Yes, but it would in principle be a less important factor than if it was heritable.
I’m not sure about that. We already observe the correlations and consequences of IQ. They need an explanation, but what needs an explanation is already known; a choice of explanation doesn’t retroactively change the observations to be of a smaller or lesser magnitude, does it? (If Mercury is observed to be 1 degree off predicted by Newton, and we finally choose relativity’s space warping as the explanation, we don’t then go back and say ‘we chose relativity therefore now we know the observations was actually 5 degrees off our predictions!’ The fact remained the same, Mercury didn’t move; it all adds up to normality.) We can say exactly that if your parents have IQ of X points the kid will average complicated-formula IQ points. If the causal factor runs through epigenetics rather than genetics, say, what does that actually change? Since we’re not discussing an exotic intervention with tailored epigenetic viruses or trying out prototype artificial wombs which might affect the actual causal pathway, just picking a mate where whichever causal pathway it is, it is active.
(I spent yesterday watching PGM videos so I’m wondering how to formulate this as a Bayesian network and d-separation problem… Hm.)
Pretty much. This falls out of the heritability research. One parent with IQ X, another with IQ Y, a known heritability of Z%, keep the environment constant, and some formulas later you have your probability distribution for the kid’s IQ.
I’m not sure about that. We already observe the correlations and consequences of IQ. They need an explanation, but what needs an explanation is already known; a choice of explanation doesn’t retroactively change the observations to be of a smaller or lesser magnitude, does it? (If Mercury is observed to be 1 degree off predicted by Newton, and we finally choose relativity’s space warping as the explanation, we don’t then go back and say ‘we chose relativity therefore now we know the observations was actually 5 degrees off our predictions!’ The fact remained the same, Mercury didn’t move; it all adds up to normality.) We can say exactly that if your parents have IQ of X points the kid will average complicated-formula IQ points. If the causal factor runs through epigenetics rather than genetics, say, what does that actually change? Since we’re not discussing an exotic intervention with tailored epigenetic viruses or trying out prototype artificial wombs which might affect the actual causal pathway, just picking a mate where whichever causal pathway it is, it is active.
(I spent yesterday watching PGM videos so I’m wondering how to formulate this as a Bayesian network and d-separation problem… Hm.)
Do we?
Pretty much. This falls out of the heritability research. One parent with IQ X, another with IQ Y, a known heritability of Z%, keep the environment constant, and some formulas later you have your probability distribution for the kid’s IQ.
Okay—in that case, I agree that if we already know the value of Z then why exactly it has that value isn’t relevant.