4 ms·
For anyone that's curious, the actual point of The Bell Curve is that modern society's knowledge economy increasingly rewards IQ, so we should be careful not to
by sebastos 8y ago
For anyone that's curious, the actual point of The Bell Curve is that modern society's knowledge economy increasingly rewards IQ, so we should be careful not to create a world that conflates "high IQ" with "good person", since you don't really get to choose your IQ. Then it goes on to try and bolster the case that picking winners based on IQ could have unforeseen negative consequences. In that setting, there is a chapter that claims that this IQ-centered culture would disadvantage certain races more than others, since there are differences between average IQ when you divide people into groups based on their race. The question of heritability arises because the authors want to argue that these differences might persist for an arbitrarily long time if they are not strictly environmental. To support this possibility, they give their interpretation of the data on this topic: that the expressed IQ differences are likely at least partially genetic.
The article OP posted argues that this is not a parsimonious explanation of what we see. The author points out that heredity and genetic determination are not synonymous, and reminds that there are a number of results that suggest environmental effects can strongly influence group IQ. Further, we cannot de-conflate the effects of passive vs. active covariance between environmental and inherited effects. An example of the these two categories is the difference between:
(Passive) "Children with high musical ability tend to be born into families that will give their child lots of music lessons"
and
(Active) "Parents tend to give their children more music lessons if they demonstrate high musical ability early in childhood"
where in both these cases an inherited trait (natural musical ability) created a double-whammy environmental effect. The former is easier to control for, but the latter requires a domain-specific hypothesis to isolate. (I'm not sure I completely followed the article here, but this is the best summary I can give.)