Why Sam Harris is Wrong About Race, Genes and IQ

In which we explore some introductory genetics and learn from Sam Harris’s mistakes.

Please consider supporting me on Patreon:

There’s already a sneak peek of an upcoming video available for patrons.

Footnote 1: Although Harris doesn’t explicitly state that the racial difference is (partly) genetic in this introduction, that is the implication that runs throughout the interview. He explicitly says so in the follow-up interview with Ezra Klein.

“[Charles Murray’s] simply suggested that there’s good reason to believe that genes and environment both play a part. That is a safe assumption for basically everything we care about physically and mentally. That is as safe an assumption in behavioral genetics as can be made. … We don’t know how much, I’m not saying that we know that the differences between various groups in IQ is all genetic, or even mostly genetic. But it’s certainly prudent to assume that genes are involved for basically every difference we’re going to find.”

Sam Harris Podcast and response:

Further Reading:

“Not In Our Genes: Biology, Ideology and Human Nature”, Steven Rose, Richard Lewontin & Leon J. Kamin 1990

[jamesspacer]

Pure bullshit starts at about 3:20.
Of course IQ is a proxy, no one is arguing it is not. It’s not like the IQ number itself is what’s heritable, IQ simply measures something that happens to be considerably heritable. Just like the speedometer needle is not what actually causes a car to go slower or faster, it’s simply a measurement of the speed. I love watching the mental gymnastics though.

[scottkidder]

Okay let me see if I understand. Sam claims that differences in intelligence between groups can be, at least partially, explained by differences in genes. You think this is incorrect for three reasons. 1) heritability is an establishment of a correlational relationship between genes and human traits, not a causal one. Therefore, heritability is not a valid measure of genetic causes and we cannot know whether differences in IQ are due to genes or to the environment. You used the example of our ability to predict whether or not a person will where glasses based on their genome. The trait of wearing glasses is heritable in the sense that you can predict it fairly accurately using genetic information. However, just because you can predict the trait of wearing glasses using genetic information does not mean that it’s the genes a person has that makes them wear glasses. After all, if you simply removed a person’s access to glasses, the heritability of glasses would become 0 despite the person’s genome staying exactly the same.
2) Heritability tells us nothing about what effect environment can have on any trait. Just because something is highly heritable does not mean that environmental factors will be ineffective. Here, you used the people on the island that were able to get taller because of the discovery of a new nutrition source. Their height had a heritability of 1, but when you change the environment, their height increased nonetheless. This demonstrates that even when a trait can be predicted by genes with 100% accuracy, the environment can still play a huge role in changing the average value of any given trait.
3) Heritability does not mean that differences in groups can be attributed to differences in genes. This is explained by imagining that two equivalent groups were placed in exactly the same environment but that one group found a better nutrient source and the other didn’t. Despite both groups having exactly the same genetic composition and despite height having a heritability of 1, we would still see a significant difference in mean height between the two groups and all of that difference would be due to environmental factors.

You then go on to say that Murray and others have attempted to get around this problem by controlling for the environment the best way they can. And this is often done by comparing twins raised in the same household by the same parents. But you dismiss this as irrelevant because theoretically, no two people can have the same environment, especially two people of different races due to discrimination and more oppression being experienced by the black child in general than the white child.

Here’s where I don’t understand what’s happening. First of all, I do think heritability has at least something to do with genetic causes. In your example, you use the trait of wearing glasses. It’s obvious that that trait is not a primary trait if you will. In other words, it’s obvious that genes can’t cause anyone to wear glasses. We know that people don’t naturally wear glasses and that the reasons they do wear glasses have to do with physiological issues with their eyes that can be predicted with genetic information. In other words, the primary trait is the physiology of the eye that results in impaired vision and that is almost certainly caused by genes. To me, intelligence is not a secondary trait that results from some other intermediary cause that is highly correlated with the primary trait. In other words, measuring intelligence is not like measuring whether or not someone wheres glasses, it’s like measuring someone’s height. And if that’s the case, then there is no reason to suspect that heritability does not imply some causal relationship between genes and traits. If that isn’t the case, then why bother measuring heritability at all? If heritability tells us nothing about the causes of genes, then it’s totally useless.
Second, you seem to be equating the causal relationship between good nutrition and height with better education/lack of discrimination and intelligence. It’s not obvious to me that an increase in oppression and discrimination leads to a decrease in IQ. It’s also not obvious to me that an increase in quality of education leads to an increase in IQ. I’m operating under the assumption that intelligence is not a measure of education or of social status, but rather an inherent trait. It’s somewhat similar to being talented in music or any other field. It’s not a measure of how good of a musician someone is, but rather their predisposed ability to acquire musicianship and their natural affinity towards it. Even when a child begins playing an instrument and is still not very good at it, you can tell whether or not they are talented. And it doesn’t matter if they’re oppressed financially, socially, or politically, their talent remains the same. Even if they don’t have a good teacher or any opportunities, you can still tell if they’re talented. My understanding is that intelligence is like that only that it’s within the domain of problem solving. I could be wrong, I don’t think that better access to education is to a higher IQ in the way that better nutrition is to increased height.
And lastly, you’re right that nobody can truly control for environment. But that is why we take averages of people and don’t generalize one specific person in a twin study to the entire population. And at the end of the day, I don’t know how you could humanely control for environment any more than Murray did. I don’t think we can afford to just give up and say that we can’t derive any useful information from those studies or any study in the future and declare that we will never be able to know if genes have any causal relationship to any trait. If that’s your argument, then the entirety of the heritability literature was all a waste of time and we can derive no conclusions from any studies done in that field, including studies that bolster our points of view. It’s a perfectly valid point, but I just hope that you never cite any studies regarding heritability in the future to bolster any argument you make and that you accept that we can know nothing about heritability at all ever to any degree of certainty. I don’t think that’s true, but you can argue that if you want to.

[JamieHumeCreative]

Grateful to have been directed to your channel. Following.

[BjornMoren]

What evidence do you have for the idea that a person's IQ is affected by how society treats him? I am not talking about withholding nutritious food, or access to education, etc, simply how welcome or rejected he feels.

[Ingeniir]

Ironically, in the Harris/Murray podcast you're criticizing (#73 - Forbidden Knowledge), Harris uses the same exact hypothetical malnourished-island scenario you used to illustrate the same exact point you make. Go to 43:46 of the podcast. Did you get that far?

[jerrysmith]

What is puzzling to me is why Sam is, on one hand, stating that "...genes appear to be 50-80% of the story…", and yet also saying "...there seems to be 'very little' we can do environmentally to increase a person's intelligence...".
If 20-50% is environmental, that doesn't sound like "very little" to me.
Additionally that 20-50% figure is just an average...the environmental factor could be way higher in one individual and way lower in the next individual.
And, if it were exactly 50%, would Sam then also describe the genetic factor as "very little"?
Having said the above, I don't agree with the video's theme that Sam was making a statement about heritability, but I can understand how someone reading "very little" could then interpret the "50-80%" as being a statement about heritability. That's just not how I interpreted what he was saying, because I give him credit for being smarter than that.

[puremercury]

The increase in average height example also is ridiculous. Murray accounts for the Flynn Effect in his writing (in fact, he and Herrnstein named the phenomenon). And heterogeneity in the Flynn Effect does not explain away any of the differences.

[luqkuh]

'Heretability can tell us nothing about the cause in group differences.' he says while in another breath insinuating that discriminatory or differential treatment is potentially causal to intelligence differences when there's practically no proof for that. Not validating studies because they don't account for that is laughable. Can you even quantify preferential or discriminatory treatment? Saying that a black kid might have a lower SAT or IQ score because they were side-eyed by teachers or pulled over by cops is hilarious.
Also I don't think you can compare height as a trait to intelligence. There are hypothetical situations in which you could through enviormental factors alter a groups height on average, not so much for intelligence. There is absolutely no proof that you can increase intelligence.

[ewanhassall]

A major flaw with this video is it ironically ingnors the very existence of bell curves itself. The statistical results described in the book bell curve are assuming the data follows a bell curve in fact the statistical measures only work when there's a bell curve. In every example you gave you to supposedly disprove the work that was done you always used an example that wasn't following a bell curve. The number of heads people have doesn't follow a bell curve. On your island example everyone got the same amount of food there wasn't a distribution food following a bell curve. If food distribution was following a bell curve then the inheritability measure would have been very accurate because you would have been able to see height change across food distribution. Measuring inheritability would have given you the correct results it's only because of this absurd example where the statistical work can fail. I thought this was particularly dishonest can you show how these statistical measures fail when the data is following a bell curve??! Obviously the distribution of IQ follows a bell curve and just about everything that can of be shown to effect a person's iq also follows a bell curve and therefore the inheritability should be giving us a very good measure of the rules genetics play in our IQ. It's deeply harmful to ignore this we're increasingly living in a high IQ society and we need to be able to show compassion to those of us who through no fault of their own have lower IQs by ignoring this we are instead blaming their failures on them being lazy or unethical or somehow of bad character and this not only prevents us from solving the problems but also makes those people feel worse as if they just aren't working hard enough.

[sirgodricenwardsaier]

This smells like a strawman. Where does Sam Harris use the fact that the heritability of IQ is 50-80% to infer that group disparities are heritable? Outside of morons in YouTube comment sections, I've never heard someone defend that proposition. A much more reasonable proposition would be that, given the presence of a phenotypic IQ gap between groups and the heritability of IQ, the expected value of the genotypic IQ gap is somewhere between the phenotypic gap and zero. Of course, this argument is far from conclusive, but Harris never makes strong claims on the topic of genetic group differences, and it does place the burden of proof on anybody claiming the gap to be zero.

[SBCBears]

9:28 "Heritability can tell us nothing about the cause of group differences."
Please, can somebody square that circle?

[RLSteve]

Charles Murray isn’t the only person claiming IQ is 50-80% genetic. Robert Plomin, one of the leading geneticists in adoption and twin studies, says the same thing quite recently in his book Blueprint. Yet everyone is claiming it’s just “Charles Murray bullishly.”

[LLWTF]

Sam Harris is a proponent of profiling. He thinks if for instance black Americans commit more crimes, then that justifies discriminator treatment of the group as a whole. What Harris never answered though is what amount of increased criminality is necessary to justify such actions. Because just recently he commented that White Nationalism is a "fringe" ideology and as such never a threat.


So I would really love if someone were to pressure Harris on an exact percentage of which he thinks it's ok to start discrimination. Because white nationalist terror in the USA was almost 75% of the total compared to islam or any other kind of terror acts.

[puremercury]

The last assertion is morally and logically absurd. This entire video is intellectually disgusting.

[maggiepaul]

What bothers me is that Sam was offered to have a conversation regarding this topic among the likes of Murray and turkhiemer and a few others; Sam Harris dismissed this offer entirely and continually pushes this idea as indisputed.

[ny]

Lol talk about reaching to justifying your position, it’s pretty clear Sam is right on this one.

[unclerhombus]

“Damn that’s a lot of bullshit…”

That’s how you open your rebuttal of him? You lost me right there.

[One-gvlr]

IQ by race follows the exact same order as as Brain size by race, GPA's by race and SAT scores by race.

[bycoadodbeking]

Poor evidence, there is better videos with actual evidence on the Subject of IQ. Also the studies dont say there is not environmental impact themselves. Do better work

[capoeirastronaut]

You need this paper: A dynamical model of general intelligence: the positive manifold of intelligence by mutualism