That's good news, right? They aren't denying the findings.
Anyway, this is STILL not *direct* evidence of the HBD hypothesis, as it's possible a priori that other populations underwent similar amounts of selection. Still, it's quite suggestive.
Right. There remain a lot of potential possibilities, some more far-fetched than others, but this kind of study could have falsified the HBD hypothesis.
By "HBD hypothesis" I was thinking something like "BW IQ gap in US is >50% genetic". For this to be false, black people would need to have undergone similar amounts of selection.
If by "HBD hypothesis" you mean the 10,000 year explosion, that humans have genetically changed a lot in the last 10,000 years, then yes I agree with you.
I’m guessing humanity is going to evolve rapidly (both genetically and socially) in the age of the internet for several reasons: (a) mating preferences have changed due to vastly more available partners, (b) mimesis has resulted in everybody wanting an Instagram model, 6-6-6 dudes, etc. with the girl or boy next door no longer being good enough, (c) data shows that Millennials and after (at least in the US) have sex, get married and have kids at substantially lower rates than previous generations, and (d) on the left there is a lot of pressure not to have kids at all because wE’rE KiLLiNg the PlAnET!!111. I am guessing from a genetic standpoint people going forward will be better looking and taller on average at a minimum since those are the people who apparently get laid these days, and from a social standpoint, presumably the types of people who are still having lots of kids (religious types, many immigrant groups, e.g.) will come to make up larger and larger parts of the gene pool and their social views will be more influential as they get passed down to their kids.
Oh also, I suspect that the widespread and instantaneous access to porn makes jerking off a much more accessible substitute for sex (like a less inferior good in an economic sense) than it was back when it was a big deal to find your dad’s collection of 3 dog eared Playboys and a Penthouse if you were lucky (speaking mostly about men here). It seems like a mathematical certainty that there is some set of men who would have been marginally likely to reproduce in the days before easy access to porn who won’t reproduce these days because in addition to the factors outlined above, it’s not worth the effort of trying to get laid in light of instant access to self-gratification. So that set of genes will also tend to disappear from the gene pool.
There is so much to refute in that argument that I don't know where to start. "Vastly more mating opportunities because of the internet"? Really? Vastly more access to the highest quality porn for sure, but mating opportunities? Have you ever gone on any of those "dating sites"? "A prettier and brighter population to come"? Islam is the fastest growing religion because Muslims are the fastest breeding culture. Islam suppresses intelligence among its followers in many ways. The way to bet is for the future to be dumber, darker and less fulfilling assuming that we don't start WWIII along the way.
If you don’t think there are vastly more mating opportunities now than there were in the days when your entire social circle consisted of people you knew within a 10 mile radius, I don’t know what to tell you.
In the first place, my circle of friends and potential mates has a radius of about 100 miles not 10. In the second place if you think the opportunities to find a mate have improved, I can't believe you are actually looking for one. You have about a 1% chance of hooking up with anyone on the web if you make less than $250,000 per year (or can fake it really well) and less than 1% of those hookups will be with anyone you would seriously consider as a mate. The much quoted statistic that 80% of the women online are only interested in 20% of the men totally obscures the fact 99.9% of the people of either sex online are not worth doing anything but masterbating to. Grow up, get real and immerse yourself in online dating culture before you talk so positively about something you seem to have no real life experience with.
Great, so now in addition to people that you had access to before, now you also have anyone who you might meet through the Internet. It is a mathematical certainty that the set of the people you have access to without the Internet plus the people you have access to with the Internet is greater than just the set of people you have access to without the Internet. You are wrong.
Just how old are you? You seem to think that the internet is the real world and netizens are real people. When was the last time you met someone from the internet face to face? How did it go? Ever hooked up with anyone you met online? Were they mate material?
This is a bit of an illusion because the same communication/transportation/social liberation/whatever changes that make 'more' partners available to you also make those same partners more available to everyone else. So the appearance of increased supply is offset by the reality of equally increased demand. At the end of the day, mating is a zero-sum game, perhaps the ultimate zero-sum game: no one can have more without someone else having less.
How the apparently widened horizons of everyone's mating prospects will work out is not known for sure yet, but a few outlines are appearing.
First is that the "more available partners" are not really available to everyone, but only to the righthand side of the each sex's attractiveness bell curve, especially for males. So unless polygamy (polygyny in practice) becomes legitimized and institutionalized, the pseudo-increased partner availability will not lead to more beautiful/intelligent/athletic/attractive people in the future. It might lead to only the beautiful/intelligent/athletic/attractive people reproducing, which would lead to about the same number of beautiful/intelligent/athletic/attractive people in the future, while the number of ugly/stupid/inept/repellant people decrease, so there may be a relative improvement though not an absolute one. There is some evidence this may be happening among First World whites and Asians, though it is too soon to say definitively. Unfortunately there is at least as much evidence that everyone else is procreating in a manner not at all calculated to maximize beauty/intelligence/athleticism/attractiveness, and worse, that the space ceded by the shrinkage of First World whites and Asians is being snaffled up by the expansion of Third World dependent populations.
So pessimism is justifiable, while optimisms needs to explain itself.
I'm reading "Who We are and How We got Here" right now. I read "The 10,000 Year Revolution" when it came out. The more we learn definitely about the past the more interesting the story becomes.
Re "Who We are and How We got Here" -- My first recommendation to interested people without (or with) a background in genetics. The first three-quarters of the book is superb, and still timely (though published in 2018) because it gives a fine sense of what's important to those who study ancient DNA, what tools they use, and how they think.
Reich and his wife/co-author then pivot, devoting the final chapters to attempting the cancellation of retired NYT science writer Nicholas Wade for thoughtcrime. Unseemly behavior for a great scientist, but Many Such Cases.
Did you ever hear of the New Testament fact given in 2nd Peter, chapter 3, verse 8. It is God's Standard Time as it relates to this present earth-epic: "But, beloved, be not ignorant of this one thing, that one day is with the Lord as a thousand years, and a thousand years as one day."
Going by that Standard, the Creation saga detailed in Genesis chapters one and two entails a total of 8,000 years, including God's Day of Rest. Time elapsed so far on the Hebrew calendar: approx 5784 years. The Hebrew calendar begins with the creation of eth'ha adam of Genesis chapter 2, verse 7. So. Total elapsed time since God said, "Let there be light:" Approx 13,784 years.
That's easily within the time-frame of the study outlined in the Abstract. Like the old saying say, "A coincidence is God's Way of saying hello." What a time to be alive!
Andrew Sullivan discussed Reich’s results today in his Substack, The Weekly Dish. The title of the article is “The Use and Abuse of Science.” He is pretty rough on Gould, calling him “not only wrong , … but grotesquely so” and later referring to “the delusions and falsehoods of Gould et al.”
This isn't rocket science, and it's something that even I (a mediocre novice re anything to do with genetics & evolution) worked out ("noticed") as a kid. Organisms evolve based on the environment they happen to be in. Humans evolved differently as they spread, and adapted to different environments. This in turn drove the culture of the different human groups- and then that culture then drove further evolution/adaption. Isn't it obvious? An example? "Civilization" (lots of people living together and cooperating in the last 10k years, for whatever reason) meant that laws and morals had to be developed - and those that didn't abide didn't get to pass on their genes a lot of the time. Why is this considered such a breakthrough idea? I guess because now we can prove it?
Because certain people have been allowed to live who wouldn't have in the past. And the anti-science crowd contain, or at least support those groups that disproportionately contain those individuals.
Steve -- I believe you understand Reich's methodology and findings, but you need to do a better job of translating these into plain English for your readers. Just regurgitating the gobbledygook abstract is low energy. If you can't explain it in plain English why would you expect the biased MSM make the effort. Anyway, I'm not trying to be a dick, but just offering a constructive note.
I think you did a good job in that comment, but I think the fact that most people don’t understand statistics puts both you and Steve at a disadvantage.
It's a very approximate measure, because gene markers so far let you predict just a modest part of the IQ variance. If you have like 100k marker sites for DNA variations - which is like the current tool box - the best you can do is to test for linear effects. You can't do combinations because it will explode the number of parameters tested in the multivariate regression analysis. You might even have some problems if you want to test heterozygosity vs homozygosity as a screen for all 100k markers. For example, what if you have a gene X in its two gene copies - and one copy has marker A at the gamma site and the other copy marker B at the delta site, but their phenotypes - if combined as AB - have a similar cellular effect as homozygosity for one of them like AA or BB?
> "Why no news coverage yet of David Reich's epochal paper?"
MSM journalists are kind of dim and don't understand any actual science. As you surmise, though, at some point they will realize that this paper has implications for "RACISM!1!!1!!", and then they will address it within the very narrow arc that their wokeness blinders allow, so prepared to be acutely disappointed by their response.
I tried posting this paper on an anthropology sub on Reddit and it was removed almost immediately with the mods claiming it was flawed and "uncited". All my comments attempting to explain it were downvoted to oblivion. It was obvious that not a single one of them actually bothered to read it. I wonder if they have any idea who David Reich is.
Also, the paper doesn't actually test on "West Eurasians" as a whole but specifically Europeans despite them using the term West Eurasian. I think there would be noticeable differences between Europeans and Middle Easterners in selection for those traits, so I wish they'd just been more specific and said European anyway. They made the title very vague too. Both of these decisions were probably motivated by a desire to mitigate controversy to some extent. Political correctness...am I right?
I find it quite odd that the people trying to suppress this information the hardest are white people. I get why, but it just feels weird to me as a non-white person.
That's not my impression at all. A lot of "white apologists" to be sure but a lot of the non whites they influence as well. As an example, most of Thomas Sowell's critics are people of color from the Race Industrial Complex in this country.
Similar to how women view other women, no one hates whites more than other whites (and white-adjacents, e.g., light-skinned Jews, Hispanics, S. Italians, S. Greeks, et al.).
> "Jim Simon’s Renaissance Technologies, that rare hedge fund that doesn’t get rich charging clients fees (in fact it has very few clients because why should the employees share their genius with clients?)."
Instead it gets rich by snatching money from regular people investors just trying to save for retirement. At least the clients agreed to the fees. The regular investors don't realize that part of their trading cost is Jim Simon's invisible tollbooth and obviously they wouldn't agree if they understood it.
I think the "problem" is that Steve and his audience are emphasizing something that isn't the main finding of the paper (or, arguably, even a finding of the paper) which is that the genes for intelligence have clearly evolved through selection over the past 10K years. Further, (I can't recall if Steve stated or merely implied this) that this is a breakthrough because previously we thought 10k years wasn't enough time.
That latter idea appears to be wrong based on the first reference above from 2010. The authors clearly state that we know evolution has occurred on that time frame and they give several examples.
The real finding of this paper is that they created a new methodology to measure evolutionary selection over that time period, and that their method works, and is going to be a very promising research tool going forward.
Someone below asked for a human readable explanation of what they did. I'm not the best person to try to explain or simplify it and I don't know if anyone likes long comments on someone else's blog. I did spend the morning trying to figure it out so I will share. This could all be wrong. I am not an evolutionary biologist, though i have a strong molecular bio background.
So here goes:
Scientists have long thought that human genes and phenotypes evolved based on selective pressures over the past 100K years and even over the past 10k years. Previously they tried to show and measure this selection by looking at differences between current populations; what the Reich paper calls "being trapped in the present and studying the scars left by selection on the genomes of descendants".
The ancient DNA he works on provided the opportunity to measure the changes in gene frequency over time, and then decide if the rate of change indicates selection or is explainable by randomness.
There is the further issue of what you measure to prove selection. Per the first paper I link above previous work focused on the dominant hypothesis of population biology, the "Hard Sweep". I think this means that you have a gene that exists in a population at a certain frequency relative to its variants, and that over time that relative frequency changes, and stabilizes into a new equilibrium.
So let's say there was a gene that coded for a protein called smartin. If you have the regular smartin (let's call it smartinA) you have an IQ of 105. If you are fortunate enough to have a slightly different copy of smartin gene smartinB (with one or two base pairs different) then you are going to have an IQ of 120. The earlier methodology see that the relative frequency of smartinA vs smartinB was 99 to 1 at the dawn of civilization and now it's 5 to 1. Then you do some kind of math to show that it's extremely unlikely this would happen without selection for the trait (smartness).
Unfortunately, they have found very few genes that have been selected to equilibrium this way. The paper I linked above also mentions "soft sweeps", which I will ignore for now, and polygenic adaptation. Many more traits like height and intelligence are going to be due to polygenic adaptation than "hard sweeps" of single genes.
The breakthrough of this paper, near as I can tell, is they took advantage of increasing available ancient genomes (sample size) and used a method to show that the polygenic scores were changing over time in a way that indicated they were slowly being selected for. That is, they did not have to wait for the polygenic scores to achieve equilibrium. They just had to show that the scores started changing in a direction and kept on changing in that direction. They included some sophisticated data cleaning techniques (which I have no desire to try to critique so I just trust them that they nailed those :))
To test their technique they tried it out on polygenic scores that a betting man would assume were selected for over the past 10K years like intelligence. They showed that the polygenic scores were indeed slowly changing in a particular direction (up to whatever extent it makes sense to say that about a polygenic score) over a long period of time, consistently and this, coupled with their math techniques mean those polygenic scores were changing due to selection, not randomness.
Is this consistent with the intuition of pretty much everyone on this comment section, that genes for intelligence were selected for higher intelligence especially since the dawn of agriculture/civilisation? Absolutely!
Does it prove it? Not by a long shot. If you want to discuss why I say that, we can get into it, but I think I've blathered long enough and I'm not confident I clarified anything (or that I am even right, although I did read the references twice)
Anyway I think that might be part of why it hasn't gotten a lot of mainstream media attention. It's a very cool technique that promises a lot of advances in the field. I'm sure evolutionary biologists will be excited about it but it doesn't have a definite exciting conclusion for the average reader just yet.
Thanks for the 2nd paper, the Nature one. The surrogate for intelligence - school achievement at 16 - had a polygenic score explaining 15% of variance. Now, for each individual alone, the predictive effect would be modest - 15% of the variance. However, if you could do averaging across 400 individuals, you could start getting very robust effects. Because your confidence intervals would be shrinking with the square root of n individuals...
Of course you would still have to prove that the later population B arose from the earlier population A. If B are not the descendents of A you are not dealing with selection inside the same population pool.
I don't understand your point. Do you mean 400 individual SNPs or people? If it's people, the prediction of 15% of variance would have had to be done at the population level, so I don't see how you propose to increase the predictive effect by averaging it again over a small population. Always willing to assume I am missing something
The polygenic score is BUILT on the SNPs they have analysed, but it is just the starting point.
I meant people. Confidence intervals (CI) are derived from variance (V) and the number of people (n). So 2SD Confidence intervals (95%) are about = +/- 2* sqrt(V/N). The number of people analyzed in a sample always plays a critical role. The whole point about CIs is that it informs you about where the "real average based on infinite number of data points" could lie when your finite data point average is where it is.
The reason I am going so poetic about confidence intervals is because no-one is "really interested" in polygenic score X. What everyone is interested is how changes in that score X impact the average IQ. And for that, you need to know the confidence intervals that are based on the sample size...
I think I understand. You mean the relevant question is "if I know your polygenic score is X and that means your IQ is 115, what is the plus or minus for that? Are we 95% confident your IQ is between 110 and 120 or are we 95% confident it's between 100 and 130?
Exactly. Even though, I would guess with the polygenic score moved from like 80 to 100, that would mean maybe - with a 15% of variation (sqrt(15%)*20=8) explained - there is an expected increase in 8 basis points on the IQ score. So the question is how confident can I be that there is an increment in the IQ score* above 0? Hence the question on confidence intervals. If there are enough individuals analysed, possible that the 2SD confidence intervals will be less than 6 or so. And then you are in the clear, with 2SD confident IQ improvement between 2 and 14 basis points...
Of course, for the score to move from 80 to 100 is a humongous effect in itself...
* I forgot of course - we are not talking about IQ, we are talking about scholastic achievement at 16. The reason we are talking about scholastic achievement at 16 and not IQ is that these data are easy to plug in from millions of individuals without raising some anti-racist eye brows.
The Razib Khan podcast you linked is very helpful because he explains how top-rated the authors' reputations are. For example, Nick Patterson is a genius mathematician, and the ability of Alkes Price approaches 'Von Neuman level, not of this world.' So the findings are probably impeccable, and the the-pro mediocrity establishment will have no choice but to ignore them.
That's good news, right? They aren't denying the findings.
Anyway, this is STILL not *direct* evidence of the HBD hypothesis, as it's possible a priori that other populations underwent similar amounts of selection. Still, it's quite suggestive.
Right. There remain a lot of potential possibilities, some more far-fetched than others, but this kind of study could have falsified the HBD hypothesis.
Right, like any good scientific hypothesis, HBD is falsifiable.
It seems to me that other populations having undergone similar amounts of selection would be a strong argument *in favor of* the HBD hypothesis.
By "HBD hypothesis" I was thinking something like "BW IQ gap in US is >50% genetic". For this to be false, black people would need to have undergone similar amounts of selection.
If by "HBD hypothesis" you mean the 10,000 year explosion, that humans have genetically changed a lot in the last 10,000 years, then yes I agree with you.
I’m guessing humanity is going to evolve rapidly (both genetically and socially) in the age of the internet for several reasons: (a) mating preferences have changed due to vastly more available partners, (b) mimesis has resulted in everybody wanting an Instagram model, 6-6-6 dudes, etc. with the girl or boy next door no longer being good enough, (c) data shows that Millennials and after (at least in the US) have sex, get married and have kids at substantially lower rates than previous generations, and (d) on the left there is a lot of pressure not to have kids at all because wE’rE KiLLiNg the PlAnET!!111. I am guessing from a genetic standpoint people going forward will be better looking and taller on average at a minimum since those are the people who apparently get laid these days, and from a social standpoint, presumably the types of people who are still having lots of kids (religious types, many immigrant groups, e.g.) will come to make up larger and larger parts of the gene pool and their social views will be more influential as they get passed down to their kids.
Oh also, I suspect that the widespread and instantaneous access to porn makes jerking off a much more accessible substitute for sex (like a less inferior good in an economic sense) than it was back when it was a big deal to find your dad’s collection of 3 dog eared Playboys and a Penthouse if you were lucky (speaking mostly about men here). It seems like a mathematical certainty that there is some set of men who would have been marginally likely to reproduce in the days before easy access to porn who won’t reproduce these days because in addition to the factors outlined above, it’s not worth the effort of trying to get laid in light of instant access to self-gratification. So that set of genes will also tend to disappear from the gene pool.
There is so much to refute in that argument that I don't know where to start. "Vastly more mating opportunities because of the internet"? Really? Vastly more access to the highest quality porn for sure, but mating opportunities? Have you ever gone on any of those "dating sites"? "A prettier and brighter population to come"? Islam is the fastest growing religion because Muslims are the fastest breeding culture. Islam suppresses intelligence among its followers in many ways. The way to bet is for the future to be dumber, darker and less fulfilling assuming that we don't start WWIII along the way.
If you don’t think there are vastly more mating opportunities now than there were in the days when your entire social circle consisted of people you knew within a 10 mile radius, I don’t know what to tell you.
In the first place, my circle of friends and potential mates has a radius of about 100 miles not 10. In the second place if you think the opportunities to find a mate have improved, I can't believe you are actually looking for one. You have about a 1% chance of hooking up with anyone on the web if you make less than $250,000 per year (or can fake it really well) and less than 1% of those hookups will be with anyone you would seriously consider as a mate. The much quoted statistic that 80% of the women online are only interested in 20% of the men totally obscures the fact 99.9% of the people of either sex online are not worth doing anything but masterbating to. Grow up, get real and immerse yourself in online dating culture before you talk so positively about something you seem to have no real life experience with.
Great, so now in addition to people that you had access to before, now you also have anyone who you might meet through the Internet. It is a mathematical certainty that the set of the people you have access to without the Internet plus the people you have access to with the Internet is greater than just the set of people you have access to without the Internet. You are wrong.
Just how old are you? You seem to think that the internet is the real world and netizens are real people. When was the last time you met someone from the internet face to face? How did it go? Ever hooked up with anyone you met online? Were they mate material?
LOL, if simple math is too much for you I see no point in continuing this discussion. You are factually mistaken and there are no two ways around it.
> "vastly more available partners"
This is a bit of an illusion because the same communication/transportation/social liberation/whatever changes that make 'more' partners available to you also make those same partners more available to everyone else. So the appearance of increased supply is offset by the reality of equally increased demand. At the end of the day, mating is a zero-sum game, perhaps the ultimate zero-sum game: no one can have more without someone else having less.
How the apparently widened horizons of everyone's mating prospects will work out is not known for sure yet, but a few outlines are appearing.
First is that the "more available partners" are not really available to everyone, but only to the righthand side of the each sex's attractiveness bell curve, especially for males. So unless polygamy (polygyny in practice) becomes legitimized and institutionalized, the pseudo-increased partner availability will not lead to more beautiful/intelligent/athletic/attractive people in the future. It might lead to only the beautiful/intelligent/athletic/attractive people reproducing, which would lead to about the same number of beautiful/intelligent/athletic/attractive people in the future, while the number of ugly/stupid/inept/repellant people decrease, so there may be a relative improvement though not an absolute one. There is some evidence this may be happening among First World whites and Asians, though it is too soon to say definitively. Unfortunately there is at least as much evidence that everyone else is procreating in a manner not at all calculated to maximize beauty/intelligence/athleticism/attractiveness, and worse, that the space ceded by the shrinkage of First World whites and Asians is being snaffled up by the expansion of Third World dependent populations.
So pessimism is justifiable, while optimisms needs to explain itself.
I'm reading "Who We are and How We got Here" right now. I read "The 10,000 Year Revolution" when it came out. The more we learn definitely about the past the more interesting the story becomes.
I saw Reich talk his book at Harvard four yrs ago. I'm really looking forward to his next one Whenever it comes out.
Re "Who We are and How We got Here" -- My first recommendation to interested people without (or with) a background in genetics. The first three-quarters of the book is superb, and still timely (though published in 2018) because it gives a fine sense of what's important to those who study ancient DNA, what tools they use, and how they think.
Reich and his wife/co-author then pivot, devoting the final chapters to attempting the cancellation of retired NYT science writer Nicholas Wade for thoughtcrime. Unseemly behavior for a great scientist, but Many Such Cases.
Thanks for warning me. I'm only halfway through the book.
Did you ever hear of the New Testament fact given in 2nd Peter, chapter 3, verse 8. It is God's Standard Time as it relates to this present earth-epic: "But, beloved, be not ignorant of this one thing, that one day is with the Lord as a thousand years, and a thousand years as one day."
Going by that Standard, the Creation saga detailed in Genesis chapters one and two entails a total of 8,000 years, including God's Day of Rest. Time elapsed so far on the Hebrew calendar: approx 5784 years. The Hebrew calendar begins with the creation of eth'ha adam of Genesis chapter 2, verse 7. So. Total elapsed time since God said, "Let there be light:" Approx 13,784 years.
That's easily within the time-frame of the study outlined in the Abstract. Like the old saying say, "A coincidence is God's Way of saying hello." What a time to be alive!
Andrew Sullivan discussed Reich’s results today in his Substack, The Weekly Dish. The title of the article is “The Use and Abuse of Science.” He is pretty rough on Gould, calling him “not only wrong , … but grotesquely so” and later referring to “the delusions and falsehoods of Gould et al.”
This isn't rocket science, and it's something that even I (a mediocre novice re anything to do with genetics & evolution) worked out ("noticed") as a kid. Organisms evolve based on the environment they happen to be in. Humans evolved differently as they spread, and adapted to different environments. This in turn drove the culture of the different human groups- and then that culture then drove further evolution/adaption. Isn't it obvious? An example? "Civilization" (lots of people living together and cooperating in the last 10k years, for whatever reason) meant that laws and morals had to be developed - and those that didn't abide didn't get to pass on their genes a lot of the time. Why is this considered such a breakthrough idea? I guess because now we can prove it?
Because certain people have been allowed to live who wouldn't have in the past. And the anti-science crowd contain, or at least support those groups that disproportionately contain those individuals.
Steve -- I believe you understand Reich's methodology and findings, but you need to do a better job of translating these into plain English for your readers. Just regurgitating the gobbledygook abstract is low energy. If you can't explain it in plain English why would you expect the biased MSM make the effort. Anyway, I'm not trying to be a dick, but just offering a constructive note.
I think Steve knows his audience extremely well.
I gave it a shot in a top level comment.
I think you did a good job in that comment, but I think the fact that most people don’t understand statistics puts both you and Steve at a disadvantage.
The MSM was put off by their use of "leverage" as a verb in the first sentence.
"increased measures related to cognitive performance" Is this an obscure way of saying IQ rose over time?
It's a very approximate measure, because gene markers so far let you predict just a modest part of the IQ variance. If you have like 100k marker sites for DNA variations - which is like the current tool box - the best you can do is to test for linear effects. You can't do combinations because it will explode the number of parameters tested in the multivariate regression analysis. You might even have some problems if you want to test heterozygosity vs homozygosity as a screen for all 100k markers. For example, what if you have a gene X in its two gene copies - and one copy has marker A at the gamma site and the other copy marker B at the delta site, but their phenotypes - if combined as AB - have a similar cellular effect as homozygosity for one of them like AA or BB?
No media until Reich proves Africans got even smarter!
> "Why no news coverage yet of David Reich's epochal paper?"
MSM journalists are kind of dim and don't understand any actual science. As you surmise, though, at some point they will realize that this paper has implications for "RACISM!1!!1!!", and then they will address it within the very narrow arc that their wokeness blinders allow, so prepared to be acutely disappointed by their response.
I like Reich's paper because it supports my hypothesis that phenotypic change is always first and foremost driven by behavioral change.
One might say that environmental change drives behavioral change, but environment is not always necessary or sufficient to do so.
When it comes down to it, life becomes what it does, and humans are no exception. I like this idea because it puts genes in their proper perspective:
Not as deterministic agents in their own right, but rather as dutiful servants of the will to live.
I tried posting this paper on an anthropology sub on Reddit and it was removed almost immediately with the mods claiming it was flawed and "uncited". All my comments attempting to explain it were downvoted to oblivion. It was obvious that not a single one of them actually bothered to read it. I wonder if they have any idea who David Reich is.
Also, the paper doesn't actually test on "West Eurasians" as a whole but specifically Europeans despite them using the term West Eurasian. I think there would be noticeable differences between Europeans and Middle Easterners in selection for those traits, so I wish they'd just been more specific and said European anyway. They made the title very vague too. Both of these decisions were probably motivated by a desire to mitigate controversy to some extent. Political correctness...am I right?
Or was "Eurasian" a nod to many of the co-authors?
It's a little odd that we're the race being studied by others--Planet of the Asians.
I find it quite odd that the people trying to suppress this information the hardest are white people. I get why, but it just feels weird to me as a non-white person.
That's not my impression at all. A lot of "white apologists" to be sure but a lot of the non whites they influence as well. As an example, most of Thomas Sowell's critics are people of color from the Race Industrial Complex in this country.
Similar to how women view other women, no one hates whites more than other whites (and white-adjacents, e.g., light-skinned Jews, Hispanics, S. Italians, S. Greeks, et al.).
Eurasian is because back 10k years a big proportion of genes that flowed into the European population were still hovering over the Asian steppe.
Nothing is so deniable as a fact plain in front of one's face.
> "Jim Simon’s Renaissance Technologies, that rare hedge fund that doesn’t get rich charging clients fees (in fact it has very few clients because why should the employees share their genius with clients?)."
Instead it gets rich by snatching money from regular people investors just trying to save for retirement. At least the clients agreed to the fees. The regular investors don't realize that part of their trading cost is Jim Simon's invisible tollbooth and obviously they wouldn't agree if they understood it.
https://youtu.be/xkbdZb0UPac?t=710
I've given the paper a second read and also, for background the following: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2994553/ (from 2010) and
https://www.nature.com/articles/s41380-021-01348-y (from 2021, to remind me where we we're WRT Polygenic scores).
I think the "problem" is that Steve and his audience are emphasizing something that isn't the main finding of the paper (or, arguably, even a finding of the paper) which is that the genes for intelligence have clearly evolved through selection over the past 10K years. Further, (I can't recall if Steve stated or merely implied this) that this is a breakthrough because previously we thought 10k years wasn't enough time.
That latter idea appears to be wrong based on the first reference above from 2010. The authors clearly state that we know evolution has occurred on that time frame and they give several examples.
The real finding of this paper is that they created a new methodology to measure evolutionary selection over that time period, and that their method works, and is going to be a very promising research tool going forward.
Someone below asked for a human readable explanation of what they did. I'm not the best person to try to explain or simplify it and I don't know if anyone likes long comments on someone else's blog. I did spend the morning trying to figure it out so I will share. This could all be wrong. I am not an evolutionary biologist, though i have a strong molecular bio background.
So here goes:
Scientists have long thought that human genes and phenotypes evolved based on selective pressures over the past 100K years and even over the past 10k years. Previously they tried to show and measure this selection by looking at differences between current populations; what the Reich paper calls "being trapped in the present and studying the scars left by selection on the genomes of descendants".
The ancient DNA he works on provided the opportunity to measure the changes in gene frequency over time, and then decide if the rate of change indicates selection or is explainable by randomness.
There is the further issue of what you measure to prove selection. Per the first paper I link above previous work focused on the dominant hypothesis of population biology, the "Hard Sweep". I think this means that you have a gene that exists in a population at a certain frequency relative to its variants, and that over time that relative frequency changes, and stabilizes into a new equilibrium.
So let's say there was a gene that coded for a protein called smartin. If you have the regular smartin (let's call it smartinA) you have an IQ of 105. If you are fortunate enough to have a slightly different copy of smartin gene smartinB (with one or two base pairs different) then you are going to have an IQ of 120. The earlier methodology see that the relative frequency of smartinA vs smartinB was 99 to 1 at the dawn of civilization and now it's 5 to 1. Then you do some kind of math to show that it's extremely unlikely this would happen without selection for the trait (smartness).
Unfortunately, they have found very few genes that have been selected to equilibrium this way. The paper I linked above also mentions "soft sweeps", which I will ignore for now, and polygenic adaptation. Many more traits like height and intelligence are going to be due to polygenic adaptation than "hard sweeps" of single genes.
The breakthrough of this paper, near as I can tell, is they took advantage of increasing available ancient genomes (sample size) and used a method to show that the polygenic scores were changing over time in a way that indicated they were slowly being selected for. That is, they did not have to wait for the polygenic scores to achieve equilibrium. They just had to show that the scores started changing in a direction and kept on changing in that direction. They included some sophisticated data cleaning techniques (which I have no desire to try to critique so I just trust them that they nailed those :))
To test their technique they tried it out on polygenic scores that a betting man would assume were selected for over the past 10K years like intelligence. They showed that the polygenic scores were indeed slowly changing in a particular direction (up to whatever extent it makes sense to say that about a polygenic score) over a long period of time, consistently and this, coupled with their math techniques mean those polygenic scores were changing due to selection, not randomness.
Is this consistent with the intuition of pretty much everyone on this comment section, that genes for intelligence were selected for higher intelligence especially since the dawn of agriculture/civilisation? Absolutely!
Does it prove it? Not by a long shot. If you want to discuss why I say that, we can get into it, but I think I've blathered long enough and I'm not confident I clarified anything (or that I am even right, although I did read the references twice)
Anyway I think that might be part of why it hasn't gotten a lot of mainstream media attention. It's a very cool technique that promises a lot of advances in the field. I'm sure evolutionary biologists will be excited about it but it doesn't have a definite exciting conclusion for the average reader just yet.
Thanks for the 2nd paper, the Nature one. The surrogate for intelligence - school achievement at 16 - had a polygenic score explaining 15% of variance. Now, for each individual alone, the predictive effect would be modest - 15% of the variance. However, if you could do averaging across 400 individuals, you could start getting very robust effects. Because your confidence intervals would be shrinking with the square root of n individuals...
Of course you would still have to prove that the later population B arose from the earlier population A. If B are not the descendents of A you are not dealing with selection inside the same population pool.
I don't understand your point. Do you mean 400 individual SNPs or people? If it's people, the prediction of 15% of variance would have had to be done at the population level, so I don't see how you propose to increase the predictive effect by averaging it again over a small population. Always willing to assume I am missing something
The polygenic score is BUILT on the SNPs they have analysed, but it is just the starting point.
I meant people. Confidence intervals (CI) are derived from variance (V) and the number of people (n). So 2SD Confidence intervals (95%) are about = +/- 2* sqrt(V/N). The number of people analyzed in a sample always plays a critical role. The whole point about CIs is that it informs you about where the "real average based on infinite number of data points" could lie when your finite data point average is where it is.
The reason I am going so poetic about confidence intervals is because no-one is "really interested" in polygenic score X. What everyone is interested is how changes in that score X impact the average IQ. And for that, you need to know the confidence intervals that are based on the sample size...
I think I understand. You mean the relevant question is "if I know your polygenic score is X and that means your IQ is 115, what is the plus or minus for that? Are we 95% confident your IQ is between 110 and 120 or are we 95% confident it's between 100 and 130?
Exactly. Even though, I would guess with the polygenic score moved from like 80 to 100, that would mean maybe - with a 15% of variation (sqrt(15%)*20=8) explained - there is an expected increase in 8 basis points on the IQ score. So the question is how confident can I be that there is an increment in the IQ score* above 0? Hence the question on confidence intervals. If there are enough individuals analysed, possible that the 2SD confidence intervals will be less than 6 or so. And then you are in the clear, with 2SD confident IQ improvement between 2 and 14 basis points...
Of course, for the score to move from 80 to 100 is a humongous effect in itself...
* I forgot of course - we are not talking about IQ, we are talking about scholastic achievement at 16. The reason we are talking about scholastic achievement at 16 and not IQ is that these data are easy to plug in from millions of individuals without raising some anti-racist eye brows.
The Razib Khan podcast you linked is very helpful because he explains how top-rated the authors' reputations are. For example, Nick Patterson is a genius mathematician, and the ability of Alkes Price approaches 'Von Neuman level, not of this world.' So the findings are probably impeccable, and the the-pro mediocrity establishment will have no choice but to ignore them.