Hacker News Reader

Hacker News

Comment by NoImmatureAdHom | Hacker News Reader

Full thread

NoImmatureAdHom

In these comments: people doing apologetics for a vision of the world where everyone is born with equal potential.

I am a scientist, I work on these things, and I am here to tell you that:

--Intelligence is largely (~70-80% of variance) inherited

--Your personality is largely determined by genetics and is pretty much settled in to what it will always be by ~30

--If you don't learn to count right by ~4 you never will

--There are, in fact, many people who could never learn calculus

--Sex differences in behavior are driven by genetics and the effects are large

The broad strokes here are not up for debate, and denying them makes you analogous to a flat-earther. The details will change.

There are huge portions of the educated populace who will go to great lengths to spin stories about how it ain't so. They do this because it is comfortable to believe given the normative ethics we were all programmed with as children. Trying really, really hard to believe manifest falsehoods has a lot of practical drawbacks, though. Like the constant, expensive performance of searching for just the right way of teaching calculus to people who will never learn it--public policy that hasn't a chance in hell of ever working.

If you've read this far, and you're feeling upset or angry or uncomfortable with the facts I've listed, I encourage you to reflect on which problems in the world are really important to you. Write them down. For each one ask, "Is continuing to believe in a broken model of how things actually are going to help me solve this problem? Or would I be better off facing an uncomfortable truth and doing what I can with it?"

Thank you for reading my rant, I'll be here all day :-D

View on HN

Forgive me, but my understanding is that the heritability of IQ at ~70% is limited to those of higher socioeconomic status, and that it's closer to ~10% for those of lower socioeconomic status.

http://ibg.colorado.edu/cdrom2016/franic/Moderation/Lit/Turk...

https://www.taylorfrancis.com/chapters/edit/10.4324/97802038...

https://journals.sagepub.com/doi/abs/10.1177/095679761348839...

(I should note that beyond an undergraduate degree in linguistics I have essentially no professional qualifications in anything related to neuroscience. Most of my understanding comes from having read some Sapolsky and Steinberg cover to cover a few times, with a limited deep dive into late adolescent risk calculation and risk-taking related to a homicide case I was on.)

Is my understanding mistaken? Why?

Not OP, but have genetics background in other fields (not intelligence). The following is a bit of a mix

A way you could look at this would be to think as heritability here as the cap of your intelligence. No matter how hard you try, barring some sort of magical medical intervention, your maximal intelligence is limited by your genetics. However, that's the maximum. There are many ways to reduce IQ (lead a famous example here, nutrition another).

I think it's not an unreasonable hypothesis that those in lower socioeconomic brackets might increasingly encounter the damaging effects which might have an outsized effect on IQ. More factors than just heritability might create that effect you've mentioned.

>Intelligence is largely (~70-80% of variance) inherited

Also a scientist here....

Intelligence, however measured, is largely inherited. However, that inheritance is not genetic, but social.

It's now been rigorously demonstrated that common variation in genetics has very little to do with intelligence as generally conceived.

Just for some examples of well-powered studies:

>A genome-wide polygenic score constructed from the GWA results accounted for 1.6% of the variance of intelligence in the normal range in an unselected sample of 3414 individuals, which is comparable to the variance explained by GWA studies of intelligence with substantially larger sample sizes.

https://www.ncbi.nlm.nih.gov/pubmed/29731509

>in 2016, a second meta-analytic GWAS analysis with a sample size of 294,000 identified 74 significant loci. This analysis produced a GPS, EA2, that predicted 3% of the variance in years of education [a proxy for intelligence] on average in independent samples.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5985927/

Our inability to identify the genes responsible for intelligence doesn’t equate to them not existing. Our understanding of genetics is still extremely nascent.

I'll tell you what genes are responsible for intelligence if you can tell me which machine code instructions are responsible for Mario. ;-)

(

I might need to expand on that a little:

I predict that "the genes responsible for intelligence" are not a thing. I think that genes tend to work together to make a living organism similar to how machine code instructions work together to make a program run.

So for example: I'm pretty sure that looking at the design of eg. a 6502 can't directly predict the existence of Mario:

https://www.biorxiv.org/content/10.1101/055624v2.full

If you can't easily predict what neurons will do based on their layout, what chance do you have to predict the exact outcome of the genes which specify that layout?

You're going to need more data than just the genes alone.

)

The analogy doesn't work as we have indeed identified genes for many biological traits. For example, whether someone is vulnerable to certain diseases. So if iq is a biological trait, like eye color or height, it is odd that we haven't yet found any iq genes.

The analogy does work: we know the specific bits that are responsible for the color of Mario's hat.

I bet one could change one or more of several different sets of bits in several different parts of the program that would change the color of mario's hat.

(Perhaps thinking of it in terms of "what bits you should change" is asking the wrong question, instead it might be handier to look for the sprite definition, or to look for the sprite plotting subroutine)

This isn't a strong 1:1 mapping. Genes are not horoscopes!

No gene works in a vacuum. Genes code for proteins that work together to perform diverse functions. Sure: in some situations you can indeed directly identify that a gene knock out or substitution might affect certain functions, but this is rather misleading!

Genetics is actually very much Turing complete and Genes might best be looked at as a kind of software (this might be considered an understatement).

For comparison: I'm sure we agree that a single machine code instruction by itself is meaningless. However, if you knock out or add a JSR (Jump to SubRoutine aka function call) at the right spot: sure you can claim that the JSR codes for a particular functionality. But: JSR is 3 bytes on a 6502, would you really believe someone who told you that 3 bytes is what it takes to "make Mario jump"?

In reality there's a lot more instructions behind it (with perhaps calls to further subroutines, and more subroutines past that). Genes are not quite 6502 code of course, if anything they're rather more sophisticated.

To look at some of your examples:

You mention eye color which is determined by quite a number of genes working together (eg. location, color select, pigment) .

Height is affected by ... almost everything all at once over a longer period of time (eating, sleeping, age, eating at a particular age, metabolic rate : which itself is fairly sophisticated... there's wall charts that span an entire room). This to the point where some people give up and say "height is actually mostly environmentally determined". (Of course in reality it's an interaction between Genes and environment)

Vulnerability and immunity to diseases is very interesting indeed, if you eg. look at the role of somatic hypermutation in acquired immunity.

And all of these would obviously not work without working cells with DNA transcription, metabolism, membranes, replication, etc. ;-)

In short: be a little bit careful with claims in newspapers that say people have found a "Gene for X".

These studies are highly powered statistically. One conclusion you can definitely draw is that common variation in intelligence is not primarily caused by common genetic variation.

Evidently, our genes determine our brain, and are what makes us more intelligent than other animals. But between individual humans, genetic variation contributes to only a few percent of variation.

Or variation in intelligence is genetically controlled in ways we don’t understand yet because our understanding of genetics is nascent.

Yeah, sfblah is right. We will be able to predict based on our knowledge of the little molecules inside you at some point, we just can't do it now.

The 70-80% is heritable and not via parenting. It is contingent on an environment that provides things like food and air, and the basic idea is that we control for the environment in some simplistic way via twin studies.

To illustrate, there is almost no difference between the genetics of humans 20,000 years ago compared to today. Yet there is a vast difference in what people understand and how they interact with the world, i.e. there has been an increase in "intelligence" in the past 20k years without there being any alteration to genetics.

It is also shown by the so-called "Flynn effect" [0], which is a substantial rise in measured IQ by over 1 SD in numerous populations during the 20th Century alone. This is comparable to the mean difference in intelligence between individuals, and absolutely cannot be accounted for by changes to genetics.

[0] https://en.wikipedia.org/wiki/Flynn_effect

> To illustrate, there is almost no difference between the genetics of humans 20,000 years ago compared to today.

There is also almost no difference between genetics of humans and chimpanzees. No difference, except in a few areas that happen to matter a lot.

As it turns out, 20 000 years is plenty enough for natural selection to make dramatic change in the genetic makeup of the population. For a most obvious example, look at the spread of lactase persistence mutation, which occurred and started spreading less than 20 000 years ago.

> It is also shown by the so-called "Flynn effect" [0], which is a substantial rise in measured IQ by over 1 SD in numerous populations during the 20th Century alone. This is comparable to the mean difference in intelligence between individuals, and absolutely cannot be accounted for by changes to genetics.

The Flynn gains are hollow, though. They are not on g factor, and it’s the g factor that’s responsible for the predictive validity of IQ.

Imagine a society where people use human foot length to measure distances. After a century or two, they observe that everything is getting shorter, a sort of anti-Hubble effect. Nobody noticed, however, that people have become taller on average, and so their feet became longer, so the actual change is only in the used metric, not the latent variable they are trying to measure. Flynn effect is like that.

No, it is quite clear that, between people, there is only a very small contribution of variation in genetics to variation in intelligence.

This is independent of understanding how genes might affect intelligence.

I've never heard a number as high as 70%. Where does that come from? Otherwise: any idea how intelligence was operationalized?

I also think we can only accept that we're born different, but human. If we should aspire to anything, it should be to make life good for every individual, taking their capacities into account, not to give everyone an academic degree.

Here's a relatively recent study: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4270739/

I agree with your sentiment in the second paragraph. Human gets a little fuzzy, though...I think you really mean to point at personhood. There are humans who definitely aren't people, like those born without brains. And with the computers our grandkids will install--oh my! I'll cross that bridge when I get to it.

The percentage without context does not give any actual information. E.g. one could argue that environment affects 99 percent of intelligence as for example if someone dies due to environment there's no intelligence at all.

So there must be some poor environment A and good environment B, and some poor genes A and good genes B that a comparison must be made between to understand exactly what this percentage means and of course there must be a definition of intelligence and a way to measure it.

Thanks for asking, I should've explained. The percentage here is percent of variance (https://en.wikipedia.org/wiki/Variance). What it means, basically, is the extent to which you can predict a child's intelligence just knowing the parents' intelligence--not assuming the child was raised by those parents.

In your second paragraph, you've actually hit on one of the apologetic clichés that's popular now. People will say, "Yes, 70-80% of variance...GIVEN OUR CURRENT ENVIRONMENT". And they will act as if they've said something profound. Yes, that's true, and we don't know about the heritability of intelligence in environments that don't exist...but we can make really good guesses. Further, from a public policy perspective, we've got to work in the environment that we've got. And in that environment, the number is 70-80%.

No, from a public policy perspective, the important question is what environment can we create that benefit as many people as possible. The current environment, from prenatal care to availability, cost and quality of higher education is something that is created by public policy. It is certainly not anything given.

> No, from a public policy perspective, the important question is what environment can we create that benefit as many people as possible.

The problem is that since there already is substantial variance in environment that different people experience today, but nevertheless the amount of variance explained by shared environment is very, very low, this means that if we want to increase intelligence using environmental interventions, you can’t designate some environment that some (eg. well off people) already experience, and get everyone into that environment, because it simply won’t work. If it did, it would already show in the amount of variance explained by environment.

Instead, what you need to do is to intervene by putting people in environments very few if any people today experience. This is the only approach that can work, as it can overcome the low amount of variance explained by environment. Suffice to say, nobody yet figured out what those unusual environments actually would be like, as you can’t simply copy whatever well off or genius people grow up like. If you do come up with something, it might end up being a hard sale too, given that this will be necessarily stranger than what people normally expect. But, if it works, you’ll improve the world massively and make history. Chances are against you though, as millions of people in the education industry worldwide have already tried almost anything you could come up with, and nothing really works.

> what environment can we create that benefit as many people as possible

That can't be researched (effectively). Policies affect the current situation, not the desired one.

You create the desired situation by affecting the current situation.

But you can't base the decision that should lead to the desired situation on any present knowledge, and it's in practice impossible to research. You can't know the desired outcome, let alone the path towards it.

So it could be interesting to look at the environment side then?

Edit: Also, what do you call intelligence?

There's some intelligence difference between population groups, let's say Group A and Group B. Lets say that Group B is measured to be on average 5p more intelligent based on whatever arbitrary measure we took.

The debate here is about whether we should try to change the environment in such way that Group A and Group B would be more similar. We know that Group A in general has poorer environment however there's an argument that improving those conditions won't bring Group A to Group B level anyway since it's mostly genetical. Progressives want to change the environment, conservatives think it's a futile effort.

How would you determine in the following scenario whether levelling the playing field would cause the averages to be the same? Would 70-80% imply that, yes, it's likely and under what circumstances?

If the variance for a single person's measurement could be around 40p, then if genetics is causing 75% of it, then it would be responsible for 30p of that and environment 10p, then can we deduce that Group A given similar environment would actually be 5p higher in intelligence?

In this case while genetics describe 75% of the variance, environment is still enough to make naturally more gifted group less successful. So without knowing the other variables the 70-80% is still not meaningful.

70% is actually on the low side of today’s results. Observed heritability is lower in early childhood (meaning the amount of variance in intelligence among children explained by genes is “only” 20%-50%, “only” in quotes because it’s still large amount, as far as phenotypical traits are concerned), but grows to 80%+ in adulthood, even higher than that if you correct for attenuation (i.e. measurement error, which will always push the heritability down), and consider g factor instead of IQ scores (which is how intelligence is usually operationalized). IQ has been found to be exceedingly good vehicle to study the notion of intelligence, as the (most replicated finding in psychology) existence of positive manifold (google it) and g factor means that all non-contrived ways of measuring what humans usually understand intelligence is, yield pretty much the same thing, especially once you perform factor analysis.

Here is an overview paper by one of the leading researchers in the field (and one of the most distinguished researchers in psychology at large) Robert Plomin: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4270739/

Unlike the rest of psychology, psychometry as a field does not have any kind of replication problem: all major results have been replicated many, many times on many different data sets. You can go to Google Scholar and find probably hundreds of replications (keyword is “heritability of intelligence”). Here is for example one from Japan: https://link.springer.com/article/10.1007/BF01067719 which finds heritability to be 50% among 12 year olds, here is one from Norway https://emilkirkegaard.dk/en/wp-content/uploads/On-the-quest... finding 70% among military recruits, you can go on and on.

> I also think we can only accept that we're born different, but human. If we should aspire to anything, it should be to make life good for every individual, taking their capacities into account, not to give everyone an academic degree.

Clearly, but if the society denies reality, the result is expansion of higher education among people who can’t benefit from it, bullshit degrees, student debt, and general misery.

Robert Plomin is not without controversy: https://blogs.scientificamerican.com/beautiful-minds/there-i...

Well, yes, because he studies intelligence and doesn’t mindlessly repeat socially acceptable nonsense. This necessarily means controversy. You’ll find that most researchers in that field are “controversial” in some way, but it really tells you more about how society views the results, not the people obtaining them.

I just want to say thank you to xyzzyz for the detail I was too tired to provide :-D

It's interesting that you cite that Plomin article, since it actually concludes that only a small proportion of intelligence is genetically inherited. To quote:

>Using years of education rather than intelligence per se, the meta-analysis mentioned earlier with 125 000 individuals in a discovery sample yielded a GPS that accounted for 2 and 3% in two independent samples.16 More variance in intelligence is likely to be explained with GPS derived from larger samples, whole-genome sequencing and more novel strategies such as using networks of functionally linked genes.

Accounting for 2-3% of variance in intelligence through genetics is pretty weak support for genetics being an important contributor to variation in intelligence (however measured).

The denialism here lies with those who wish to believe that intelligence is largely genetically determined, rather than socially and environmentally determined.

You completely misunderstand the part you quote. What Plomin is saying is not that genetic heritability accounts for only 2-3% of the variance, but rather the GPSs, that is, the polygenic scores.

What that means is that we can now point to concrete genes that contribute to intelligence, and assign them weights that tell us the significance of this contribution. This means that this 2-3% figure is not about reality of genetic contribution to intelligence on the ground, but rather about our explicit understanding thereof in 2014, when this article was originally published.

Since then, our polygenic scores have significantly improved: for example, Lee et al in 2018 constructed polygenic scores that describe 11-13% of variance, a significant improvement over 2-3% result Plomin cites. There are probably even better results now, as the progress in the area is quick.

Of course, 11-13% is still short of 80%+ that we already know is determined by genes, but our explicit understanding is slowly getting there. Importantly, as we get there, the goalposts of critics are constantly shifting: when we couldn’t point out to specific genes, the entire notion of heritability was attacked. Now that we can, the alleged problems are with spurious correlations resulting from population stratification etc. Alas, the science moves forward, despite fierce opposition.

https://www.nature.com/articles/s41588-018-0147-3

> The denialism here lies with those who wish to believe that intelligence is largely genetically determined, rather than socially and environmentally determined.

I think your inability to distinguish between heritability and polygenic scores makes you thoroughly unqualified to make statements like that.

Polygenic scores are measures of the contributions of many genes to hereditability, what else do you think they are?

There are two questions here:

1) What genes contribute to human intelligence? Evidently there are some, as the structure of the brain humans are uniquely endowed with at birth is mostly determined genetically.

2) Do common variations in these genes contribute to common variations in human intelligence?

For 2) the answer is categorically "a few percent at most", even according to the sources you provide yourself. It's pointless denying this fact. Repeating that ~80% of variation in intelligence is genetic is displaying a limited understanding of inheritance.

To answer 1) it may be possible to identify those genes that contribute to intelligence by performing large scale genetic studies to identify the rare cases where a small signal is detected from one or several of these (a polygenic test, for example, as you are fond of).

Once these genes have been identified, one could then in theory investigate how they function, etc, in order to better understand the genetic basis of human intelligence, but it is unlikely to help understand the basis of the difference in intelligence between individuals.

> Polygenic scores are measures of the contributions of many genes to hereditability, what else do you think they are?

They are not “measures” in the same sense measuring tape is a measure of length. They are like credit scores: they are useful to predict likelihood of defaulting on a loan, but they do not give you full understanding of the entire reality on the ground. With better constructed credit scores you can get better at predicting defaults, just like Lee et al got better at predicting educational attainment than the 2-3% figure Plomin quoted. No sane person would say that Experian credit scores give you complete understanding of given persons credit ability and their future credit behavior, and no sane person will tell you that polygenic scores of intelligence/educational attainment give you complete picture of a person’s intelligence or education. Do you understand it now?

If you’re still confused, try to think about this: if what Plomin meant, like you suggest, that genes only explain 2-3% of variance in educational attainment, how is it possible that only 4 years later, Lee et al exhibited PGS explaining 11-13% of variance? Have genes got 4 times stronger in 4 years?

> For 2) the answer is categorically "a few percent at most", even according to the sources you provide yourself.

No, it is not. I encourage you to put some effort into understanding what is being talked about here.

> --Intelligence is largely (~70-80% of variance) inherited

Citation needed, especially given that there are enormous differences in intelligence between people who acquired speech and those who didn't.

I think the number of people that haven't acquired overt communication is so low, it can hardly affect the average.

My point is that if you can produce twins, one of which will be taught how to speak, read and write and the other not and the difference in intelligence between them will be so significant, then the genetic component can't possibly amount for most of intelligence.

Yes, it's the nature vs nurture argument all over again, but I think that regarding intelligence it's far from settled.

Well, that's why I wanted to know how intelligence was operationalized. But even in your Gedankenexperiment, there are ways to estimate heritability of intelligence: you'd need 100s similar cases, and, e.g., test them on puzzle solving and learning skills.

Edit: sorry, I asked that question in another thread...

Yeah, as the other child comment says I'm not sure what you're getting at with speech acquisition. Here's a study I plucked: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4270739/

Here is a quote from the study:

> However, for intelligence, heritability increases linearly, from (approximately) 20% in infancy to 40% in adolescence, and to 60% in adulthood. Some evidence suggests that heritability might increase to as much as 80% in later adulthood but then decline to about 60% after age 80.

It does not corroborate the claim "Intelligence is largely (~70-80% of variance) inherited".

That sounds right on the money to me?

Variance in all traits is higher in childhood. The idea is that you grow into your mostly pre-determined self.

No, it's not on the money. Do you really not see the numerical discrepancy?

I hope this need for a citation comes from genuine curiosity and not some misguided ideological want to uphold the notion of equality in intelligence.

It just doesn't make sense given a more macroscopic view of things. If inheritance obviously controls the difference in intelligence between the human brain and the dog brain, or the cat brain, or the monkey brain, what black magic makes it so that among humans themselves inheritance controls very little of intelligence?

The 70% - 80% statistic seems like a ballpark estimate he pulled out of a anecdotal conclusion based off of a broad spectrum knowledge of the related scientific research. You likely aren't going to find a citation that proves this claim definitively. It does not mean that his point is incredulous nor does it mean that his point isn't part of common sense.

His point makes sense. It makes a lot of sense. The same sense as the fact that the covid vaccine protects against the virus even though for the longest time it was NEVER FDA approved (aka no strong citations). The conclusion was obvious despite not meeting the stringent verification of the FDA and many people were able to arrive at this conclusion without the science. Yet many people refused to believe the efficacy of the vaccine due to their political ideology.

I believe there's a similar biased attitude coming from the far left side of thinking. The unwillingness to see that the science and common sense points to an inescapable conclusion. There is no unseen magical force that makes all races, all genders and all peoples equal. There are differences, and the differences are biological and obvious.

Like literally genes control how ugly and how tall you are but not intelligence at all? By some crazy magic intelligence is the one thing that isn't at all influenced that much by genetics? Seriously, which viewpoint is more realistic here?

Seriously the same ultra left people who were calling out the anti-vacc people as stupid and unintelligent for refusing the vaccine are unable to see the exact same kind of ideological stupidity within themselves.

Yep, I agree with basically all of this. A reasonable rhetorical move when speaking to science-deniers about inheritance is, "Do you have a principled way of telling people from rocks on the basis of intelligence, then?"

People like me would be more convinced by your rant/flamebait if you provided sources for your claims. You claimed that the "broad strokes here are not up for debate" so it should be trivial for you to provide sources. Since you don't, it's hard to take your arguments seriously. Everyone can make unsubstantiated claims, providing evidence is the difficult part. I'm sure you understand that the "trust me, I'm a scientist" argument doesn't fly around here. :)

> People like me would be more convinced by your flamebait if you provided sources for your claims.

To be honest, from what I've read in these comments: I doubt it. There's a lot of goal-post moving and deflection ("but what even is intelligence", "but what if we never taught people to speak", "but why should it even matter") that I believe Betteridge's law of headlines still holds.

I don't think denigrating others contributes to a healthy debate.

I didn't read it as denigration, it was a legit point. We make inferences based on the type of discussion.

Yeah, as I said elsewhere this is a reasonable request. I was writing in a moment of passion very late at night. I am supposed to do a variety of "real work" today but I'll see what I can gin up.

The things I listed are "the sky is blue"-type claims to people who do evolution stuff.

Regardless of your credentials, as far as I am aware the scientific consensus on this topic is that genetics is nearly impossible to separate from other causal factors. Mainly other causal factors that we have clearer and more direct evidence of having an influence on cognitive ability.

So, quite frankly, I don't care if you spend a lot of time thinking about it, it's irrelevant. The consensus is clear.

Yes, exactly like that!

If you would be so kind to explain to me why I should believe you, a random person on the internet claiming to be a "scientist" who "spends a lot of time thinking" about this subject, over thousands of peer reviewed research articles and case studies stating that even if genetics is a factor in intelligence then it is so closely tied to other causal factors that it is effectively impossible to control for, then I am happy to listen.

So please, I am begging you, tell me why I should believe you over everyone else.

That is not, in fact, what "thousands of peer reviewed research articles" state. You might be confusing the % of variance we can predict from genetics (e.g., knowing your ATGCATAGCCGTAG code), which at present is maybe 4%(?), with what we can show is due to heritability (70-80%). We didn't even know about DNA when people started measuring this kind of stuff with twin studies in the early 20th century.

So, to reiterate because I'm realizing that para wasn't so clear: guessing how smart you are based on your genetic code is something we're only just learning how to do. We're getting better at it though. Estimating how smart you are based on how smart your parents are is a totally different game that we've been playing for a century.

Why shouldn't intelligence be inherited while other traits are? It makes little sense to me, especially because intelligence is so important for survival.

> "spends a lot of time thinking"

Why is this in quotes when it isn't a quote?

There was another comment I had just finished reading and merged them in my head. I'll go back through and edit, sorry for any confusion

How does your science control for differences between parenting, access to education and nutrition though? I would expect doing so to be unethical and therefore not allowed.

Twin studies is the oldest, best answer. People have fancy statistical trickery these days too.

Would you mind elaborating on exactly how many studies you think are engaging in "statistical trickery"?

"exactly how many"--I have no idea. I didn't mean to put those studies down, the statistical trickery can be quite cool! Hard to do right, though.

I think at least some of those statements are independent of parenting, access to education and nutrition.

Give everyone perfect parenting, a great education and lots of nutritious food and some of them still won't be able to learn calculus.

Give everyone the perfect genes and lack of environment can also cause them not to learn calculus.

You are misunderstanding. Calculus is just a placeholder for something that requires a particular level of intelligence.

It's about whether it's possible or not, not whether you actually do learn the thing.

Someone with an IQ of 120 may not ever learn calculus, but they probably have the ability to do so. You cannot say the same of someone with an IQ of 80.

Yeah, but the whole debate is about how much is genetics responsible for intelligence vs the environment. There was a claim that genetics is responsible for ~70-80% of variance. Pick any intellectual ability and lack of environment can cause for an individual to not be able to achieve that, so how can genetics be responsible for that much variance?

The question is where and how does the 70-80% come? How is it meaningful and how is it measured?

It's 70-80% conditional on (some sampling from) our current environment. Obviously if your environment is boiling sulfuric acid you aren't gonna be smart.

But why only calculus? Why the focus on a single subject? This is a very limited view of intelligence.

I was continuing the example the OP gave. Calculus is just a placeholder for something that requires a particular level of intelligence.

It wouldnt matter what subject was picked, the point they are making that 70% of your ability in a domain is going to be genetic. It could be distance running, calculus or marble sculpting.

Thanks for your rant! It was wells stated. I did not know anything of these things in the sense of having any academic knowledge about this stuff.

I didn't know about the counting one and am not surprised by this. In a thread on another post I wrote that there is a lot of brain washing that occurs in the teaching of low level mathematics. It's easier to brain wash children in this sense than it is for adults.

You've made four claims and thus far provided a source for the first one. Any chance for the other four?

It's a very reasonable ask. I'm not on my work computer, and anyway I don't have the classic papers at the tip of my tongue, so I'd have to look them up. So I guess I've got nothing for you but an apology. Perhaps when the sun rises.

In this day and age, it is easy to get the actual degree and be called a 'scientist'.

In these Internet days, it is even easier to claim you are a scientist when you are anonymous online, and make all sort of claims.

>In this day and age, it is easy to get the actual degree and be called a 'scientist'.

>In these Internet days, it is even easier to claim you are a scientist when you are anonymous online, and make all sort of claims.

OK, but what about their comment do you disagree with or find in error?

> Intelligence is largely (~70-80% of variance) inherited

I'd like your definition of intelligence.

"g" is the way contemporary scientists think about it: https://en.wikipedia.org/wiki/G_factor_(psychometrics)

I haven't read the wiki article and so can't vouch for it

Cool, so malnourishment throughout life will only account for at most a 30% decline. Or is that "inherited" because it's environmental, but determined by parents?

The assumption is your environment is pretty similar to everyone else's environment, which means you don't die of hunger or get boiled in acid.

But you raise a very important point, and you're right: as far as I'm aware, latest evidence says that being malnourished as a child does significantly impact intelligence as it does height (stunting). Though I believe the last piece of work I read on it claimed the effect was smaller than previously believed. That was ~2 years ago.

Yes and no, and yes again. And no.

For any given trait X, the question about how much it is "inherited/genetic" or "environmental" is a commonly used shorthand, even amongst professionals, but it can be subtly misleading. All traits are 100% inherited, and all traits are 100% environmental. What do I mean by this? How is the ability to speak Russian genetic? How is the colour of your hair environmental?

Well, put it this way: is the difference between a human child brought up in a Russian family and their pet dog who has lived with them for the same time genetic or environmental? And what is the difference between a pair of identical twins, one of whom bleached their hair with peroxide–genetic or environmental?

The above are of course exaggerations, but they're just obvious examples. The correct way of phrasing the question is this: given a population A, what percentage of the variability of trait X is inherited and what percentage is environmental? Given the population of mammals, is the ability to speak Russian mainly genetic? Yes. Given a pair of identical twins, is any difference in their hair colour environmental? Also yes.

So, to the question as to how much of a decline in intelligence can be accounted to malnourishment throughout life (and how much to other environmental factors, genetics, etc.) we need to specify what population we're talking about. I'd be willing to bet that:

1- the variability in intelligence due to malnourishment in the population of middle-to-upper class children is negligible,

2- the variability in intelligence due to malnourishment in the entire child population of a developed country is detectable,

3- ...and the variability in intelligence in the due to malnourishment in the worldwide child population is significantly higher.

Reply on news.ycombinator.com