[rheide]

The article assumes that if 1000 positive gene changes are responsible for increasing intelligence, you'll get 1000 increases of intelligence. It's much, much more likely that a lot of those gene changes overlap, meaning combined changes won't affect the IQ any more than either/or changes would.

[gwern]

> It's much, much more likely that a lot of those gene changes overlap, meaning combined changes won't affect the IQ any more than either/or changes would.

No, it's not. Please see Hsu's extended comments on the topic of interaction and non-additivity at the top of the page, and note the support from twin studies & GCTA. The point of his suggested strategy is that you're targeting the subset of gene changes in total heritability which don't do that.

[NoMoreNicksLeft]

It's even possible that some are mutually antagonistic/incompatible.

If gene A or gene B increase intelligence, A + B might cancel out or even lower it below baseline. It all depends on what mechanisms are responsible and how the genes alter those.

[vixin]

Possible to check this out. Society is polarizing and smart people are now tending to associate with other smart people to an extent that did not occur earlier. How are their progeny performing?

[T-A]

Like this http://archive.wired.com/wired/archive/9.12/aspergers_pr.htm... and this http://www.economist.com/node/4032638 ?

Those pesky genes tend to affect more than one thing...

[ap22213]

Isn't it true that children regress toward the norm? For example, two very tall people will more likely have a child that is shorter. Or, is that wrong?

[bhouston]

> Isn't it true that children regress toward the norm?

There isn't perfect heritability of all traits because they are influenced by multiple genes. Two tall people likely have different genes that are making them tall, and there are likely interactions happening with other genes in their genome. When you get a child, they have some genes from each parent but they also get lot of novel interactions between the genes of tallness and the rest of their genome. This tends to reduce the heritability of traits that are modulated by complex interacting gene networks.

But it isn't a regression to some imaginary norm, it just means that that genes are not simply additive, they context dependent. But if you were to isolate 100 tall people from the population and had them breed for generations, and 100 short people and have them breed for generations, their offsprings likely will continue different in height still, they will never regress to the same "norm" because there are more tall genes or short genes in their respective populations.

[rjsw]

Diet has changed a lot over a few recent generations, a child that should be a bit shorter than its parents due to genetics may still be taller.

Other multivariate genes may provide a better example of regression.