Baba Brinkman’s latest salvo in his quest for a fact-based justification for his proposal to select meanness out of the human race by not sleeping with it really boils down to a question most members of my generation will likely remember from a childhood saturated in “Sesame Street”: Who are the people in your neighborhood?
We’ve come to this question because Brinkman has finally discovered that there is, in fact, data that might suggest genetic variation contributes to variation in “meanness”—even if he couldn’t be bothered to cite it in connection with the campaign up to now:
In his new post, Yoder’s argument is not that male violence isn’t an adaptation; rather, he argues that our violent tendencies have been so completely drilled into us by natural selection that they show insufficient genetic variation for selection to act on …
He’s right that a complete lack of individual genetic differences in proneness-to-violence would be a death-blow for my campaign, but luckily for me and all the other peaceniks who support the DSWMP credo, Yoder simply didn’t bother to look up any of the evidence.
You have to love how, after implicitly conceding the factual point—that in his first attempt to shore up the scientific basis of DSWMP, he cited data that has nothing to do with the question at hand—Brinkman chides me for not doing my homework. In fact I’ve acknowledged at every step of our little back-and-forth that there is a body of research which suggests there’s some genetic contribution to variation in what we might call “meanness.” My argument isn’t that this genetic contribution doesn’t exist—it’s that this genetic contribution is pretty much meaningless from the perspective of an individual person’s dating life.
Geneticists typically differentiate the total variation in a phenotypic trait into components due to genetic variation, and variation created by environmental effects. If you’re a biologist who’s really keen to understand the genetic contribution to variation—or to artificially select for a desirable trait—you spend a lot of effort ensuring that you’re working with a population that has the smallest possible environmental variation. If you’re studying plants, you grow your sample in the most non-variable environment that you can construct in a common garden or a greenhouse or a climate-controlled growth chamber. If you’re studying humans, you have to get more creative.
You can sample only people with very similar upbringings. Or you can statistically “control” for the effects of different environments by including environmental differences as a covariate in your data. Or you can compare identical twins with fraternal twins; because the former share 100% of their genetic code and the latter share only 50%, if identical twins are more alike than fraternal twins, genetics may be responsible for the similarity. However, this last method really isn’t very good at comparing genetic and environmental variation, unless it also includes data from pairs of twins who were raised separately [PDF]. Much of the literature on human behavioral genetics, including what Brinkman cites, lacks this kind of data.
So there are, indeed, studies that demonstrate some genetic contribution to diagnosed psychopathy or measures of aggression. But these are estimates of genetic variation found via methods that aren’t very good at comparing it to the effects of environmental variation, or that deliberately minimize environmental variation in the studied sample. The relative contribution of genetics and environment to variation in the human population as a whole—the people that you meet, when you’re walking down the street, or browsing OKCupid—is likely to be very different.
Brinkman cites one study that demonstrates this point very neatly. Here’s the relevant quote from the New York Times article Brinkman links to, with emphasis added:
Kevin Beaver, an associate professor at Florida State University’s College of Criminology and Criminal Justice, said genetics may account for, say, half of a person’s aggressive behavior, but that 50 percent comprises hundreds or thousands of genes that express themselves differently depending on the environment.
He has tried to measure which circumstances — having delinquent friends, living in a disadvantaged neighborhood — influence whether a predisposition to violence surfaces. After studying twins and siblings, he came up with an astonishing result: In boys not exposed to the risk factors, genetics played no role in any of their violent behavior. … In boys with eight or more risk factors, however, genes explained 80 percent of their violence.
Neither the Times nor Brinkman provides a citation of the specific peer-reviewed article presenting that result, but I’m pretty sure it’s this one, published in The Journal of Adolescent Research in 2011—and it does indeed find that, in boys experiencing more risk factors for criminal behavior, there is a stronger effect of genetics on whether or not they actually committed crimes.
Having read the study, I suspect that some of this effect is actually because environmental variation is simply smaller in the higher-risk group—as they have more risk factors in common, their upbringings become more similar. But even if we take Beaver’s interpretation of his results at face value, it’s actually pretty devastating to the “Don’t Sleep with Mean People” scheme.
That’s because, if genetic variation is only strongly predictive of variation in criminality for people who’ve experienced many environmental risk factors, then you’d have to exclusively date people with those risk factors (this is a filter I haven’t found on OKCupid yet) in order to know that the variation in their personalities is probably due to genetics. And any time you date people who don’t have those risk factors, your perception of their “meanness” will be entirely unrelated to whether or not they carry gene variants associated with risk-factor-triggered criminality—you could end up reproducing with someone whose genome is chock-full of variants for “meanness” that are never apparent in the right environment.
(Confidential to BB: If you don’t stop referring to data that makes my point for me, people are going to think we’re in cahoots.)
Back when I first raised a fuss about “Don’t Sleep with Mean People,” I was about 75% sure that the response I’d get—if any—would be to tell me that the whole thing was done tongue-in-cheek, and I shouldn’t be such a stickler for accuracy. And, you know, “Don’t Sleep with Mean People” isn’t such a big deal as a joke, or even as a one-off novelty rap. In fact, it’s a joke anyone with even a passing understanding of natural selection has made, whether chuckling over some unfortunate recipient of a Darwin Award or watching “Idiocracy”.
But I’ve been flabbergasted to discover that Baba Brinkman actually seems to be quite serious—he thinks that individual dating choices could actually make future generations perceptibly, genetically nicer. As a social movement, “Don’t Sleep with Mean People” is like a meditation philosophy built around contemplation of the Eternal Puzzle of Why the Chicken Crossed the Road—with the added tone-deafness of an implication that people need to be reminded not to have sex with partners who are antisocial and violence-prone.
Regardless, we’ve now got to the point where I’m essentially repeating the same argument I made from the beginning, only with more words and greater exasperation. Until Brinkman actually shows some sign of understanding how absurd and counterproductive “Don’t Sleep with Mean People” is, I’ve got nothing further to say.
Beaver, K. M. 2010. Environmental moderators of genetic influences on adolescent delinquent involvement and victimization. Journal of Adolescent Research 26:84–114. doi: 10.1177/0743558410384736.
Coccaro, E. F., C. S. Bergeman, R. J. Kavoussi, and A. D. Seroczynski. 1997. Heritability of aggression and irritability: a twin study of the Buss-Durkee aggression scales in adult male subjects. Biological Psychiatry 41:273–84. doi: 10.1016/S0006-3223(96)00257-0.
Viding, E., R. J. R. Blair, T. E. Moffitt, and R. Plomin. 2005. Evidence for substantial genetic risk for psychopathy in 7-year-olds. Journal of Child Psychology and Psychiatry 46:592–7. doi: 10.1111/j.1469-7610.2004.00393.x.
Schönemann, P. 1997. On models and muddles of heritability. Genetica 99:97–108. doi: 10.1023/A:1018358504373.