?" etc.
> Africa is extremely genetically diverse, but I think most of the countries measured in the paper, including Malawi, are some variety of Niger-Congo speakers, who I don’t think are that much more diverse than white people or anyone else. The really interesting African ethnicities, like the Khoi-San, don’t show up as much at a national level.
Languages don't follow genes. Both Indians and Brits speak PIE-derived languages, and they're not super genetically related; inversely, Turks only have a 21.7% East Asian (incl. Central Asian) contribution, because most of their ancestors are Anatolian Greeks with the Mediterranean genes. In Africa, you can even see this visually - the Twa are a pygmy people who speak a Bantu language, so obviously their short genes didn't get swamped out by Bantu immigrants. The spread of Niger-Congo languages came with the spread of Niger-Congo culture, technology, etc, but that doesn't mean they overwhelmed of the native populations' genes. IIRC, in Cabinda they did, but mostly African populations' genes are substantially "ancestral," i.e. predating the Bantu expansion.
In the last comments thread it was explained why that is incorrect. Most people in SSA are descended from the Niger-Congo population who replaced (killed, out-bred, genocided) hunter-gatherers across most of west, east and south Africa. There is not much phenotypic variance within SSA populations or between them and their (17%) admixed American descendants.
No, they interbred with the hunter-gatherer populations as they went. There is significant admixture with locals among basically all Bantu-speaking populations, which can and has been identified. They did not engage in some kind of proto-anti-miscegenation law in the year 2000 BCE.
I’m not informed about the genetic and linguistic makeup of the Niger-Congo people, but in prehistory it’s relatively common for an invading people to almost completely genetically replace the preexisting people.
Off the top of my head there’s the Dorset culture (which has literally no genetic presence in the modern Inuit), the Mesolithic European population replaced by the Anatolian farming peoples (which has very little genetic presence) and the Jomon of Japan (who only represent ~10% of the modern Japanese genome depending on the area).
I don’t know about the Niger-Congo expansion, and how much interbreeding actually happened (I’m sure it varied dramatically), but I wouldn’t be at all surprised if there was quite a bit of genetic similarity between those people, as you typically don’t spread a language without spreading at least some of your genes in ancient history.
"GDP only depends on the real IQ", hmmm perhaps reread Henrich:) as not all key traits leading to successful societies like obedience, conformism, cooperation, etc, are correlated to intelligence (some even against it:). Dr Jan van de Beek did interesting research into which immigrants in the Netherlands had a net positive contribution.to its society. The cultural indicator proved to be defining. >
https://x.com/demo_demo_nl?ref_src=twsrc%5Egoogle%7Ctwcamp%5Eserp%7Ctwgr%5Eauthor
1. I wonder whether there might be a "spare capacity effect"? If you're using all your wits to survive your brain has rewired to do that and you don't have much spare capacity to handle new problems.
2. in farming (subsistence or otherwise) you're used to thinking long term, you have a small number of very important decisions with plenty of time to think about them. That doesn't lead to a brain that will do well in IQ tests.
3. What psychologists used to call rigidity (maybe they still do). If your culture is very conservative you're probably programmed to resist new ideas and have never learnt to solve problems creatively.
> but that sometimes you can get IQ through other means (like an extreme pro-education push by a poor country), and this doesn’t itself cause development
But why wouldn’t it? If levels of education improve cognitive ability is that what matters, even if the general IQ stays the same. Ability at intellectual tasks is the ultimate prize, regardless if the cause is good IQ hampered by bad education, or bad IQ improved by good education.
Keep seeing mentions of an article(s) in "Aporia".
For those of you unfortunate enough not to be European entomologists, I should point out that Aporia is the scientific name of the butterfly, Black-veined White.
I'm fascinated by the global iodine map, I thought that surely the results for Ireland had to be wrong - we're an island! Fish on Fridays was the norm in the hey-day of Catholicism! Seaweed was used as fertiliser!
But no, we're mildly deficient:
https://www.medicalindependent.ie/in-the-news/conference/mckenna-lecture-addresses-need-to-fortify-iodine-in-irish-food-products/
"Dr Mullan said this is important for “our women and our babies” as iodine deficiency is a contributor to low IQ. A 2013 analysis of results from the Avon Longitudinal Study of Parents and Children showed a link between the low IQ of children and low iodine levels in their mothers’ first trimester, she said in her lecture.
High levels of iodine are often linked to a good intake of milk, yoghurt, and white fish. However, changes in farming practices, the dissolution of the milk marketing board in 2002, and a rise in “alternative” plant-based milk add to the decrease in iodine levels, according to Dr Mullan.
Cow’s milk intake has declined in women and girls, and “young women don’t drink any more than a quarter of a pint per day”. Plant-based milk alternatives are much lower in iodine than conventional and organic cow milk. There is also low availability of iodised salt in supermarkets on the island of Ireland, according to a 2019 study from the European Journal of Clinical Nutrition, which Dr Mullan was involved in."
And this really intrigues me: breathing in air from seaweed increases your iodine intake. Clearly, I was smarter when I lived right beside the sea as a child 😀
https://imj.ie/2743-2/
"Results for UI from both schoolchildren and adults in the present study confirm that living in a coastal environment does not automatically ensure adequate iodine intake, as the apparent iodine deficient status of the Irish population might not have been predicted on an island where few live more than 200Km from the sea. Iodine deficiency has been reported in many coastal areas while recent studies have shown that coastal communities on the Atlantic islands of Madeira and the Azores have median UI values lower than on the Portuguese mainland. These islands support very little seaweed growth (Limbert, Private Communication). Thus the conventional wisdom that iodine intake in coastal regions exceeds that in inland areas needs to be re-evaluated. What appears to make a significant contribution to the coastal iodine environment from this study is the presence of abundant seaweed growth as evidenced by the higher median UI values observed in both schoolchildren and adults living in coastal areas close to abundant seaweed growth (hotspots). The most significant manifestation of this was the % of individual UI values >150µg/L which was 45.6% in schoolchildren and 43% in adults living near a seaweed hotspot compared with lower prevalences (2.3-16%) in subjects living in inland or low seaweed abundant coastal areas. While factors other than living near seaweed hotspots may explain the differences in UI observed, the findings suggest or at least do not exclude, the possibility of seaweed derived gaseous iodine inspiration through respiration making a contribution. Consumption of iodine rich foodstuffs might have applied when food consumed was almost exclusively locally produced but is less likely in the era of supermarkets with widely sourced products. Thus the findings in this study do not support the thesis that living in close proximity to the sea is in itself a guarantee of adequate iodine intake. This may require a seaweed rich environment, allowing for intake by respiration of seaweed derived gaseous iodine. Possible confounders to the results include the failure to obtain detailed dietary information and data on possible exposure to seaweed derived gaseous iodine on all subjects and findings are therefore dependent on the geographical location of sample collection."
So bring your kids down to the beach where there is abundant seaweed and let them play amongst it for the benefits of gaseous iodine!
While I do accept that there is a thing we call "intelligence", which we can measure, and stick fancy labels on (whatever 'g' turns out to be), I am coming more and more to the opinion that IQ tests are measuring a plethora of things, which get confused.
Yes, one thing they're measuring is something we can roughly call intelligence (though I suspect it's more 'mathematical capability' and we can have a whole other fight about "does 'being smart' mean 'good at maths' only? and what does that mean for times in history when we didn't have or need the mathematical talents for, say, computer programming which are what is now deemed a measure of being smart, successful, and rich? What about the coming of AI, which will take over those mathematical tasks for us and do them even better than any human could, will humans then be deemed 'stupid' even with the same IQ results?")
But I think another thing it is measuring is "how good are you at taking tests? how practiced are you?"
One comment on the original post talked about the Raven's Matrices style tests and that the solution to one question was " Like I recall there being a problem that ended up being a geometric version of a square root."
Now, my reaction to that was "You can *do* that?" (and hard on the heels of that, "*Why* would you do that?") but mainly, you could sit me down in front of that question from now to Doomsday and (speaking as a white chimpanzee, thank you Charles Kingsley!) it would never, ever occur to me that the collection of lines and black and white circles was representing a square root. And I've *heard* of square roots.
What if you're living in, yes, Malawi and you never heard of a square root in your life? Maybe the really smart can figure out the pattern is "if this blob is X, and that blob is Y, then the next blob should be Z", but if you're having to invent the entire concept of square roots from scratch, I don't think you'll do well on a timed test.
Conversely, if you've trained on these tests and can recognise the shortcut that "oh yeah this one is about square roots", you're going to score better.
I wonder if most of us would score well on a test asking us, for example, "which of these pictures of cereal crops is sorghum?" but a Malawian subsistence farmer would ace a test about crops and planting times. Who then would be the fools and who the smart?
https://en.wikipedia.org/wiki/Agriculture_in_Malawi
"The main economic products of Malawi are tobacco, tea, cotton, groundnuts, sugar and coffee. These have been among the main cash crops for the last century, but tobacco has become increasingly predominant in the last quarter-century, with a production in 2011 of 175,000 tonnes. Over the last century, tea and groundnuts have increased in relative importance while cotton has decreased. The main food crops are maize, cassava, sweet potatoes, sorghum, bananas, rice, and Irish potatoes and cattle, sheep and goats are raised. The main industries deal with agricultural processing of tobacco, tea and sugar and timber products. The industrial production growth rate is estimated at 10%."
It would seem that Malawi also lacks mineral resources, the soils are depleted, and rainfall is unreliable, with a period of good rainfall (and thus adequate and even abundant crops) later replaced by droughts (and hence shortages and famines). I think all of this has to have an effect on the population, and we need to clear away the effects of starvation, malnutrition, poverty, lack of schooling, etc. before we can even get to testing "are Malawians really all dumb?"
" "Does 'being smart' mean 'good at maths' only? "
The answer would probably be that math abilities correlate reasonably well with language and reading skills, and even musical aptitude (the latter bothers me most). However, the correlation comes apart in the ends as it should, so when you look at university professors, the linguists are not incredibly good at math or vice versa.
Although, for some reason, kids with better social skills had lower math skills. This contradicts the otherwise nicely correlating set of all cognitive abilities. https://www.nature.com/articles/srep23011
I am also sceptical about GDP, given that Ireland bumped up its GDP so surely that means we're all now smart and successful? Yeah, well, there's the GDP as recorded and the real GDP, which is the one where you strip out all the multinationals and get what we're actually producing.
An easy way to boost GDP is to become a tax haven for American FAANGs. That has an amazing effect on the perception that everyone in the country is now rich 😁
https://www.ntma.ie/business-areas/funding-and-debt-management/investor-relations/irish-economy
"The last few years have highlighted the unique structure of the Irish economy. Ireland has a standard domestic retail/services economy which is augmented by a highly productive and fast-growing multinational sector. The two-speed nature of the Irish economy means that volatility can be expected in the usual measures of economic activity.
Following a growth rate of 8.6% in 2022, GDP fell by 5.5% in 2023. In Budget 2025, the Department of Finance forecast GDP to fall by -0.2% in 2024.
However, given that Irish GDP is distorted by the presence of multinationals, Modified Domestic Demand (MDD) is a more accurate measure of underlying domestic growth. Final MDD grew by 2.6% in 2023 and is predicted to grow by 2.5% in 2024. In Q4 of 2024, MDD increased by 4.1% compared to the same period last year. This growth in the domestic economy is largely driven by positive trends in the labour market and in personal consumption."
>"Yeah, well, there's the GDP as recorded and the real GDP, which is the one where you strip out all the multinationals and get what we're actually producing."
The latter figure is GNP (gross national product), what's produced by the Irish; GDP (gross domestic product) is what's produced in Ireland. There was a switch over time in which one economists focused on because it's much easier to measure GDP than GNP.
You're spot on that FAANG causes the two to diverge far more in Ireland than in most (all?) other countries.
Real GDP is what you get by subtracting inflation from nominal GDP (only NGDP is actually measurable, and there are multiple ways to construct an estimate of inflation). Real vs. nominal GNP would have a similar relationship, but harder to calculate due to multiple currencies and their distinct inflation rates.
What if IQ has three components: a genetic component, a nutrition/disease/parasites component, and an education component which is less "real" than the other two? Then you couldn't boost GDP with education if there's still a high disease burden making everything worse.
I think you're onto something, nature/nurture seems to conflate too much; it might help to refine things even further; e.g.,
g = f(genetics, nutrition, disease, parasites, …)
IQ = f(g, education, test bias, …)
"Key point is basically: it really isn't true that an IQ 60 person can run a farm functionally."
I would be hesitant to put much stock in this, because I don't know exactly how many smart people you need to run a farm but I'm sure it's not a lot. As an analogy, consider a car: you need reasonably smart people designing the car, but assembling the car can be done with people who are much less smart. Likewise, you need some (maybe just one per farm) smart people to organize the farming, but much of the work can be done without much intelligence.
Scott's responses to many of the comments about point 2 ("but they don't appear disabled...") seem self-contradictory to me. On the one hand, he wants to say that we're badly calibrated to what a low IQ without other disabilities should "appear" like. On this view, Malawians "really" have a low IQ, and the estimate of say, 60, is valid. On the other hand, he also says a few times that the notion of IQ itself falls apart in these cultural contexts. On that view, Malawians don't "really" have a measurable IQ at all, and estimates of 60 are not valid. Both of these can't be right.
I have lived and worked for decades in Africa including Malawi. Malawian farmers are clearly and obviously not retarded or stupid. They are on the whole, lovely, engaging, kind, religious and hard-working people and manage well in a subsistence agriculture world. At the same time they are much less efficient and effective farmers than south-east Asian or mid-west farmers.
And they are not very good at abstract thinking, pattern recognition, strategy, foresight, mathematics. They are not winning Field's prizes. They generally have a great sense of humour and can be very warm and kind. They can also be superstitious, quick to anger, and easily fooled by bad actors.
They are diverse and the population includes some brilliant, nerdy, introverted, eccentric people. But averages mean something and this does not describe the average Malawian.
I support agricultural, medical and educational initiatives in Malawi and elsewhere and encourage others to do so too. At the same time, I don't think Africa or the world will really benefit from millions or billions more Malawians who are a net cost to every high income country they migrate to. And inside Malawi there are very strong patriarchal and social norms which keep society peaceable and friendly (although inequitable and elitist) and when Malawians leave these norms break down which can be bad for them and bad for their hosts.
"At the same time they are much less efficient and effective farmers than south-east Asian or mid-west farmers."
I have commented before about how, in dealing with my 135+IQ wife vs ~115ish me, one of the fundamental tension is her attempt to force various systems of hers on me to produce "efficiency" for me (which are not efficient due to baseline differences in intellectual capacity). Indeed, I think I've read that love of efficiency and desire to promulgate that around may be the hallmark of high-IQ people but not sure where that comes from.
So I could actually believe in an average IQ of low 60s manifesting itself not in people being "obviously stupid or retarded" but in them just being simply "much less efficient and effective" in a similar task.
60 is an outlier, an estimate for one country in one study.
It would be more helpful to think about 70 as an average. There are neighborhoods in the United States where the local average isn't all that much above 70.
Is there data on this? What neighborhoods would you guess this to be true for?
http://lagriffedulion.f2s.com/city.htm (from 25 years ago) finds an IQ of Blacks in (majority Black) Baltimore City of 76. Given variation in neighborhoods, probably some are lower than that and some are higher.
Wow! I guess this is caused by migration in/out, right? Basically a mix of "boiling off" (people with alternatives leave, and smarter people have more alternatives) and "condensation" (people with few alternatives end up there and stay)?
That's right, the article discusses it. In the case Baltimore, with its shrinking population, it's probably much more a matter of selective emigration than selective immigration.
What are the implications if IQ does vary and it is mainly genetic and it correlates strongly with ethnicity?
Discussions about this topic always end up arguing about whether this is true and why. Here, we find an argument that IQ differences by country are true. The racists want to say that it's mainly genes, the liberals want to say that it's mainly environment. Even if it's 50/50 that's a big role for genes. What if it's true?
I think this is the most important problem that society needs to come to terms with. Even though the 'racist' view seems to be mostly advocated by conservatives, it's not actually clear to me that the implications lead to solutions conservatives like.
If people have low-IQ mainly due to bad genes, and you can't educate them out of that, doesn't it tend to support some kind of Rawlsian leftist view of society? We need some kind of a welfare state system for them, because those people will never be able to be competitive in society. They need some alternative more lenient legal system, because they are incapable of fully appreciating the consequences of their actions to the same extent high-IQ people are. We could end up with a lot of Woke policies.
Actually, I think this is just the first stage of a broader problem. Society has gotten so complex, technologically and socially, that the low-end of the IQ distribution is incapable of navigating it. We're seeing the consequences most clearly in the groups most affected. It's a lot easier to see that as a racial issue than as an IQ issue, but it's more correlation than causation.
You might enjoy reading The Genetic Lottery. It’s a leftwing take on what to do about IQ being genetic
What's the tldr?
I feel like you've just completely glossed over the other solution, which is eugenics. Though, it might not actually be a good idea to select for high intelligence across the entire population. Optimally you want a caste system where the laborer castes are selected for moderate intelligence and high obedience/low agency, and the learned castes are selected for high intelligence.
This seems to stretch the word 'optimally' beyond all possible reason for most people, for whom 'optimally' and 'caste system' are basically not compatible.
You will also note that claims to simply "be curious" to the contrary, the "race scientists" are social activists. As I noted in another thread this is a reason to demand rigour.
You're not just spouting off about which sports stars are best or which entertainers are the most attractive here.
There's a really important thing to remember when you're talking about how much of IQ differences are genetic vs environmental: it's not constant across changes in environment. In environments where substantial numbers of people are malnourished or have hookworm or something, the heritability of IQ will be lower than in environments where basically nobody is malnourished or parasite-ridden.
dIQ/dEnv is a function of environment, not a constant!
I'm not an expert in any of this. But presumably "intelligence" (or whatever IQ measures) boils down to doing a better job of taking in sensory inputs, processing them, and producing outputs, where better means doing things faster, with more complex inputs, processing, and outputs. This implies that the sense organs and brains of people with high IQ differ from those of people with lower IQ. Has this ever been observed experimentally? If so, how do those sense organs/brains differ? More neurons? More connections between neurons? Faster neurons/synapses? Any/all of the above concentrated in certain functional parts of the brain? Something else entirely?
The obvious one is larger brains. There is a clear association between brain size and intelligence. The current best guess for the correlation between whole brain size and intelligence (g) in adults is about r = .30
The obvious question, then, is whether or not groups with higher average IQ have larger average brain sizes?
Generally that's probably the case:
"Regarding the size of differences, the Black-White, Hispanic-White, and Other-White whole brain volume differences are, respectively, 0.99d, 0.54d, and 0.36d. "
https://www.emilkirkegaard.com/p/modern-neuroscience-confirms-race
This might be a clue why our average IQs are presumably low in Africa. We are not politically correct as kids, so we just nicknamed people with big heads 'Big Head'. I suspect most kids made sure their heads remained small.
"The best fit between brain traits and degrees of intelligence among mammals is reached by a combination of the number of cortical neurons, neuron packing density, interneuronal distance and axonal conduction velocity—factors that determine general information processing capacity (IPC), as reflected by general intelligence. The highest IPC is found in humans, followed by the great apes, Old World and New World monkeys. The IPC of cetaceans and elephants is much lower because of a thin cortex, low neuron packing density and low axonal conduction velocity. By contrast, corvid and psittacid birds have very small and densely packed pallial neurons and relatively many neurons, which, despite very small brain volumes, might explain their high intelligence."
https://pmc.ncbi.nlm.nih.gov/articles/PMC4685590/
Does that hold when comparing different individuals within the same species?
Biology is not my area of expertise, I'd just remembered hearing about the research at the time.
That said, I see no a priori reason it *wouldn't* hold intraspecies, although the ranges of neurological parameter values & individual intelligences would be much narrower than interspecies, making the relationship noisier (id est, sensitive to measurement errors and/or species-specific factors).
The evidence-based explanation for why petrostates don't convert GDP-per-capita to IQ at the same rate as their GDP peers is that their kleptocratic governments don't convert GDP per capita to years of schooling at the same rate, not that their people are genetically less capable of converting years of schooling into IQ. Qatar, Kuwait, UAE, Bahrain, Brunei form a very obvious cluster below the curve here: https://ourworldindata.org/grapher/average-years-of-schooling-vs-gdp-per-capita
The selective collapse of intellectual standards on this is really a sight to behold.
And yet educational attainment scores in petrostates show that kids in the exact same classes in the exact same schools get different scores according to their ethnic background. By some strange, and never properly explained coincidence, east Asian kids score highest, followed by caucasians, with arabs far behind.
In the last commentsn thread it was explained why that is incorrect. Most people in SSA are descended from the Niger-Congo population who replaced (killed, out-bred, genocided) hunter-gatherers across most of west, east and south Africa. There is not much phenotypic variance within SSA populations or between them and their (17%) admixed American descendants.
Obviously, there's much left to unpack about what IQ tests measure and how they do it.
Two thoughts -
1) The comment about "IQ of a 3rd grader" statement is in complete ignorance of how IQ is measured. '3rd graders on average in comparison to adults', gets closer, but here too the context is missing - US 3rd graders? compared to US adults? There are proportionally equivalent 3rd graders with IQs on an equal spectrum as there are adults IQs on a spectrum. IQ isn't a measure of absolute, it's a measure of relative. (another aside - don't assess the 3rd graders you personally know as being representative).
2) the question of IQ and development level... Perhaps IQ is more related to the complexity of society as a whole, and development level then is a consequence of the complexity of society. Thus, 5-year-old IQ assessments still remain valid despite 5-year-olds having perhaps less exposure to development, but with substantial exposure to complexity.
"Malawians are 60-80% subsistence farmers."
Off topic somewhat, but I find this shocking. What a way to live in the 21st century. This could be the subject of many long discussions.
"One possible answer is that the causal pathway is high GDP → lots of education → lots of practice with abstract reasoning → high abstract/symbolic IQ. I don’t think this can be the whole story, because some countries that “cheated” to get high GDP (eg oil sheikhdoms) can’t translate it into IQ points at the same rate as everyone else."
It's probably not that they can't, but that they won't. If your high GDP is largely a matter of resource extraction, then you have very little incentive to create a robust educational system to prepare the citizenry for manipulation of complex symbols and performance of abstract reasoning. You just need them to know what a mine cart is. (Unless, like Qatar, you prefer to handle that part via "guest workers" who can never become permanent residents.) The map and table here https://data.worldbank.org/indicator/NY.GDP.TOTL.RT.ZS?view=map shows what countries have most of their GDP via resource extraction.
There's this video I've always found really interesting called "How I kept a job at McDonald's with an IQ of 70" where the guy talks about his life experience with surprising reflectiveness. I'd always assumed it must be fake or his IQ test was bungled or something, but now I'm wondering if it really is legitimate. Even after the examples given of how people with 70-ish IQ can seem normal though, some of the things he says in the video seem a bit too abstract for me to buy it.
Here's a link: https://youtu.be/fjDXvXACIEA
My three favourite quotes about IQ and race, blank slateism, gender, feminism, culture wars etc etc.
"MAGNA EST VERITAS ET PRAEVALEBIT "
Coventry Patmore:
For want of me the world's course will not fail:
When all its work is done, the lie shall rot;
The truth is great, and shall prevail,
When none cares whether it prevail or not.
And best of all Stevie Smith:
"Agreeing with that Latin writer, Great is Truth and will prevail in a bit"
Scott: You still seem to be avoiding the issue that Lynn used shitty IQ data for many of his countries, and he "estimated" the data for other countries. I don't understand why you take Lynn's national IQ dataset seriously.
This exposes a more general problem with Bayesian analysis. Unless the researcher performing the Bayesian meta-analysis takes a careful look at the original datasets (and frequently, they're unavailable), it could be garbage in equals garbage out.
Did you read Cremieux very good post dealing with this? He quite clearly shows that Lynn's data was not perfect but good enough. Remember it went through several iterations and while the earlier did depend on extrapolations and interpolations, later versions had more and more good studies to back them up.
And also, all the IQ data is aligned with other data such as the educational scores, the attainment scores, the number of scientific papers per capita, the patents, innovations, prize winners, historically great figures etc etc.
It is all completely consistent. Ashkenazi and East Asians best, followed by Caucasians, a long way behind Indians and Arabs, far behind Africans, at the absolute last Australian aborigines, pygmies, negritos. We are not making this up. This has been shown over and over again by hundreds and thousands of studies. Ignoring this really does show some amazing talents in self-deception.
Cremieux has his axes to grind. He loves to quote studies that agree with his biases and ignores those that don't.
Does anyone know what his academic background is? I suspect he's an example of one of those hobbyist people that Sasha Gusev described in his Abhorrent Science article — "who, with no credible academic affiliations, typically publish in journals that they themselves have founded and review. The underlying studies, to the extent they are conducted, are low quality and highly repetitive because the primary goal is to generate memes for political blogs and forums."
https://theinfinitesimal.substack.com/p/on-abhorrent-science-and-the-weaponization
I suspect he is someone who scores very high on openness, very high on conscientiousness, and very high on disagreeability. It is extremely hard to break group consensus and think against the herd. He is obviously intelligent and organised.
Probably a hobbyist. There are more than enough degrees of freedom to come to preordained conclusions on a topic like this. Causal inference is 99% a shitshow and no fancy structural model will save you. Professional x-metricians know this and generally don’t make confident assertions like he does.
Knock off the ad hominem bs.
The data isn't perfect but we work with what we have. The answers are still informative even if they have large implied error bars. What axe do YOU have to grind that you're upset that anyone is even talking about it? If you have a problem with his methods then contradict him with evidence.
One random thing that stuck out to me, mostly unrelated to the broader point.. the iodine deficiencies in Cuba that they seem to have not covered up for some reason is harrowing.. it makes me wonder just how bad all their other metrics actually are. I’ve been there and it was just awful.
I’d just like to see to see some of this research repeated by somebody who wasn’t a scientific racist OR one of his scientific racist colleagues. We wouldn’t spend so much time talking about an anti-vaxxer’s report on vaccine-related harms.
What do you mean by racist? Do you mean someone who thinks that populations who were separated for thousands and tens of thousands of years and were subject to different evolutionary pressures evolved different characteristics and that evolution did not stop at the neck but also influenced the brain where 25% of human genes are expressed?
Or do you mean someone who has an irrational hate of those who come from a different ethnic background and does not want to treat them fairly or decently no matter how intelligent, kind, friendly, useful and willing to integrate they are?
Or do you mean someone who notices patterns?
If you help me out with your understanding of racist I can try and explain where I am coming from. It is not a place of irrational hate.
He’s probably referring to this stuff (from the article):
He wasn't someone who had "personal racist opinions" that he kept as "personal racist opinions". He was the editor-in-chief of Mankind Quarterly, a white supremacist journal that was founded by people like:
Henry Garrett an American psychologist who testified in favor of segregated schools during Brown versus Board of Education, Corrado Gini who was president of the Italian genetics and eugenics Society in fascist Italy, and Otmar Freiherr von Verschuer who was director of the Kaiser Wilhelm Institute of anthropology human heredity and eugenics in Nazi Germany. He was a member of the Nazi Party and the mentor of Josef Mengele, the physician at the Auschwitz concentration camp infamous for performing human experimentation on the prisoners during World War 2. Mengele provided for Verschuer with human remains from Auschwitz to use in his research into eugenics.
It's funded by the pioneer fund, an organization he was a board member of and that has been classified as a white supremacist hate group, with one of its first projects being to fund the distribution in US churches and schools of "Erbkrank", a Nazi propaganda film about eugenics.
He's not just called racist, he *is* racist, he even describes *himself* as a racist.
Yes, it really depends on whether he meant racist in the terms of "I would like to round up gypsies and jews and work them to the point of death and then gas and incinerate them" or if he meant racist in terms of "I believe in evolution". The term racist could apply to either. My understanding is Richard Lynn was a racist of the second type.
The paragraphs I quoted are about Richard Lynn though. Personally I agree with Scott that Lynn being a total racist doesn’t mean we can’t still use his estimates, but yes he was very racist.. in the first sense of the word you mentioned.
I have listened to him and read some but not all of his work and I disagree. I think he would be totally horrified at the round them up and kill them racism, but would probably have bought into the HG Wells, Galton, progressive racism that was the default for the progressive liberals in the 1930s and a bit before and after.
So you're not really a racist unless you're also willing to commit mass murder?
Quite a take.
Sorry, I don’t know who you are and I’m not trying to insult you. I don’t think arguing about the definition of racism is a good use of anyone’s time.
>”We wouldn’t spend so much time talking about an anti-vaxxer’s report on vaccine-related harms.”
Actually, we would!
https://www.astralcodexten.com/p/failure-to-replicate-anti-vaccine
I think this is a good lesson. If someone is making insanely false statistical statements about the world, it shouldn’t be *that* hard to correct them simply by gathering the correct statistics yourself.
Several species of ants are capable of subsistence farming. I'm unconvinced success at subsistence farming is highly correlated with IQ.
To be really high decoupling here, I don't understand why people say "how can a society with an average IQ of 70 even survive?" when we see troops of chimpanzees and bonobos surviving quite well. Of course they aren't farming but they do use tools and maintain territory and form coalitions and have politics and take care of their young and have friendships and do lots of other quite complex things. Inventing farming is hard, but farming itself in a subsistence world is not a lot harder than maintaining status in a troop of primates.
Though, their ability to farm was coded into them through natural selection over an extremely long period of time. As far as I can tell, human farming is a purely adaptive behavior. They don't have any hardcoded methods to work with.
Humans are very good at imitation. Human subsistence farming is mostly just imitation of previous generations. Not much abstract thought or math involved. The imitation drive is very deep in the homo lineage.
Epistemic status: I'm an interested amateur, not an expert. I've thought about this a bit but could definitely be wrong.
Compare IQ scores with implicit bias scores. In both cases, there are very smart people with a nice coherent underlying theory doing experiments and measures and coming up with test scores that are pretty consistent across time for the same person. The only difference is that IQ scores meaningfully predict outcomes at school, at work, and in life, whereas implicit bias scores don't seem to meaningfully predict anything.
We know IQ scores predict things because of observations on people in first-world countries, overwhelmingly raised under first-world conditions. Just about nobody in the populations used for those studies were ever malnourished, they basically all got their shots and went to school, they overwhelmingly grew up in 20th century first-world conditions and so had stuff like TVs and music and books and such. The studies that have found the correlation between IQ and outcomes we care about are the only reason why we should care more about IQ scores than implicit bias scores.
As far as I know, we don't have studies telling us about relationships between IQ score and school, work, or life outcomes in Mali or Liberia or Kenya or wherever else. My guess (and it's just a guess) is that we would probably not see the same relationship between IQ score and those outcomes for the populations we're talking about. At a guess, that is probably a mix of modern industrial society requiring more abstract thinking, and also encouraging more abstract thinking by presenting you with situations that need it every day of your life.
If that is true, it's consistent with what we see. When you give someone in a first world industrial democracy an IQ test, you get something that's useful for predicting outcomes, and so seeing someone with an IQ of 70 implies some very bad things for that person's future. You can look at that and say that, assuming the score is correct, he's probably going to have a very hard time graduating high school, and he's not going to be able to do things as mentally demanding as, say, reading a five-page set of instructions for how to mix the cleaning fluid properly to get the floor clean without breaking the floor waxing machine. Perhaps an IQ score of 70 in some kid who grew up in a mud hut in Tonga and never saw the inside of a schoolhouse just doesn't mean the same thing.
How would we tell whether this is right?
I have lived and worked in Africa and other developing regions for decades. Many of the commentators seem to have no understanding of contemporary Africa. It is not a country of people in leopard-skin loin cloths with bones through their noses. Most of modern Africa has gone to primary school and is connected to radio, tv, smartphone and/or internet and has been for a quarter of a century.
Only a tiny minority of Africans have never seen a schoolhouse. An African with an IQ of 70 means someone who is probably very pleasant, talkative, interested in you, speaks a couple of languages (wants to sleep with you if you are a girl). But they can't do hard math, they haven't read good literature, they have no interest in abstract discussions, they are annoyed and bored over intellectual pursuits, they have a deep distrust of everyone, think people into shrimp welfare are so crazy as not even to be human.
Many Africans have an IQ of 70. They are often really nice, kind, hard-working, religious people. It really seems to be genetic and no evidence it isn't. We can be quite sure it is not famine or cultural deprivation because of the Dutch hunger, adoption studies etc.
Why do they have a deep distrust of everyone? I'm not convinced that's going to be IQ related.
...Looking at this comments section, I think their distrust is perfectly justified.
Based.
Trust is a funny thing. Some societies which seem a bit alien to me like Swedes and Germans and mid-west have a lot of trust. I suspect they are like cheetahs and went through some sort of horrible genetic bottleneck which means they can all donate organs to eachother.
I know that in low IQ societies there is very little trust and for very good reasons. That was my "lived experience" but I am happy to see if you google studies, people with more poverty and more diversity have less trust.
This is so obvious to me I wonder why you don't grok it? Do you think with people being nasty and brutish and short it would be fun to hang out with them? Take my word, it is not fun.
But you just characterized them as often "really nice, kind, hard-working, religious people". And if they live in smaller communities, why would they have a lot of diversity? Poverty, okay. But now you've switched from "they're really nice" to "they're brutal, impoverished scum."
That is very unkind of you to call them brutal impoverished scum. I have lived in Africa for decades and that is so far from the truth.
You said that. I didn't.
There are a bunch of measures of trust (how people play in various experimental economics games, for example) that differ across societies. I think you should expect that culture/upbringing/experiences in your society will drive this, independent of IQ. ISTR that Russians come out very low on trust, for example. Russia isn't a low-IQ population, but there's a culture and history and current environment that would plausibly push you toward skepticism and suspicion.
"Many Africans have an IQ of 70"? You got this from talking to them for decades in developing regions?
Of late one of the biggest problems is young people who are educated but cannot find jobs because the economy is not big enough to accomodate them. Grade inflation has risen dramatically with access to e-learning and practice tests. So it's now competitive to get into the colleges. It's fair to say 'high IQ' is no longer translating to the good outcomes. Thinking that 'high IQ' would translates to higher national GDP is ridiculous and far-fetched.
All the evidence suggests that IQ predicts good things in the modern world/ west/ global north/ high income countries (e.g. educational attainment, marital stability, health, income, longevity, non-criminality etc) just as well for Africans and Asians living in modern Africa and Asia as IQ clearly predicts good outcomes for people born in high income countries. IQ is the single best single predictor of all of these nice outcomes. Yes, better than parental income or SES.
There is some evidence that people with "smarts" are also better hunters, gatherers, "survivors on a desert island" than people who are obviously more dumb. But it also seems there are some skills useful in a hunter-gatherer lifestyle which are not highly correlated with IQ.
Having spent many decades in developing countries I have a lot of first hand experience working with populations around an average IQ of 70. Yes there are many great, hard-working, religious, serious people I am happy to count among my friends. No they would never have got to an industrial revolution.
Something funny and special happened in NW Europe and especially England. Similar things were in the air around parts of Asia. Greg Clark has interesting things on how England ratcheted up IQ through primogeniture. HBDchick had interesting things on how within the Hajnal line marriage patterns led to high IQ.
Good fun reading this and so glad it has escaped the Overton window.
I was in a small village in west Africa, doing blood tests for an endemic disease, and I left my bag with all the money for the whole study like 500 USD on a post in a field. And the African villagers just took the money back to me like you idiot you forgot it. They were making less than 50 USD a year and they just hurried back to me saying you forgot your wallet.
Since I missed the original, I just want to say: You are very, very brave. Ave Alexandrum, nos qui celare te salutamus.
The topic of racial or national differences in IQ has always tied me in knots. I’m pretty sure my knots are not unique so I thought it might be helpful to just name them here. There are 3 knots:
The first is that I grew up in the south, and absorbed what you might call the soft version of southern racism. My family thought it had been a mistake to integrate southern schools, because “nigras are just different from white people, honey.” They would certainly not have been in one of the groups of enraged whites protesting one little black girl entering the white school. They weren’t rageful people, and would have had empathy for the little girl. They just thought integration was a mistake because nigras were different. And what they meant by different was dumber, more prone to violence, and more sexual. On the other hand, my father had been parented mostly by a black nanny named Jenny. His own mother was a brittle socialite. He loved Jenny, and in fact when she was dying, when my father was in his 40’s, he traveled to DC to be with her and was actually holding her in his arms when she died. So blacks were seen as more primitive and animal, but for also as people who had a lot to give in contexts where earthiness and practicality were important. Anyhow, I did not question what I was taught about black people, and actually did not have much interest in the subject. I am now much better informed about race, and my feelings about the matter are much more nuanced. But a shrunken version of the old, of-course-they’re-inferior point of view is still drifting around in me, and occasionally I’m in touch with it.
The second is that I have always thought of myself as a smart person, and when I was in my teens things like IQ and SAT scores loomed large in my mind. That side of me has shrunk greatly over the years but it is not gone and I am ashamed of sometimes feeling better than other people because I can tell that in that IQ test, SAT kind of way I am superior to them. Also, there was an occasion in my life when the rubber really met the road as regards race and intelligence: I decided to adopt a foreign-born child, and had to decide what countries to consider adopting from. I ruled out Africa, and ended up adopting from China. I told people I was ruling out Africa because I did not want a child who would have to deal with the black/white issues in the US, and while that was true, it was also true that I was afraid an African child would have a lower than average IQ. I feel guilty about that.
The third is that I get occasional glimpses, sometimes on here, of people full of truly venomous hatred of blacks, people who sound like they would like to exterminate them. I feel guilty about participating in discussions that indirectly support the ideas of the venomous, and that might nudge neutral people closer to the venomous attitude.
None of these knots make me wish Scott had not posted his piece. But they interfere with my thinking about the subject. There are so many aspects of it that make me feel guilty. And I don’t feel able to distinguish clearly between being racist and being realistic.
...Don't feel too guilty about any of this. This is just a tiny forum of outcasts, divorced from social realities. History will move on, with or without us.
I am worried about many things but changing the course of history is not one of them.
I really like this comment, it's very honest on a subject where it's hard to be.
I layer on top of your personal guilt, which I share, a dose of suspicion -- why are some rationalists, the people who are most impressed by iq, are most proud of their IQs and are most convinced that bands of high iq engineers are going to bring about a possibly utopian singularity, the same people who are so keen to make group iq differences an area of focus?
I think i'd rather hear about the subject from people for whom iq is not already a major part of their identity.
Yes, agreed! And some of those people
also seem blind to certain forms of excellence not
captured well by IQ tests. I’ve seen a few tech guys post on here that we don’t need poets any more because AI can produce a poem on any subject with correct rhyme and meter. Yeah, but it’s doggeral! Somebody put up a post saying they didn’t believe anyone really liked Shakespeare— what’s really going on is that people pretend to in order to win culture points. Good grief!
Thank you for this very open-hearted comment.
Link to the post where Scott goes into detail about the tendency of correlations to stop working at the extremes.
https://slatestarcodex.com/2018/09/25/the-tails-coming-apart-as-metaphor-for-life/
people "with" low IQs, but who seem canny in most regards, might just have not tried very hard on the test and now are subject to... I want to say "contract-lawyering about the results", but there's really no arguing even happening. It's more like, some people see an IQ test result and expect not to be challenged if they treat the person like that's what their IQ is, and the person with the low result either doesn't think it's their place to contest the result or is letting society martyr them because they have nothing better to do
https://www.cremieux.xyz/p/national-iqs-are-valid title self explanatory 😅
I'm one of those people who thought these "national IQs" of 70 had to be wrong, and what I'm seeing in this discussion isn't changing my mind. Defenders are saying, IQ 70 doesn't mean that whole nations are retarded, in fact the average IQ of the USA itself was 70 in 1900. Is this really consistent with the historic usage of IQ, or are we seeing equivocation between different ways of defining IQ? I already know that at some point there was a shift from "ratio IQ" to "deviation IQ" and that this impacts the meaning of higher IQs; is there a similar rescaling of lower IQs too, that leads to a "softening" of what it means to be IQ 70 or IQ 50?
The gains in IQ from 1900 were mostly not on g or innate intelligence but on IQ subtests that are more related to 'being modern' e.g. getting better at taking tests, Ravens, weird stuff people in the Victorian age weren't used to. Subtests on vocabulary or general knowledge or counting backwards haven't changed much. If you met your great-great-grandfather they would probably have talked to you very intelligently about literature or politics or history but been very bad at reading comics or doing sudoku.
And people with an IQ of 70 are really perfectly fine and nice and interesting and good people. Just they don't do abstract thinking and have no interest in shrimp welfare or the Monty Hall problem. Can't do hard math, don't read poetry for fun, tv people.
What's the difference between IQ 70 and IQ 100?
It's not clear to me you're properly calibrated here. What you just described are "average" people, whose IQ is supposedly 100 in the U.S. An 85 IQ is the bottom 20% of Americans. In my experience the bottom 50% have no orientation whatsoever towards intellectual matters and can't do even easy math problems. A 70 should be something else entirely.
The people you are describing are not 70s and might very well be 90s.
"it seems like there's some disconnect between do-well-on-tests intelligence and run-a-subsistence-farm intelligence, and the abstract/concrete and novel/cultural distinctions are the best explanation that I can think of"
Exactly- people would be 80% less annoying about IQ if we called it "abstract intelligence"
Having worked with subsistence farmers in Africa and Asia for decades I disagree. The farmers who are better at tests and maths and vocabulary are better at subsistence farming. Plenty of studies show that good predictors of more profitable and productive subsistence farming are younger, better educated, smarter, and male.
"Africa is extremely genetically diverse, but I think most of the countries measured in the paper, including Malawi, are some variety of Niger-Congo speakers, who I don’t think are that much more diverse than white people or anyone else. The really interesting African ethnicities, like the Khoi-San, don’t show up as much at a national level."
there are two issues
- the average genome of an african is more 'diverse' than a non-african. so an african has 15-20% more SNPs (~5 million) in the genome (3 billion base pairs) than a non-african (a bit over 4 million). this is for an individual
- but, the pairwise between *populations* the genetic distance between populations is not always that great; that between various bantu-speaking ethnicities for example is similar to that between european nationalities. so not that high
naively, you could then make the cause that african IQ variance should be higher in the aggregate (since 15-20% more genetic variance), but the between-group differences should be like those in europe (i think the range is 95-105 except for a few outliers like albanians).
The idea that African subsistence agriculture is highly optimized reminds me of an anecdote in this recent Economist article. They discuss a very effective intervention for improving dairy yields: have the farmers stop starving their cows.
https://www.economist.com/special-report/2025/01/06/africa-is-undergoing-social-change-without-economic-transformation
And then your population would grow to consume the surplus. Malthus is mostly dead these days, but he can crop up in the subsistence-farming parts of the world.
Flynn himself has suspected that the eponymous effect might be due to more time spent in learning abstract thinking. He also pointed out that earlier IQ tests had more practical life questions and that IQ tests became more abstract and symbolic over time. Finally... It appears that modern populations do worse on these practical questions than the same countries did 100 years ago.
"First, he corrected my misconception about Reich on ancient European cognitive evolution. Reich had said that pre-agriculture Europeans were “2-3 standard deviations” below moderns. I had interpreted that as IQ deviations of 15 points, making them genetic IQ 55-70, which would have been pretty crazy. Stone tells me he actually meant PGS deviations, each of which was about 3-4 IQ points, so he’s claiming that pre-agriculture Europeans had genetic IQ of 90 (they probably also had lower IQ for environmental reasons)."
This is not quite right either. The real answer is that we don't know how to scale the PGS deviations from ancients to moderns. It is unwise to assume the modern correlation and use that. Think about it this way. If the real gap moved 45 IQ points (3 SD), and we use a PGS with validity of 0.33, this would mean the PGS would move 1 SD. There is no real way to interpret these PGS gaps in terms of real IQ points, or heights, for that matter.
Lyman Stone is still wrong about National IQs. His chief mistake, ironically, is trusting each datapoint as accurate. Single studies for countries are not that reliable as indicators of the mean IQ of that country. If one tracks them over time, one can easily get the impression nothing is reliable. This is the wrong conclusion to draw. I think in modern times Ron Unz was the first to do this kind of data digging, so we may aptly name it the Unz fallacy. In any case, Seb Jensen already replied to Lyman Stone. https://substack.com/home/post/p-155059660
As a numbers person, this is obvious about everything. If you want to know the average milk yield of cattle globally you will find a bunch of studies. In high income countries you likely find many, representative, large studies. In low income countries you will find fewer studies often with problems in sample size, selection methods etc.
And you will find that cows in America and Australia give most milk and cows in Africa give very little. The global milk yield of cattle looks strangely similar to global IQ or patents or homicide or anything else nice or nasty.
The few bad studies on milk yield from Malawi and Chad will likely show a lot of variation study to study and year to year. But the overall trend is obvious. It is just the same with IQ.
https://ourworldindata.org/grapher/milk-yields-per-animal
Harmonized Learning Outcomes use tests of learning outcomes, not psychometric tests. That preprint “updates” Lynn’s numbers by averaging them with the learning outcomes data. Do you see the problem with using a measure of development to estimate national IQ, then turning around and using that as evidence that national IQ causes development?
"learning outcomes" are really just intelligence tests, which they call PISA etc.
That’s maybe most true of PISA reading questions (it also has math and science which are pretty unambiguously functional competency) but generally a mischaracterization. TIMSS is math and science, most of the others are math and literacy/reading comprehension. They very obviously test and are designed to test education quality.
Please bear with the longish comment.....
There are many ways to talk about poverty in Africa, or in Malawi in particular. Malawi is one of the poorest countries in Africa. So you can choose how to describe the level of poverty. You can use hard economic measures, health, or education as defined by an alphabet soup of international organization. Education also has many different metrics, from basic literacy, years in school, tertiary education to the quality of education when attended. Alternatively, you can come up with statistics about IQ. (Btw I'm neither confirming nor contesting the validity of IQ testing as a measure of intelligence, the validity of those tests in Malawi or the accuracy of data in your map in the event it's actually valid). I'm just sharing a perspective as an African.
>(Sidenote - there is more genetic variation in what you call 'sub-Saharan' Africa than between Asians and Europeans)
The ethnic groups in Malawi overlap into Zambia. Malawians have also been significant migrants to Zimbabwe and South Africa for over a century and chiChewa is now an Official language in Zimbabwe. Ethnic groups in Zimbabwe also overlap into Zambia, Botswana, South Africa and Mozambique. The Batswana people live in Botswana, Zimbabwe and South Africa. I grew up in Zimbabwe learning and playing with kids of Malawian heritage before spending most of my adult life in South Africa. I suspect I'm qualified to claim that the Bantu Peoples are one big family. The genetic evidence is overwhelming and our different languages are obviously related. Maybe just anecdotal, bit there's nothing inherently inferior about Malawians or their intelligence. If the data is actually valid, I suspect they just measured a proxy of poverty - as if we needed other people to say it AGAIN.
African countries have structural macroeconomic challenges which have been studied extensively for decades. The extreme poverty in Malawi is certainly not because of low IQ. Abit of history before I make an ASK:
Firstly, Zimbabwe (as Rhodesia) and South Africa were the last bastions of European settler colonialism. For over a century, Europeans settlers knew Africans were capable. That is why they reserved lucrative skilled jobs for the European settlers. They knew that if they competed equally on merit, Africans will get the jobs despite 'native' education being underfunded by a factor of 10!
Secondly, Africans have been subjected to degrading studies to prove their inferiority for centuries. There're still thousands of human remains in European archives used for these studies. So far, they have all been proven to be nonsense and some of us thought Europeans and others would finally let it go. Africans were also turned into spectacles for the amusement of Europeans and even exhibited in zoos. Each time they protested the maltreatment.
The poverty in Malawi is a very serious matter and not a spectacle for your amusement. So. Will it be asking too much if I ask you to take down your articles? If that is too much, can I ask you do your studies about IQ and whatever other variables using data from other continents in future. Malawians are the most hard working and kindest people I know. Perhaps they behaved that way because they were migrants in Zimbabwe. I wonder where the'd rank in kindness stats???? So let me make a final request - If Africa is irresistible for you, can you use Zimbabwe instead? You have my permission. It is also a poster child of notions of African inferiority in online circles and will likely generate a lot of engagements too.
The natural state of humanity is poverty. Malawi is not poor because of colonial studies but because all humans were poor until an unusual sub-population in Europe put together innovations and institutions that moved them out of the brutal and grinding poverty that characterised the lives of my European ancestors.
I don't want to be snide, and I know it is a very common mis-perception that Africa is only poor because someone (who?) stole something (what?) from them. But economists have been studying this seriously for decades and this is absolutely incorrect. Noah Smith is a good populariser.
People can absolutely be hard-working, kind, very religious, have a low IQ and be poor. It seems that what keeps populations poor is not their propensity for hard-working or kindness but a combination of low IQ and strongly associated factors such as low executive function, high impulsivity, high discount values, little ability to do hard math, can't abstract, doesn't read or write poetry for fun, doesn't care about shrimp welfare, believes strongly in magic and god, good sense of humour, explosive strength in athletics, friendliness etc etc many which have a substantial genetic contribution.
That is the dismal truth. But the world as a whole will have worse outcomes if we refuse to recognise it.
I did NOT suggest that Malawi is poor because someone stole from it. Apologies if I gave you the impression that I’m looking for the reasons or bold insight that shrimp welfare and poetry appreciation are the missing links to poverty alleviation. I do not mean to snide.
My experience and all of the literature suggests Malawians are not just otherwise identical to north-western Europeans except that they happen to be poor. There are many differences between them and some of the differences are highly correlated. Loving Shakespeare does not make you rich, nor does ability to do advanced calculus. But understanding complex language and maths is closely related to high IQ. Countries with low average IQ are going to find it difficult to escape the poverty which is the natural state of humanity.
I myself have Autism, and as a kid I took an IQ test and they scored me at 126, IIRC, and told me I was in the 96th percentile or so - smarter than 95 other people in the average 100 person room. But I really don't feel that smart in day to day activity. So I think IQ just isn't very well correlated with day to day stuff at all.
All of this IQ discussion is making me curious. Is there any way for me to take the test without going into a psychiatrist office? Or, how do I convince my psychiatrist I need to take this test?
That is a very good one, and free. It only doesn't measure verbal intelligence.
https://is.gd/4JrTiG
"Even a "normal" low-IQ person cannot do the implied math and long-term planning involved in this kind of farming. And in fact, economists routinely find that African small-plot subsistence agriculture is actually highly optimized; farmers make very precise choices about where to plant which seeds, which fertilizer to use, etc. Key point is basically: it really isn't true that an IQ 60 person can run a farm functionally."
Makes sense to me. IQ tests are timed, so they partly measure speed. I'm pretty sure, I could figure out all questions correctly on a normal IQ test, if given enough time with pen and paper. That would not mean, I have the highest possible IQ at all!
Similarly, farming runs on seasonal cycles and you also get every incentive to get it right. If you spend an entire day on removing weeds, being 60 IQ probably means it takes you hours to figure out the optimal wheedwhacking pattern, that someone at 140 IQ spots within seconds and someone with 100 IQ spots within minutes. But with 60 IQ, you probably still spot it after a day, if you've got nothing else to think about. And then remember and tell your low IQ compatriots. They can perhaps not learn to fluidly navigate truckloads of abstract information from text or video, but you don't need to be comfortable with that amount of information that the farming lifestyle provides. Especially, if it's communal farming where knowledge/skills/responsibilities can be distributed, so no single person needs to understand the entire emergent workflow, which may in total end up very sophisticated and complex.
Except African small-plot subsistence farming is not highly optimised. It has extremely low productivity which is directly related to inadequate use of inputs and poor management. This is not just costly inputs like fertiliser and hybrid seeds but studies have found African subsistence farmers would greatly increase yields just by weeding (labour no input) in societies where people are chronically under-employed and not time scarce (except at short intervals around harvest etc.). Likewise if African farmers would give their cows enough water to drink they would greatly improve milk yields. Recent decades have seen African agricultural yields increasing due to cultivating more land (bad for ecosystems) and not through increasing productivity which is woefully low.
Iodine levels across the globe appear to have advanced considerably since the 2012 chart above - https://ign.org/app/uploads/2024/09/MAP-GEN-GP-SAC-2023.pdf