Not to be persnickety, but I think too many people over-index on heritability estimates: what they are, how large they are, whether they're extrapolate across populations.
Heritability estimates are population specific and don't necessarily represent some fixed reality. Heritability doesn't tell us how innate a trait is (e.g. heritability of having five fingers is ~0 but it's wholly innate). Heritability doesn't tell us how modifiable a trait is. It's just the portion of the variance in a population explained by genetic variation.
Additionally, one of the more eminent population geneticists (David Reich) openly acknowledges that the likelihood of population differences across many traits is highly likely. It's just unlikely that these are large differences on average and that they accord with common or racist stereotypes.
Of course, I'm happy to talk with you about it (can take to DMs to sort out). I agree with many of Dr. Gusev's critiques but not necessarily with all his conclusions about the genetic contribution of neuro/psych traits. So it's a middle road position which I think is often tacitly held by many academics in these fields.
The heritability of intelligence is hard to estimate (everything is modeling, at the end of the day), but the largest GCTA estimate corrected for attenuation is still 48% (https://gwern.net/embryo-selection).
Has anyone written a direct response to Sasha's sources (he claims 15% heritable) that you think is best?
Some less relevant thoughts:
Is it possible that intelligence gaps are 48% heritable when comparing individuals, but that racial gaps are more environmental? I was looking at some SAT data that showed whites outperforming Asians in Washington DC, and Hispanics performing at the same level as whites in West Virginia, and blacks sometimes performing better than Pacific Islanders and Native Americans, and sometimes worse.
On the other hand, if there was a 10 point IQ gap between individuals, could that level of variation be possible? For example: two people take IQ tests at random times. How much does IQ vary per test (maybe due to bad sleep or an emotionally traumatic event like an argument or relationship breakup), what are the chances that a 10 point IQ gap would close? I would expect large groups like races to be less subject to "bad days" than individuals.
I'm guessing the hereditarian position would be "West Virginia whites are subject to brain drain and dysgenics, or environmental toxins from coal mining." My question would be, how quickly did the WV-white vs average white gap form? Was that due to a founder effect from England (Albion's Seed) in 1600? Or the result of de-industrialization since 1970? If we can quantify the rate of change, then we could say that IQ can be increased or decreased at that rate. Perhaps as slow as 1 point per 100 years, or as quickly as 1 point per 10 years. That would make racial gaps a question of timescale -- a range between 150 to 1500 years.
I'm working on an article expanding on the idea of medieval population replacement of non-elites by elites, and projecting that into the future. The basic idea is if average people adopt Seoul fertility (0.5), and elites have a TFR of 5 (I have to explain how that could happen), then elites would genetically replace the population within a few centuries, creating a new sort of racial division, without any new technology being needed. Potential technology would accelerate this process.
>Has anyone written a direct response to Sasha's sources (he claims 15% heritable) that you think is best?
Not an exhaustive critique.
>whites outperforming Asians in Washington DC
I assume cases like those are due to selective migration.
>On the other hand, if there was a 10 point IQ gap between individuals, could that level of variation be possible? For example: two people take IQ tests at random times. How much does IQ vary per test (maybe due to bad sleep or an emotionally traumatic event like an argument or relationship breakup), what are the chances that a 10 point IQ gap would close?
The test-retest correlation of IQ tests is lower than the internal reliability, which is not what is usually observed for most psychological tests, where the internal reliability is lower than the test-retest correlation. I assume most of that fluctuation is biological noise like mood/sleep/eating patterns/thirst and whatnot.
>The basic idea is if average people adopt Seoul fertility (0.5), and elites have a TFR of 5 (I have to explain how that could happen)
Elites having a TFR of 5 would be interesting. Would have to be some kind of cultural change I assume, as the genes/money will stay the same.
The biggest legal aspect for elite fertility seems to be divorce and alimony law. Bezos seemed unable to out-lawyer his wife. Abolish marriage and alimony, replace with individual contracts, and TFR goes up for rich people, is my current theory. On the other hand, maybe it decreases for the middle class because without divorce and alimony law women will be afraid to have children and feel less stable.
Not to be persnickety, but I think too many people over-index on heritability estimates: what they are, how large they are, whether they're extrapolate across populations.
Heritability estimates are population specific and don't necessarily represent some fixed reality. Heritability doesn't tell us how innate a trait is (e.g. heritability of having five fingers is ~0 but it's wholly innate). Heritability doesn't tell us how modifiable a trait is. It's just the portion of the variance in a population explained by genetic variation.
Additionally, one of the more eminent population geneticists (David Reich) openly acknowledges that the likelihood of population differences across many traits is highly likely. It's just unlikely that these are large differences on average and that they accord with common or racist stereotypes.
I don't find that persnickety. Would love to interview you to give your thoughts a fair shake.
Of course, I'm happy to talk with you about it (can take to DMs to sort out). I agree with many of Dr. Gusev's critiques but not necessarily with all his conclusions about the genetic contribution of neuro/psych traits. So it's a middle road position which I think is often tacitly held by many academics in these fields.
It's heritable but not as heritable as some think. Still, that position is unacceptable to the prevailing mainstream narrative.
The heritability of intelligence is hard to estimate (everything is modeling, at the end of the day), but the largest GCTA estimate corrected for attenuation is still 48% (https://gwern.net/embryo-selection).
Has anyone written a direct response to Sasha's sources (he claims 15% heritable) that you think is best?
Some less relevant thoughts:
Is it possible that intelligence gaps are 48% heritable when comparing individuals, but that racial gaps are more environmental? I was looking at some SAT data that showed whites outperforming Asians in Washington DC, and Hispanics performing at the same level as whites in West Virginia, and blacks sometimes performing better than Pacific Islanders and Native Americans, and sometimes worse.
On the other hand, if there was a 10 point IQ gap between individuals, could that level of variation be possible? For example: two people take IQ tests at random times. How much does IQ vary per test (maybe due to bad sleep or an emotionally traumatic event like an argument or relationship breakup), what are the chances that a 10 point IQ gap would close? I would expect large groups like races to be less subject to "bad days" than individuals.
I'm guessing the hereditarian position would be "West Virginia whites are subject to brain drain and dysgenics, or environmental toxins from coal mining." My question would be, how quickly did the WV-white vs average white gap form? Was that due to a founder effect from England (Albion's Seed) in 1600? Or the result of de-industrialization since 1970? If we can quantify the rate of change, then we could say that IQ can be increased or decreased at that rate. Perhaps as slow as 1 point per 100 years, or as quickly as 1 point per 10 years. That would make racial gaps a question of timescale -- a range between 150 to 1500 years.
I'm working on an article expanding on the idea of medieval population replacement of non-elites by elites, and projecting that into the future. The basic idea is if average people adopt Seoul fertility (0.5), and elites have a TFR of 5 (I have to explain how that could happen), then elites would genetically replace the population within a few centuries, creating a new sort of racial division, without any new technology being needed. Potential technology would accelerate this process.
>Has anyone written a direct response to Sasha's sources (he claims 15% heritable) that you think is best?
Not an exhaustive critique.
>whites outperforming Asians in Washington DC
I assume cases like those are due to selective migration.
>On the other hand, if there was a 10 point IQ gap between individuals, could that level of variation be possible? For example: two people take IQ tests at random times. How much does IQ vary per test (maybe due to bad sleep or an emotionally traumatic event like an argument or relationship breakup), what are the chances that a 10 point IQ gap would close?
The test-retest correlation of IQ tests is lower than the internal reliability, which is not what is usually observed for most psychological tests, where the internal reliability is lower than the test-retest correlation. I assume most of that fluctuation is biological noise like mood/sleep/eating patterns/thirst and whatnot.
>The basic idea is if average people adopt Seoul fertility (0.5), and elites have a TFR of 5 (I have to explain how that could happen)
Elites having a TFR of 5 would be interesting. Would have to be some kind of cultural change I assume, as the genes/money will stay the same.
The biggest legal aspect for elite fertility seems to be divorce and alimony law. Bezos seemed unable to out-lawyer his wife. Abolish marriage and alimony, replace with individual contracts, and TFR goes up for rich people, is my current theory. On the other hand, maybe it decreases for the middle class because without divorce and alimony law women will be afraid to have children and feel less stable.