Evolutionary Psychology and Behavioral Genetics

Dr. Miller’s and Dr. Kanazawa’s Ten politically incorrect truths about human nature is everywhere these days. I discussed it over coffee with Daniel Nexon (of The Duck), Sean Meade (of Interact) emailed it to me, and it has appeared both on Thomas P.M. Barnett :: Weblog and South Dakota Politics. Like evolutionary psychology (of which this article is a manifestation), it is useful in that it helps smash the Standard Social Science Model, but incomplete in that it does not fully embrace social sciences.

The Standard Social Sciences Model (SSSM) is the overall research program of social sciences since World War II. It is most notable for ignoring biological factors, especailly at the group level, as causes of variation in human behavior. So ancient stone axes are described as “ceremonial” (the idea that weapons are for violence being seen as biological reductionism), and racial variations are not even mentioned as possible hypothesis when looking at racial gaps in intelligence or attention span. The SSSM essentially put half of all variables in taboo, hobbling social science to this day. The Evolutionary Psychologists, and the sociobiologists before them, have been tireless opponents of the SSSM, opening the door to real social scientific research for the first time in generations.

However, the exclusion of biological factors from social sciences for half a century did its work in limiting the utility of early biological explanations. The central tools of social science, regression and correlation in explaining variation, are underused by EP and SB because they were relatively new to social science at the time of the taboo began. More scientific approaches to biological factors have now appeared, and these generally go by the name of behavioral genetics. The Evolutionary Psychologists and Biopsychologists ultimately did not prevail, but took the damage that allowed more scientific approachesto flurish.

So back to the original article, Ten politically incorrect truths about human nature. Twenty years ago the authors would have been hounded out of academia, because they dare believe that biology influences behavior. Nowadays there specific claims are dismissed, because of weak operationalize and overbroad generalities.

That’s progress. That’s science. That’s the search for knowledge.

There is now real debate. Men like Edward O. Wilson and John Tooby are to thank for that.


Some links: The twin blogs, Gene Expression and gnxp, are amazing sources for the latest in behavioral genetics. I first learned of the EvolPsych/behavgen split from Steven Pinker. And at Dreaming 5GW, I examined two cases where dangerously presented half-truths are worse than no truth at all.

Update: Per a request from Sean, my uninformed impressions of the specific claims are below the fold:

Men like blond bombshells (and women want to look like them)

Men should be expected to be attracted, and women should be expected to try to emulate, any hard-to-fake sign of reproductive fitness. As skin color and intelligence are generally correlated with moderate climates, this would imply that the idela female type should be skewed towards signs of moderate climate. (In all populations, women’s skin tends to be lighter than men’s, for perhaps this reason). Blond hair is a particularly European mutation, however, so this specific claim seems unlikely as a human universal.

Humans are naturally polygamous.

Better to say men historically have higher variance in the number of reproductive partners they have than women. The last universal male ancestor was much closer to our time than the last universal female ancestor for just this reason.

Most women benefit form polygyny, while most men benefit from monogamy

Indeed. Monogamous societies are male guilds, where competition for females is limited for the bettermen of the average men. In the same way, the professions with the greatest “merit pay” relative to standard wage (academia, hollywood, professional sports) are worst for average workers but best for the best performing. See The Right Nation for more on this.

Most suicide bombers are Muslim

This is an objective fact, so within the claim itself there is no debate.

My own research indicates that genetic variation might have more to do with the particularly Arab (or perhaps more accurately, Semite) form of terrorism we see today. But I don’t know. This is a really open question, and there is a lot of money to be made in implementing an answer.

Having sons reduces the likelihood of divorce.

The claim and explanation both sound reasonable. No argument.

Beautiful people have more daughters.

Something similar is true of deer populations: daughters of highly reproductive males tend to be under-reproductive themselves. My guess is that beauty is an adaption that is particularly useful for females, and so it should skew toward females.

What Bill Gates and Paul McCartney have in common with criminals.

They have something else in common two: risk taking behavior and focus on abstract concepts. Many of the differences between males and females may come from two tendencies which emerge almost at birth: the male preference for systems and risk over people and stability.

The midlife crisis is a myth — sort of

Sounds reasonable. The authors present a specific test of their hypothesis, so it’s up to someone to tst it.

It’s natural for politicians to risk everything for an affair (but only if they’re male)

Solid discussion of inclusive fitness v. individual fitness.

Men sexually harass women because they are not sexist

Many behaviors are sex- (or at least gender-) dependent, and misapplication of these behaviors violates social norms. (Try to punch a mouthy woman or embrace a male aquaintence to see this for yourselves). The authors extend this logic to “harrassment.” Still, it goes without saying that many “harrasing” behaviors would be non-normative in a same-sex environment.

17 thoughts on “Evolutionary Psychology and Behavioral Genetics”

  1. “Blond hair is a particularly European mutation, however, so this specific claim seems unlikely as a human universal.”

    Do Japanese and African men like blondes better than other types of women? The fact that historically they never saw one is not a reason to say yes or no to the thesis. I suspect the thesis is correct based on informal observations.

  2. Daniel,

    I read through the posts. A lot of the charges can be leveled against a variety of published articles in social science. As I mentioned in the article, I’m a fan of the evolutionary psychologists more for the good they do in breaking the SSSM than for the persuasiveness of their arguments (which is where I prefer the behavioral geneticists).


    I’ve had the pleasure of meeting researchers who link-up with remote tribes in order to conduct social science experiments [1]. It would be interesting to use that sort of expertise and bring in otherwise balanced photographs and videos of models…



    PS: All reactions in this corner of the blogosphere are much better than the blind rage exhbited on Dakota Women [2,3] and, I expect, many other SSSM adherents.

    [1] http://digitalcommons.unl.edu/politicalsciencehendricks/13/
    [2] http://dakotawomen.blogspot.com/2007/07/post-on-south-dakota-politics-causes.html
    [3] http://dakotawomen.blogspot.com/2007/07/another-reason-to-hate-summer.html

  3. The last study I read on hourglass shapes (not hair) suggested that the primary explanatory variable was degree of contact with western media, render the constructionist explanation more plausible. No idea about hair color, though.

    Dan: that’s a very odd response. Gelman demolishes the article; at the end of the day, the statistical analysis is SO bad that it doesn’t prove anything it purports to. Handwaving about general problems with the “SSSM” as you call it, doesn’t absolve the author (try finding that magnitude of mistake, for example, in Gary King’s work). Indeed, just because one likes the result isn’t a reason for affirming bad work or taking it seriously

  4. “My own research indicates that genetic variation might have more to do with the particularly Arab (or perhaps more accurately, Semite) form of terrorism we see today.”

    Dan, a quibble. “Arab” is a construct, particularly in the Levant where and Arab is just a likely to be ethnically Arab as he is to be Circassian, Turk, European, Egyptian, Greek or even Jew.

    I’m willing to believe that behavior may pass down genetically through selection, certainly culture which is ‘heritable’ also plays a part.

  5. Daniel,

    Gelman does not demolish the article, and this is not surprising, as he did not attempt to. His published criticism [1] addresses two previously published conclusions

    1. Engineers are more likely to have sons
    2. Attractive people are more likely to have daughters

    The first claim is not addressed in the “Ten Myths” article, so it is as irrelevent in “demolishing” Miller Kanazawa’s claims as previous criticism of Gelman [2] is in demolishing his demolition!

    The second claim, unrelated to the first, is addressed in the Psychology Today article. Gelman’s criticism breaks into three parts

    a. criticizing the operationalization of attractiveness
    b. criticizing the lack of the bonferroni correction
    c. criticizing the interpretation of a logistic regression coefficient

    Hopefully Kanazawa addressses these criticisms, and supplies more data one way or the other.

    You write that “at the end of the day, the statistical analysis is SO bad that it doesn’t prove anything it purports to.” This might be better written as “at the end of the day, the statistical analysis for one of the article’s ten claims is questionable.” Arguing from there than the entire article is “SO bad” or that it “doesn’t prove anything” is as reasonable as saying that published criticisisms of Gelman allow us to say his critique is “SO bad” or that they “don’t prove anything.”

    … though of course, Gelman’s critique isn’t quite so long as that. The Psychology Today appears to report prior findings, while Gelman critiques Kanazawa’s own research.

    “Indeed, just because one likes the result isn’t a reason for affirming bad work or taking it seriously”

    True, and just because one dislikes results isn’t a reason for rejecting good work or not taking it seriously. Gelman raises serious issues for one claim, number 6 of 10. The response, at least among the presumably knowledgeable readership of Andrew’s blog [3], tends toward hysteria.


    Very good points.

    Racial constructs are interesting, in that they can be meaningful even if not in the way that are presupposed. For instance, Turks in Turkey as as Turkik as blacks in America are “white.” Both are populations with a significant outside genetic contribution.

    With research on the “other half of social science variables” open for the first time in ages, there’s exciting work to be done!

    [1] http://www.stat.columbia.edu/~gelman/research/published/kanazawa.pdf
    [2] http://links.jstor.org/sici?sici=0081-1750(1995)25%3C185%3ARMSIUI%3E2.0.CO%3B2-8
    [3] http://www.stat.columbia.edu/%7Ecook/movabletype/archives/2007/07/how_should_unpr.html

  6. Let me rephrase: mistakes of the magnitude made in this claim reflect very poorly on the veracity of the evidence behind his other claims. What Gelman documents are appallingly bad methodological mistakes–ones that can most charitably be described as “incompetent.” They are a different oder than the substantive debate about Bayesian methods and, for example, priors found in the article you link to.

  7. Just to clarify a couple points:

    1. I agree with Daniel’s statement that Kanazawa’s statistics are so flawed that they do not prove what K. had claimed to prove. I also agree that Kanazawa’s claims could very well be true (and I think this is clear in my article). But I don’t think they would’ve been published as is, had the statistical flaws been realized. There are many many many possible conjectures to make, and Kanazawa’s claim in his article was that he had proven his results (to within reasonable confidence) statistically. So: “not proven,” not “false.”

    2. On to the specifics: I agree about my comments (a) and (c). I think it should be possible to address (a) with additional data. There is no way to address (c); that was simply a mistake in the presentation of the results.

    But on (b), I don’t suggest that K. do a Bonferroni correction. I agree that this is a possibility but I think that a simple regression coefficient (averaging the attractiveness ratings for all 3 waves of the study) would be the way to go. I brought up the Bonferroni correction in my article only because Kanazawa had picked such an odd comparison. Again, this concern could be addressed using new data (or, for that matter, using all 3 waves of the study). I think it’s likely the results won’t be statistically significant, but, again, that doesn’t make the claim false.

    A power analysis suggests that if the sex ratio effects are in the range of 1% (which is perhaps a priori plausible), you’d need a huge sample size to find patterns. In this case, further analysis of Kanazawa’s existing data wouldn’t show much. I think it would be fair to say that the data are consistent with K.’s hypotheses, and the data are also consistent with many many other hypotheses.

    Based on the evidence I’ve seen, I think it’s a little early to be writing a book called “Why beautiful people have more daughters” or presenting these conjectures as fact in Psychology Today. But I agree with Dan that my statistical criticisms shouldn’t be considered a reason to not study these things.

    3. I like the Raftery article that you linked to. But Raftery and I have disagreements about statistical methods. Kanazawa was making mistakes with standard methods. This is a bit different. I don’t think Raftery would disagree with my criticisms of Kanazawa’s work at all. As I said in my blog entry, I suspect that if Kanazawa had collaborated with a statistician, that these errors wouldn’t have occurred. (But then the work, with its non-statistically-significant results, probably wouldn’t have appeared in J. Theor. Bio., and we might not be having this discussion at all.)

    4. I think the hysterical (if that’s what it is) reaction on our blog was to Kanazawa’s presentation of unproved conjectures as proved. I don’t think it was a reaction to sociobiology etc. in general.

  8. Daniel,

    Ultimately the question si to what extent ad hominem attacks are valid in social science critiques. I think it’s fair to describe our current conversation as one voice believing that if some findings of one published article are wrongly interpreted,the author’s entire body of work (including both findings and literature reviews) should be conclusively presumed to be flawed. Another view is that such a black list is foolish.


    Thank you for your comment. I think we agree.

    I’m unfamiliar with Kanazawa’s line of research. While I guessed his findings are true in my original post, certainly it’s possible to argue for ways in which human females are under less evolutionary pressure than human males (far more distant universal common ancestor, etc), so who knows. Ultimately we need more data, and I hope Kanazawa presents the rest of his findings.

    However, as I read the PT piece, Kanazawa does not cite his own research in making claim 6. It seems that even ignoring the Kanazawa beauty article, he could have written #6 just the same.

    I have not purchased his book, so I cannot comment on the claims made therein.

  9. lere,

    Thanks for the link!

    From the article:

    Frost’s theory is also backed up by a separate scientific analysis of north European genes carried out at three Japanese universities, which has isolated the date of the genetic mutation that resulted in blond hair to about 11,000 years ago.

    The hair colour gene MC1R has at least seven variants in Europe and the continent has an unusually wide range of hair and eye shades. In the rest of the world, dark hair and eyes are overwhelmingly dominant.

    Just how such variety emerged over such a short period of time in one part of the world has long been a mystery. According to the new research, if the changes had occurred by the usual processes of evolution, they would have taken about 850,000 years. But modern humans, emigrating from Africa, reached Europe only 35,000-40,000 years ago.

    The “10,000 years ago” formula is one that keeps turning up. About that time, humans began living in cities, and indeed were able to live in large-scale societies even before there were farms. Modern humans and those just 15,000 ago may be behaviorally quite different from each other. [1]

    I’m not sure what “the usual processes of evolution” are that the article mentions, but I’m guessing they mean genetic drift. Natural selection became politically incorrect in the 1960s, so there’s a consistent push to explain differences in terms of random assortment than fitness maximization. Too bad, because natural selection matters now more than ever. [2]

    [1] http://www.tdaxp.com/archive/2007/05/02/review-of-before-the-dawn-by-nicholas-wade.html
    [2] http://www.tdaxp.com/archive/2007/12/14/what-if-evolution-works-15000-times-faster-than-we-imagined.html

  10. Lere,

    Very interesting!

    It would be interesting to know how much of visible human variation (eye color, height, hair patterning, etc.) is the result of sexual selection.

    Likewise, to what extent this sexual selection is the result of mate choices of men by women, as opposed to mate choices of men by the father’s of women, or even differential rates of rape victimhood, would be interesting to know.

  11. Lere,

    You bring up very interesting points!

    “Black Africans” were once a minority of Africans, who were more successful than their competitors and spread. I’m not sure how this relates to the Bantu expansion [1], but the relatively more ancient origins of the Bushmen, Pygmies, and Aboriginies is discussed by Nick Wade in Before the Dawn [2,3].

    Scientific genetics have let us begin to see the world as it really is, in spite of the myth makers of the 1970s (who denied such things as races existed [4]) and the 19th century (who viewed them as platonic forms [5]).

    [1] http://en.wikipedia.org/wiki/Bantu_expansion
    [2] http://www.tdaxp.com/archive/2007/05/02/review-of-before-the-dawn-by-nicholas-wade.html
    [3] http://www.amazon.com/Before-Dawn-Recovering-History-Ancestors/dp/1594200793
    [4] http://en.wikipedia.org/wiki/Lewontin%27s_Fallacy
    [5] http://en.wikipedia.org/wiki/The_Passing_of_the_Great_Race

Leave a Reply

Your email address will not be published. Required fields are marked *