= 27d567bc2d48d9f40e9ccc20ea370b15)
see comments on Open Thread, September 1-15, 2012 - Less Wrong
You are viewing a comment permalink. View the original post to see all comments and the full post content.
= 783df68a0f980790206b9ea87794c5b6)
Loading…= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
Powered by Reddit
= f037147d6e6c911a85753b9abdedda8d)
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (353)
Was reading up on the Flynn effect, and saw the claim it's too fast to reflect evolution. Is that really true? Yes, it's too fast, given the pressures, for what Darwin called natural selection, given the lack of anything coming along and dramatically killing off the less intelligent before they can reproduce. But that's not the only force of evolution; there's also sexual selection.
If it's become easier in the last 150 years for women to have surviving children by high-desirability mates, then we should, in fact, see a proportionate increase in the high-desirability characteristics. And since IQ and socioeconomic status are correlated, and SES is a known high-desirability characteristic, we would expect an increase in IQ accordingly, insofar as IQ is heritable.
And, in fact, there is a change in society that would do that — increasing urbanization. Not only have cities become healthy enough to have non-negative population RNIs for the first time in history, but they've also become the home of the majority of the human species for the first time in history. Studies of infidelity rates show it does, in fact, correlate fairly strongly with urbanization (probably for the logical reasons that increased population density increases opportunities and urban anonymity makes it easier to conceal from a mate).
So, the urbanization of the last 150 years increased successful infidelity. The usual models of sexual selection indicate that successful infidelity by women should result in high SES men having more children. IQ is correlated with high SES. IQ seems to be heritable in large part. And the period where we would expect high SES men to have more kids is matched by an increase in the general population's performance on tests of IQ.
I'm currently operating without good access to scientific journals to see if this has been considered and debunked, or not considered, or considered and put forward. But, at least sitting here just thinking about it without the resources to test it (or even model it effectively mathematically), it seems an increase in the genes that increase IQ as a result of sexual selection could be a plausible explanation of the Flynn Effect.
But there's also an opposing evolutionary pressure: educated women have fewer children.
The Reproduction of Intelligence attempts to quanitfy this effect:
Shouldn't they be checking for number of grandchildren?
I'd assign a low probability to this hypothesis. Most of the Flynn effect seems to occur on the lower end of the IQ spectrum moving upwards. Source. This is highly consistent with education, nutrition and diseases hypotheses, but it is difficult to see how to reconcile this with a sexual selection hypothesis.
Also, I'm not sure that your hypothesis fits with expected forms of infidelity. One commonly expected form of common infidelity would be generally with strong males while trying to get a resource rich males to think the children are there's If such infidelity is a common pattern, then one shouldn't expect much selection pressure for intelligence, if anything the opposite.
The fraction of the population which engages in infidelity even in urban environments is not that high. Infidelity rates in both genders are around 5-15%, but only about 3% of offspring have parentage that reflects infidelity. Source, so the selection impact can't be that large.
One thing worth noting though is that one of the pieces of evidence for disease mattering is that there's a correlation between high parasite load and lower average IQ, but your hypothesis would also cause one to expect such a correlation since reduced parasite load would be better correlated with better medicine and more functional urban environments in general. This is evidence in favor of your hypothesis.
I'm not aware of any obvious way to test your hypothesis. I'd be curious if you have any suggested things to look at or if anyone else has any ideas.
It reconciles quite well, actually.
The greater the genetically-determined status differential between a woman's husband and her a potential lover, the more differential advantage to the woman's offspring in replacing the husband's genes with those of a higher-quality male. So the lower the status of the husband, the greater the incentive to replace his genes with another's.
Assuming for a moment IQ is 100% heritable and IQ is linear in advantage, the woman with an IQ of 85 and a husband of IQ 85 will see her kids have an IQ of 85 if she's faithful, and 115 with a lover of 145, for a net advantage to her kids of +30 IQ if she strays. If a woman and her husband are IQ 100, the same lover will raise the IQ +22.5; her kids get less advantage than Mrs. 85. In the case of Mr & Mrs. IQ 115, the advantage is only +15. For Mr & Mrs. IQ 130, the advantage to cheating is only +7.5. For Mr. & Mrs. IQ 145, cheating with a lover of IQ 145 doesn't benefit her kids at all, while for Mr & Mrs. IQ 160, she wants to avoid having kids by a lover of IQ 145.
So, it is precisely the women on the low end that have the greater incentive to cheat "up", which we would expect would result in more cheating, and thus the low end where IQs would increase the most.
Also, the lower status the woman's husband, the easier it is to find a willing lover of higher status, and thus the greater the opportunity to replace the husband's genes with another's. Mrs. 85 can find a lover with IQ 100 more easily than Mrs. 100 can find a lover with IQ 115, even though the both have the same incentive to find a lover of +15 IQ points. Mrs. 115 has even more difficulty finding a lover of IQ 130, and so on.
So, it is precisely the women on the low end that have the greater opportunities to cheat "up", which we would expect would result in more cheating, and thus the low end where IQs would increase the most.
Assuming monogamous and assortative marriage, there's a serious limit to how high resource/high status a male a woman can marry relative to her own status. Assuming she's already landed the best husband she can manage, it is then in her subsequent interest to acquire the best genes for her kids she can. Insofar as better genes result in higher status, this would translate to favoring high-status males as lovers to produce kids supported by the best-she-can-manage husband. If status correlates better with IQ rather than strength in human societies, well, we'd expect that to select for IQ.
Ahah. My data on this was substantially out-of-date. That is a serious blow to the hypothesis.
(Hmm. Except that in modern welfare states, the government has replaced the husband as supporter on the low end of the socioeconomic ladder, so maybe the effect is now most strongly happening among the children unmarried women, which would cause a drop-off in children of infidelity corresponding to the rise of out-of-wedlock births? Meh.)
Yeah, me neither.
A test would be to look wether there is a correlation between cheating and IQ and whether this correlation is influenced by sex. Also, asymetrical incidence of STDs with respect to the sexes could also be an indicator.
How would you test this model?
There are so many other factors, you're probably getting mostly noise there. For instance: I read somewhere that depending on whether babies drink breast milk or formula, they may lose 10 points (to formula) - the reason stated was lack of omega 3. What about lead paint chips? We have banned lead, that should increase IQ - after an initial decrease when lead paint began to be used. (There'd be a similar increase / decrease cycle with the invention of formula.) The point of these two is that as we learn more, we may be preventing a lot of things that previously caused children brain damage. And then there are other health factors which we've improved. In the great depression, I read 10% of the population starved to death. Starvation, for those who survive it, can cause brain damage. Were there other starvations before this, that had stopped happening? When did helmets become popular for people riding bicycles and skateboards and such?
There are just too many factors.
Heh, and I read somewhere that here in America, the Flynn effect has stopped. O.O
1) Sure. I'm not claiming the Flynn effect is genetic; I'm disputing the common claim that it can't be genetic.
2) Whether the Flynn effect has stopped or not is an area of ongoing dispute; some studies suggest it merely paused for a while. And if it has ended . . . that might merely mark that America's reached the new equilibrium point under urban infidelity conditions.
"I'm disputing the common claim that it can't be genetic."
Oh, sorry.
I have found out the hard way, myself, that it's really best to start with a single sentence that makes one's point clear in the very beginning. Maybe that would help your commenters respond appropriately.
When people say that it's "too fast," they are making a quantitative claim. The Flynn effect is a standard deviation per generation. Under your scenario of no selection on women, this would require that the bottom half of the bell curve to have no biological children. 50% cuckoldry, perfectly correlated with IQ? Even men who think they've been cuckolded don't have that high a rate.