”Why do kids look and act like their parents?” I always felt that they kind of don’t? That they ought too look more alike, that the ought to inherit more traits and more cultural behavior - if they share genetics and milieu?
I think it's under-appreciated that there's still a lot of genetic variation within families due to the segregation of chromosomes. Especially for rare/outlier conditions (which is the type of thing we are most likely to notice by eye), even when heritability is high you would still expect most cases to come from families that did not have the condition: https://theinfinitesimal.substack.com/p/what-happens-to-heritable-conditions
I have another question, unrelated to race. Can intelligence be increased over generations? Let's assume optimal environmental conditions: individuals are well-fed, highly academic due to parental encouragement, and free from confounding factors like lead poisoning or anything else that could lower IQ. Under these conditions, can the breeder's equation still be applied to humans? There's a caveat: I'm not suggesting we should do this, but is it theoretically possible, and would it yield significant gains in intelligence in humans?
I'm usually pretty responsive to critique but this article did not actually talk about anything I wrote so there is not much to respond to. Instead, Weingard simply uses me as a stand-in to parrot out a bunch of unrelated correlative arguments that he has already made in the past and that I'm frankly not very interested in. The erroneous claims regarding between group heritability and admixture analysis have already been addressed elsewhere: https://kevinabird.github.io/2023/12/30/No-Case-For-Race-Realism.html . What I've learned is that hereditarians are mostly non-scientists, so if you present a conventional theory -> model -> data argument they will immediately ignore it and retreat to theory-free anecdotes and speculation, as happened here.
Thank you I don't fully know the subject I get confused on it very easily there is no much back end data I'm having to parse out thanks for making this simple.
So, this might be a dumb question: how phenotypic correlations and assortative matting are not circularities to each other?.
Again, probably dumb, trying to understand this:
“relatives by marriage but not by kinship (step siblings/cousins in-law), for whom resemblance is only driven by the assortative mating of their parents (m); and relatives through adoption but not kinship, for whom resemblance is expected to be zero.”
The scenario is this one?: two couple of parents mated assortatively, then one of each of their respective progenies who are not genetically related to their respective couple of parents marry each other?.
Ok, I hope, despite me being dumb, for hilarity I am getting this right:
“Interestingly, using the CTD to estimate heritability under random mating produced a value for the shared environment that fits the adopted siblings fairly well, but does not fit the other biological relatives.”
So, the adopted resemble more their family than the twins? :)
But, it got to me more hilarious:
“On the other hand, accounting for assortative mating could fit most of the biological resemblances but reduced the estimate of the shared environment to zero and thus no longer fit the adopted and in-law resemblances. “
Then again more folly from me:
´In-law relationships, of course, only share some of their environment: the component that is captured by assortative mating and the component after marriage. It therefore remains unknown how much lower the “h2” estimate could go if full environmental sharing was modeled.´
Is that not a restatement that people who come from similar upbringings: same religion, same political beliefs, same educational levels, etc., tend to mate more often among each other than with people not having similarly?
Again, sorry:
“Notably, the lowest kinship heritability estimates were from the cross-generational comparisons — avuncular and grandparental — that are less likely to share environments. The authors concluded that such patterns were not consistent with genetic dominance/epistasis alone and were likely to be driven by the shared environment.”
Meaning I am more influenced by the shared environment of my uncles and grandparents than the environment of my siblings?, despite they are likely to be more different?. Or that I look more like my siblings because our environment is more different than the one for our parents and uncles?. Not necessarily more similar...
There is also another pattern evident to me in the green-green graph: the smaller the effect, the smaller the difference, meaning only small differences can be expected to achieve consistency, i.e. no contradictions. So the modeling of kinship, genetics and environment with these sort of models can only show or find small effects which are consistent…
To reformulate: close to zero “true correlation” the absolute differences are smaller, meaning at some point, the more variables are brought in they will explain a lot: everything has an small influence... and non causal nonetheless...
So, again dumber:
“By extension, the differences in correlations seen between MZ and DZ twins, the workhorse of the Classic Twin Design, do not generalize to the correlations observed for other biological relationships.”
Does that mean twin studies are only useful to explain twin studies?.
Does this mean I am more like my more distant relatives than my own twins or other siblings?:
´This aligns with the theory of “dynastic effects” recently observed using genetic data, where offspring phenotypes were as or more significantly correlated with the genetic scores (and presumably phenotypic influences) of extended relatives than the genetic scores of their own parents.´
So the conclusions seem to me:
1.- The more data you bring in, the less influence seems to be measured on the influence of genes on behavior, in this cases called similarity.
2.- Measuring the effect of factors is not consistent: they differ in strength of influence depending on how you calculate the "independent" strength of each. Making muti-factor correlation estimates useless...
3.- Meaning there probably is no "independent strength" in this sort of studies for each factor.
I guess supported by this:
"I’ve previously written about how twin studies do not agree on heritability estimates even internally, driven by the assumptions they make about assortative mating, shared environment, and interactions. Now this study, in a single massive cohort with consistent modeling, shows that the disagreement persists for extended pedigree and adoption data as well."
Adding non-direct genetic effects as dynastic effects and epistasis: the influence of genes on each other, seems to me an example of bringing more confusion instead of more clarity, more variables, obscure ones, to fit "the data" seems like massaging data until it seems real...
And the future way probably to achieve that is again concluding everything matters, it just so happens a little bit and in contradictory, i.e. inconsistent manners.
And the way the review is presented, as is written, seems to not lend itself to the sort of identifying the contradictions but obfuscate them by looking to fit the data and speculating on why should it be?, instead of acknowledging beyond we don´t know enough, ignoring it does not make sense because it can´t make sense...
Does Eftedal or Zaitlin look into what would happen if you set aside (Monozygotic) Twins and tried fitting only the remaining relatives? Perhaps twins are just an outlier case b/c of so many potential interactions, local environments, contingencies...whatever you want to call it?
Both studies have a very small number of twins and the way the model is fit they basically do not contribute to the parameter estimates. Eftedal did a post-hoc analysis where they showed that if you estimated heritability *only* from twins, it wouldn't match the observations in the rest of the pedigree, and vice versa.
Thanks for writing on these topics. Very helpful!
That's a great figure Sasha; it really clarifies what is going on.
I keep suggesting Sam Harris have Dr. Gusev on his podcast. I hope it happens!
”Why do kids look and act like their parents?” I always felt that they kind of don’t? That they ought too look more alike, that the ought to inherit more traits and more cultural behavior - if they share genetics and milieu?
I think it's under-appreciated that there's still a lot of genetic variation within families due to the segregation of chromosomes. Especially for rare/outlier conditions (which is the type of thing we are most likely to notice by eye), even when heritability is high you would still expect most cases to come from families that did not have the condition: https://theinfinitesimal.substack.com/p/what-happens-to-heritable-conditions
I have another question, unrelated to race. Can intelligence be increased over generations? Let's assume optimal environmental conditions: individuals are well-fed, highly academic due to parental encouragement, and free from confounding factors like lead poisoning or anything else that could lower IQ. Under these conditions, can the breeder's equation still be applied to humans? There's a caveat: I'm not suggesting we should do this, but is it theoretically possible, and would it yield significant gains in intelligence in humans?
hello sasha i'm not a geneticist I don't know enough about this subject but what are your thoughts on this article
https://www.aporiamagazine.com/p/sasha-gusev-is-wrong
I'm usually pretty responsive to critique but this article did not actually talk about anything I wrote so there is not much to respond to. Instead, Weingard simply uses me as a stand-in to parrot out a bunch of unrelated correlative arguments that he has already made in the past and that I'm frankly not very interested in. The erroneous claims regarding between group heritability and admixture analysis have already been addressed elsewhere: https://kevinabird.github.io/2023/12/30/No-Case-For-Race-Realism.html . What I've learned is that hereditarians are mostly non-scientists, so if you present a conventional theory -> model -> data argument they will immediately ignore it and retreat to theory-free anecdotes and speculation, as happened here.
Thank you I don't fully know the subject I get confused on it very easily there is no much back end data I'm having to parse out thanks for making this simple.
So, this might be a dumb question: how phenotypic correlations and assortative matting are not circularities to each other?.
Again, probably dumb, trying to understand this:
“relatives by marriage but not by kinship (step siblings/cousins in-law), for whom resemblance is only driven by the assortative mating of their parents (m); and relatives through adoption but not kinship, for whom resemblance is expected to be zero.”
The scenario is this one?: two couple of parents mated assortatively, then one of each of their respective progenies who are not genetically related to their respective couple of parents marry each other?.
Ok, I hope, despite me being dumb, for hilarity I am getting this right:
“Interestingly, using the CTD to estimate heritability under random mating produced a value for the shared environment that fits the adopted siblings fairly well, but does not fit the other biological relatives.”
So, the adopted resemble more their family than the twins? :)
But, it got to me more hilarious:
“On the other hand, accounting for assortative mating could fit most of the biological resemblances but reduced the estimate of the shared environment to zero and thus no longer fit the adopted and in-law resemblances. “
Then again more folly from me:
´In-law relationships, of course, only share some of their environment: the component that is captured by assortative mating and the component after marriage. It therefore remains unknown how much lower the “h2” estimate could go if full environmental sharing was modeled.´
Is that not a restatement that people who come from similar upbringings: same religion, same political beliefs, same educational levels, etc., tend to mate more often among each other than with people not having similarly?
Again, sorry:
“Notably, the lowest kinship heritability estimates were from the cross-generational comparisons — avuncular and grandparental — that are less likely to share environments. The authors concluded that such patterns were not consistent with genetic dominance/epistasis alone and were likely to be driven by the shared environment.”
Meaning I am more influenced by the shared environment of my uncles and grandparents than the environment of my siblings?, despite they are likely to be more different?. Or that I look more like my siblings because our environment is more different than the one for our parents and uncles?. Not necessarily more similar...
There is also another pattern evident to me in the green-green graph: the smaller the effect, the smaller the difference, meaning only small differences can be expected to achieve consistency, i.e. no contradictions. So the modeling of kinship, genetics and environment with these sort of models can only show or find small effects which are consistent…
To reformulate: close to zero “true correlation” the absolute differences are smaller, meaning at some point, the more variables are brought in they will explain a lot: everything has an small influence... and non causal nonetheless...
So, again dumber:
“By extension, the differences in correlations seen between MZ and DZ twins, the workhorse of the Classic Twin Design, do not generalize to the correlations observed for other biological relationships.”
Does that mean twin studies are only useful to explain twin studies?.
Does this mean I am more like my more distant relatives than my own twins or other siblings?:
´This aligns with the theory of “dynastic effects” recently observed using genetic data, where offspring phenotypes were as or more significantly correlated with the genetic scores (and presumably phenotypic influences) of extended relatives than the genetic scores of their own parents.´
So the conclusions seem to me:
1.- The more data you bring in, the less influence seems to be measured on the influence of genes on behavior, in this cases called similarity.
2.- Measuring the effect of factors is not consistent: they differ in strength of influence depending on how you calculate the "independent" strength of each. Making muti-factor correlation estimates useless...
3.- Meaning there probably is no "independent strength" in this sort of studies for each factor.
I guess supported by this:
"I’ve previously written about how twin studies do not agree on heritability estimates even internally, driven by the assumptions they make about assortative mating, shared environment, and interactions. Now this study, in a single massive cohort with consistent modeling, shows that the disagreement persists for extended pedigree and adoption data as well."
Adding non-direct genetic effects as dynastic effects and epistasis: the influence of genes on each other, seems to me an example of bringing more confusion instead of more clarity, more variables, obscure ones, to fit "the data" seems like massaging data until it seems real...
And the future way probably to achieve that is again concluding everything matters, it just so happens a little bit and in contradictory, i.e. inconsistent manners.
And the way the review is presented, as is written, seems to not lend itself to the sort of identifying the contradictions but obfuscate them by looking to fit the data and speculating on why should it be?, instead of acknowledging beyond we don´t know enough, ignoring it does not make sense because it can´t make sense...
Does Eftedal or Zaitlin look into what would happen if you set aside (Monozygotic) Twins and tried fitting only the remaining relatives? Perhaps twins are just an outlier case b/c of so many potential interactions, local environments, contingencies...whatever you want to call it?
Both studies have a very small number of twins and the way the model is fit they basically do not contribute to the parameter estimates. Eftedal did a post-hoc analysis where they showed that if you estimated heritability *only* from twins, it wouldn't match the observations in the rest of the pedigree, and vice versa.