Hacker News new | past | comments | ask | show | jobs | submit login
Comparing ancient and modern genomes for cognitive ability variants (biorxiv.org)
51 points by gwern on Mar 1, 2017 | hide | past | web | favorite | 44 comments

For those politically on the centre and left, what are the correct responses to scientific evidence that different population groups have different cognitive abilities?

I always regarded myself as being a left libertarian sort of type. My reaction to learning about individual differences being so profoundly driven by genetics - never mind racial differences - was simple: research and subsidize genetic engineering. For everyone. To the point of being free. Level the genetic playing field. There is not the slightest bit of justice, fairness, or economic optimality to many people being born with high genetic risk for stupidity, schizophrenia, heart attacks, or anything! Anymore than there is in being born with a silver spoon in your mouth or being born female. No different from Headstart or any other program; if we can do it safely and cost-effectively, we should.

If there are racial differences, then that simply means it would be more efficient to allocate more of those funds to people in that group. (Actually, if there are racial differences, this might even be good news; the massive efforts in the USA to close the black-white difference have largely failed over the past half-century, and I have more faith in genetic engineering becoming practical than I do in another half-century of trying the same thing suddenly succeeding. Wouldn't it be great if the black-white gap was just a few genes or insufficient iodine or something? We can fix that - all it would take is money and science! Maybe not even more than a few billions, chump change! Whereas if it's structural racism...)

(Subsidising equality via the state doesn't sound 'libertarian', and Headstart has been ineffective, but I'll leave that aside)

I think your solution cannot happen because those with money and tech, and who don't have a cultural squeamishness with the idea, will get the genetic engineering first. The ultrarich. Probably the Chinese, maybe others. Then it will be a case of making the rest of the world amenable to these enhanced people. Elites don't want chaos and nor do they want anyone to threaten their dominance. So a functioning, intelligent, but malleable mass will be controlled by an elite who have the tech. Just as today the average joe doesn't have access to nuclear weapons.

I never said I was a doctrinaire libertarian and I think it's fine for the government to spend on things which pass a stringent cost-benefit test, which genetic engineering would be able to considering the many positive externalities; and I know Headstart doesn't work, but if it did.

"I think your solution cannot happen because those with money and tech, and who don't have a cultural squeamishness with the idea, will get the genetic engineering first. The ultrarich."

In some scenarios, not in others. With embryo selection, the gain is small enough that it will not create new elite classes with high solidarity, and in iterated/synthesis scenarios, an elite could try but the core technologies are widespread now and the end result gametes are easily stolen and replicated (imagine trying to do nuclear arms control if every single atomic bomb was walking around, having kids, and regularly shed skin cells which could be turned into new atomic bombs).

That said, this is of course good reason to oppose regulation, as regulatory capture is common and we already see a lot of self-serving hypocritical behavior by elites.

So, eugenics. Foss eugenics. I'm viscerally repulsed by this, but, then again, enh.

I think this is a hard truth to accept for people of any political persuasion. A lot of right wing ideology rests on the idea that anyone can pull themselves up by their own bootstraps and make it.

One could see this as a rallying cry for communism "From each according to his ability, to each according to his need".

I think the left has a tacit acknowledgement that those who cannot support themselves should be supported by society.

Does the fact that a certain group is genetically predisposed to some condition mean we should discriminate against that entire group? I think the answer is obviously not.

The question of diversity gets messy if you assume that this is because everyone is equally qualified and the current state of affairs is only due to discrimination, rather than seeing it as a way to improve society-wide outcomes.

It's a messy topic since much of today's discourse is based on the idea that we are or should be equal when that is clearly not the case, but we keep the fiction going to avoid looking too closely because it is uncomfortable for everyone.

> this is a hard truth to accept for people of any political persuasion

I think it should be welcome by people of all political persuasions. To explain inequality, the Right tends to assume epidemic laziness (wanton self-destructive behavior) while the Left tends to assume epidemic racism and hatred -- neither of which is particularly nice. We can still choose whether we want equality of opportunity or equality of outcomes, but we can make that choice with compassion and have a much better shot at effective policy to support it when we acknowledge the truth.

> A lot of right wing ideology rests on the idea that anyone can pull themselves up by their own bootstraps and make it.

The right-wing bootstraps ideology is as absurd as the left-wing equality ideology. The Randian right-wingers don't REALLY believe you can make it anywhere, otherwise they would be happy to move to Mogadishu or Kinshasa. Or even Flint, Michagan. Just like liberal white flight (revealed preference) contradicts liberals equality dogma (stated preference).

I doubt my opinion is the 'correct' line but:

That there are speculative and probably marginal upsides to basing policies on this premise and there are concrete and known downsides to not do this (Jim Crow laws, Eugenics programs have been largely judged as abominations rather than successes).

Which is to say whether or not different populations have different (and I mean different as in different, not linearly better or worse) cognitive abilities may be true without being actionable or useful.

What is correct when responding to scientific evidence is does not change with one's political views.

The correct response is the same as for any other scientific result: Was the experiment designed correctly? Was it executed correctly? And (most importantly) has it been reproduced by other investigators?

More importantly: does the result matter? In this case it's swamped by other effects. It would be interesting to a scientist, but not to an engineer.

The tiny effects claimed per genetic variant also call the scientific validity of the result into question. If you use a large enough sample and a broken enough null model, of course you're going to see "significant" effects.

Genetic variants are correlated. If you have thing 1, then we can somewhat predict thing 2 and thing 3 and so on.

Thus the effects are not really tiny. The direct effect of a genetic variant may be tiny, but it is associated with many genetic variants with similar effects.

This is most obvious in the Ashkenazi population, which has been so strongly selected for intelligence that it now suffers from numerous recessive problems. If you have a trait found in that population, the chances are good that you have other traits found in that population.

Look at the effect sizes of different factors that affect intelligence.

Child and prenatal nutrition and parasite load are much larger factors in IQ variations between population groups. Genetic variation between groups is much smaller and the effect size is very small. In practice it disappears into variation between individuals.


How do you explain the IQ difference between, say, American whites and blacks? They should have similar parasite load and nutritional status.

>They should have similar parasite load and nutritional status.

This is strong assumption. Malnutrition (wider concept than just calories) exists in US and there are large differences between ethnic groups.

Interestingly enough, from the same researchers:

Parasite prevalence and the distribution of intelligence among the states of the USA, https://www.researchgate.net/publication/229409035_Parasite_...

>In this study, we tested the parasite-stress hypothesis for the distribution of intelligence among the USA states: the hypothesis proposes that intelligence emerges from a developmental trade-off between maximizing brain vs. immune function. From this we predicted that among the USA states where infectious disease stress was high, average intelligence would be low and where infectious disease stress was low, average intelligence would be high. As predicted, we found that the correlation between average state IQ and infectious disease stress was − 0.67 (p < 0.0001) across the 50 states. Furthermore, when controlling the effects of wealth and educational variation among states, infectious disease stress was the best predictor of average state IQ.Highlights► We compared rates of infectious disease with average IQ across USA states. ► We controlled for education and wealth. ► Infectious disease was the best predictor of average IQ.

The article you linked to is paywalled, which offers the abstract, Infectious disease was the best predictor of average IQ. But the study appears to conflate infectious disease with parasite and offers no way to look at the data they used.

Correlation is not Causation: You might want to look at the prevalence of infectious diseases by race according to the CDC.


Edit: Insert Numbers;

Black People make up 14% of US population and

45% of the total of HIV/Aids

62% of women with HIV/Aids

55% of Gonorrhea

Chalmydia: Black Women 5.2x more likely and Black men 7.3 times more likely as white people

38% of Syphilis cases Black Men and Women over 5 times as likely as white people

37% of TB cases Black people, and 87% of all reported TB cases occurred in racial and ethnic minorities

You are supporting my argument with the evidence you provide.

How so?

If the states with lower IQs have a higher percentage of African Americans than the other states, they will have a higher rate of infectious diseases due to the propensity of infectious diseases among that group with respect to others.

PS: The two states with the lowest rate of infectious diseases, Nevada and West Virginia are respectivelly the 2nd and 7th lowest IQ states. #1 lowest IQ, Hawaii was deemed an outlier because of latitude and excluded from the results of the study you linked to.


Thirty Years of Research On Race Difference in Cognitive Ability, aggregates over a dozen studies from the past few decades, looking at everything from twins raised in different environments, the effects of adoption, location of the population, etc, using everything from multiple IQ tests to the ASVAB.


Its worth reading and answers your question, while refuting the OP above

I never claimed that there is no differences between ethnic groups.

I suggested looking and comparing the magnitudes between different factors and putting them into order of importance.

We know that taller people have higher intelligence than shorter people on average [1]. Compare effect sizes between black and whites and tall and short white males. Now ask if the difference between ethnic groups is important or not.

[1] : https://en.wikipedia.org/wiki/Height_and_intelligence

ps. 90% of genetic variance in human races exists in the African continent. If we clean out all environmental factors, the "smart genes" are probably in Africa. We just need to find them and distribute them.

> 90% of genetic variance in human races exists in the African continent

Counting neutral variation (the vast majority of genetic variation). You can get plenty of variation in non-coding DNA from drift without any effect on the variance in a trait. In fact, you rarely actually get effects on highly polygenic traits from drift without selection (imagine random directions chosen independently for each of the relevant alleles, and think in terms of the law of large numbers).

If you're getting this from Oded Galor or Deidre McCloskey, they're generally smart people but don't understand population genetics well.

Taller people have higher intelligence than shorter people on average because, brain size is positively correlated with height, and brain size is positively correlated with IQ.

>Studies on over 700 participants show that individuals with larger brain volumes have higher IQ scores. About two dozen studies using magnetic resonance imaging (MRI) to measure the volume of the human brain have found an overall correlation with IQ of greater than .40

>Within each race, cranial capacity correlated with IQ scores. By age 7, the Asian American children averaged an IQ of 110; the White children, 102; and the Black children 90. Because the Asian American children were the shortest in stature and the lightest in weight while the Black children were the tallest in stature and the heaviest in weight, these average race differences in brain-size/IQ relations were not due to body size.

>Overall, MRI studies show that brain size is related to IQ differences within race. Moreover, the three-way pattern of group differences in average brain size is detectable at birth. https://www1.udel.edu/educ/gottfredson/30years/Rushton-Jense...

Masai people must be super clever and Finnish people too. It must be true because I know a lot of really clever Dutch people and they are all really tall! Chinese people must be really stupid. I am so glad you gave me this insight. I am tall and I feel great now.

He/she is quoting "these average race differences in brain-size/IQ relations were not due to body size", i.e. it is an average within a population group rather than amongst population groups so you appear to be making and attacking a straw man argument.

Perhaps the NBA is an untapped source of nuclear physicists.

Yao Ming cracked cold fusion years ago!

The question is posed as though this is settled science, when it isn't. Leaving aside causes of difference, the concept of "cognitive ability" is, at the very least, fiercely debated as being variously: culturally biased, overly narrow or mechanistic, or unmeasurable.

Measurements like IQ or g are only validated by being "correlated with important social outcomes" and "correlated with job performance", uncritically accepting the current society or structure of work as natural or inevitable rather than as products of culture.

If "cognitive ability" is restated as "how likely an individual is to succeed and thrive in the dominant culture" it becomes less surprising that there are differences between different "population groups".

Irrespective of left, center or right, scientific values dictate great skepticism of these results. The authors of the GREML study do not address the possibility that performance on the intelligence tests correlated with social status, which is a highly heritable trait but for reasons completely unrelated to biology.

Essentially any evidence based on a Genome-Wide Association Study should be regarded as suspect. They use extremely simple population-genetic null models, and it's extremely rare for them to turn up novel, testable biological insights.

If I am socially conservative and fiscally liberal, where does that place me politically? What "should" my response be?

I honestly have no idea.

"Response" in what context and for whom?



As a fellow leftist, I would counter that each side clings to certain politically convenient mistruths, and not recognizing these tendencies in all political groups is indicative of thorough brainwashing.

I would be astonished.

And after that?

Sorry simple question, simple answer and probably should have only deserved some disdain as I can't see what politics has to do with it unless you are some ignorant white supremacist. My limited understanding of the actual scientific fact is that, given that any 'group' or 'measure' is fairly arbitrary and tests for cognitive ability quite subjective culturally and even by species and so on, I would be very surprised if there weren't any differences. If we are really getting into haplotypes and have some way of not skewing whatever we do with whatever biased measures then I still say 'but of course individuals and their relations have different cognitive abilities - duh! They are different people with different abilities. The sum of those people are just the sum of those people. If we can get down to this gene makes you smarter at this test then still I say: 'but what value is that test in measuring overall cognitive ability?' The chimps will say of course that they are 'smarter' (if that is really where we are meant to be going with this because they are obviously superior to us baldies by this test:


All chimps are monarchists btw.

My understanding is education and poverty are the biggest factors in IQ type scores. That is to say, they drive cognitive abilities in practice. Those two factors are tightly correlated with ethnicities and cultures, of course, so it gets very messy.

Anyway, my opinions are these:

* school k-12 should be free and available to all, with well taught and well compensated teachers, in facilities that are well maintained (posh facilities are pointless, but lead-free water is a must).

* Post-secondary education should span blue-collar and white-collar work and be free to all.

* Unions should be promoted and/or required for midsize and large employers.

* Strict regulations with relation to racist hiring and behaviors in the educational and job force.

The above will work to give everyone a reasonably equal starting point; if we can achieve that as a society as a stable fixedpoint, we've done well. I doubt I or my children will see that day.

I don't claim to speak for the Left, but I am a reasonable facsimile of a middle of the road Democrat in the US. I think that I rather like Churchill's definition of Liberalism from a century ago(slightly eliding the speech[1]):

* Liberalism seeks to raise up poverty.

* Liberalism would preserve private interests in the only way in which they can be safely and justly preserved, namely, by reconciling them with public right.

* Liberalism would rescue enterprise from the trammels of privilege and preference.

* Liberalism seeks, and shall seek more in the future, to build up a minimum standard for the mass.

It's also the case that popular analysis of population differences has historically been channeled into racial theories, which tends to get very ugly in practice. So as a non-utopian policy matter in 2017, it's best to downplay this research, because it attracts Nazis like viscera attracts flies.

[1] http://www.encyclopedia.com/politics/legal-and-political-mag...

It's great that we have access to ancient and historical genomes. Because of the development of medicine we've probably accumulated a lot of random mutations in our present-day genomes and at some point we'll need to identify them.

I only read the abstract, but the result seems counterintuitive. I would have expected that as civilization developed the selection pressure on low cognitive ability would be reduced.

Civilization also can mean high population density, far more intricate social networks and occupations, intensified warfare, possibility of better nutrition, more need for long-term planning to save up money and manage agriculture, etc. (Of course, civilization can also work the other direction as the past few centuries suggest. That there might be a long run trend to select on intelligence doesn't mean that there are no short-run trends select against it.)

Doesn't surviving and thriving w/o civilization take more effort mentally and physically? Isn't civilization a labor saving device, and viewed as such the tool user doesn't need to be as skilled or strong.

It has never been easier to kiss your bosses ass, take an early bus home and order a pizza with your smart phone en route.

Physical sure (so far as strength and endurance). Mental though, no way. Because in a populated area the competition isn't nature, it's other human beings, and outsmarting them is key.

It's important to remember that the metric of success in evolution is not outsmarting other people. It's how much you can spread your genes!

Doesn't look like they have any pre-holocene baseline unless they think there were still major non-agrarian societies 1200 years ago. Or did I miss something?

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact