On twin studies, unrelated lookalikes, behavior and personality

Posted by Sappho on August 26th, 2014 filed in DNA


Razik Khan, in No two look-alikes, points out this New York Times article on examining the genetic component in twin behavior by comparing the personalities of people who simply look alike. Are identical twins similar in personality simply because they look alike, and people therefore treat them similarly? New research suggests that’s not all there is to it, because people who merely look alike are not as similar in personality as identical twins. From the New York Times article:

For Dr. Segal’s initial study, she asked Mr. Brunelle to send questionnaires to some of his subjects, and she received completed forms from 23 pairs of unrelated look-alikes. The questionnaires yield a score based on five personality measures: stability, openness, extroversion, agreeableness and conscientiousness. The participants also took the Rosenberg Self-Esteem Scale, a widely used measure in social science research.

As she expected, the unrelated look-alikes showed little similarity in either personality or self-esteem. By contrast, twins — especially identical twins — score similarly on both scales, suggesting that the likeness is largely because of genetics. Her results were published in the journal Personality and Individual Differences.

Here, via Razib Khan is the original study comparing twins and lookalikes. Khan also recommends Whole genome approaches to quantitative genetics, for a complementary approach investigating the influence of heredity on the same traits.

But, I want to highlight Razib Khan’s caveat at the end.

Relying on the body of twin research alone as a foundation might be a shaky basis for conjecture, but now this area is going multi-disciplinary, allowing for a stool with multiple legs. Of course all it is doing is confirming modest heritabilities for behavioral phenotypes.

While Khan closes with “God does play dice” (and includes a reminder that even much of the environmental component in our behavior isn’t subject to parental or governmental control), he is not reporting that research shows that the majority of our behavior is genetically determined.

One of the things that happens when discussion of genetics intersects discussion of politics is that, if you lean at all to the left in your politics (as I do), people who lean both conservative and hereditarian (in some cases more hereditarian than the science warrants) interpret any qualifiers you offer about the influence of heredity as a statement of belief in a blank slate, full stop.

I, for one, do not believe in a blank slate. I both believe that human nature in general is somewhat bounded by heredity (that we have somewhat different behavioral inclinations on average than, say, Neanderthals would have, had they survived) and that individual differences in personality and behavior are somewhat heritable. But I tend to expect the hereditary component of such things to be on the order of, say, 20-40%. And that in some cases there’s enough evidence both pointing to heredity and pointing to environment that it’s hard to say which is contributing how much. This is the case, for instance, with gender differences, where there’s a huge body of work pointing all over the place, as far as how large they are or aren’t and how far they are or aren’t culturally determined. (Partly this happens because your results depend on how you’ve defined the question. For instance, men on average are ready for sex on shorter acquaintance and with less commitment than women on average, but how far this is true depends hugely on how you’ve set up the experiment and what your operational definition of “interested in casual sex” is. Is it whether you’d have sex with an attractive total stranger right now? Whether you think, when asked in a survey, that you’d like to have sex with a hot relative stranger if certain risks and social judgments were removed?) But if I say, we don’t know how far gender differences are culturally determined, someone less feminist than me will often interpret this statement as me denying science and insisting, full stop, that it’s All About Culture.

Anyway, for the record, I tend to expect the heritability of our behavior to be about 20-40%. I think that sexual orientation is about 20-40% hereditary. I think that personality traits are about 20-40% hereditary, on average. I suspect that gender differences are about 20-40% genetically driven, and otherwise heavily influenced by culture. I’m open to adjusting my views up or down for the hereditary component of any given trait, or for the influence of heredity in general. But if you start to advance your argument by proposing that the only available choices are “mostly innate” and “not innate at all,” then I, as someone who sees, in what I’ve read about current research, a lot of support for modest inheritance of behavioral traits, will start to suspect that you’re making more of a political than a scientific argument.

That said, the links Razib Khan supplies are an interesting look at the influence of heredity, and are worth checking out.



One Response to “On twin studies, unrelated lookalikes, behavior and personality”

  1. Wired Sisters Says:

    Apparently I have a double who hangs out in some of the same places, and is involved in some of the same activities as I am. But, apparently, her politics are more radical than mine. We have never met.