March 11, 2014

Are conserv-lib diffs genetically rooted? Not as much as Mooney implies

Chris Mooney/Wikipedia
Chris Mooney, he of "motivated reasoning" discussion, reviews here two books that discuss this issue. They are "Predisposed: Liberals, Conservatives, and the Biology of Political Differences," by John Hibbing et al, and "Our Political Nature: The Evolutionary Origins of What Divides Us," by Avi Tuschman.

The biggest caveat? Mooney never claims that evolution does select for political differences, but let's make that clear, because he wants to have his cake and eat it on this issue in general, specifically concerning political conservativism. The titles of the two books should indicate that.

So, again, let's be clear.

Evolution does not "select" for political differences. Psychological differences that are part of political differences, like anxiety/neuroticism, openness to experience, etc., may well be selected for, but political experience is not. That's just like religious belief is not "selected for," but rather, things like pattern recognition and agency imputation, which likely are selected for, become part of an overarching system that develops into religious belief. Mooney does halfway acknowledge this, as I note below, but, it's cursory, and probably not even halfway.

And, in terms of sexual reproduction, evolution certainly selects for many things, not just five personality issues.

Let's start right here with what's wrong:
Again and again, when they take the widely accepted Big Five personality traits test, liberals tend to score higher on one of the five major dimensions—openness: the desire to explore, to try new things, to meet new people—and conservatives score higher on conscientiousness: the desire for order, structure, and stability. Research samples in many countries, not just the U.S., show as much. And this finding is highly consequential, because as both Hibbing et al. and Tuschman note, people tend to mate and have offspring with those who are similar to them on the openness measure—and therefore, with those who share their deeply rooted political outlook. It’s a process called “assortative mating,” and it will almost certainly exacerbate our current political divide.  
First, men and women pair up for many reasons other than political inclinations. That includes perceived health by facial and bodily similarity, fat deposits in certain parts of the body (ahem), perceived ability to be a provider and much more. Note that those items I all spelled out are all purely physiological. No psychology involved.

So, once again, per the eponymously named (by me) intellectual shortcut, Mooney appears to be "pulling a Mooney."

If political inclinations were a primary driver of human mating, we'd see a lot more division than we currently do.

This is more nuanced than some of Mooney's writing on this issue, but not fantastic.

So, on to more counterpoints.

First and foremost?

What's heritable is a tendency, nothing more. And, while 40 percent (if correct, and that number is nothing more than a guesstimate) is significant, it's still less than half of 100 percent. There are a number of traits we know of, that a genetic tendency toward which has a 40-percent basis. And, I'm sure that 40 percent is high. If we take the five-type personality profile, as an example, and insist that the composite of all five types has a 40 percent heritability factor, each individual factor would have to be more than 80 percent heritable, if we're using the basic "and" conjunctiveness. (Do the math yourself.) Even if we allow for some looseness of overlap, and not a strict "and," still, for all five types to be 40 percent heritable as a chunk, the individual ones would still have to be at least 65 percent heritable, and NO human trait is that heritable.

Even if I re-read Mooney's reading of Hibbing to claim that each factor individually is 40 percent heritable, I'm still skeptical. Read the next few paragraphs for details.

Related to that is the fact that the five-type test, while perhaps better than the MMPI, is not the be-all and end-all of psychological research on personality types.

Related to that? None of the studies in either of the books Mooney reviews appear to make any allowance for epigenetics. None. In the case of neuroticism, a large part of which is anxiety-related, we know from lab animal experiments that there's an epigenetic, as well as a genetic, component to heritability of anxiety.

Indeed, that's an issue with twin studies, which are supposed to be the gold standard. Identical twins (setting aside chimera issues) not only share the same genes, but, vis-a-vis even fraternal twins, depending on the date of twin division of the original single egg, identicals may share a more similar womb environment, if it includes a single placenta or single amniotic sac. (Identicals, depending on date of division, may have separate amniotic sacs and placentas, one sac but separate placentas, or share both.)

Related to that? Even on the genetic side, we're still in the Neolithic, or at best, the Early Bronze, on pairing specific sets of genes (no single-gene claims for personality factors need apply) to heritability of personality factors.

This is a big failure of proponents of personality testing in general. It's another reason why, while I say that the five-type personality spectrum is better than the MMPI, it's not the be-all and end-all. It's probably not that on how it divides personality and what it considers as the most important aspects of personality. It's certainly not that on its explanation of how these differences arise. Wikipedia's article on the Big Five, while not exhaustive, is a good starting point. It notes, outside the "criticisms" section, that the five factors appear to not be total cultural universals, despite implications otherwise on Mooney's story.

Mooney does caveat the evolutionary angle, to be sure:
Moreover, in evolution, some things happen for an explicitly Darwinian “reason”—traits become more prevalent or fixed in populations because they advanced organisms’ chances of survival and reproduction in a particular environment—while others happen more accidentally. Some complex social traits may emerge, for instance, because they are a fortuitous by-product of other, more fundamental traits laid down by Darwinian evolution.  
He notes religion as a good "spandrel" example.

However, Hibbing then claims conservativism is probably the "default." Why? Because Hobbes, and life is "nasty, brutish and short."

It was 5000 BCE, after the invention of agriculture, to be sure. But, not to read too much into "The Gods Must Be Crazy," maybe it was NOT so in, say, 25,000 BCE. That's a failure of Hibbing more than Mooney, but it's also a failure of Mooney for not making this observation himself.

He then goes on to Tuschman's book. Problems are to be found here, some of which Mooney notes. The biggest is that Tuschman may be advocating group selection as a partial explainer. As an educated layperson, I'm not dead against the idea, and think some evolutionary biologists who are, are likely wrong. But, it's still controversial. But, Mooney notes it's not clear that he does advocate this.

Mooney does say:
In the end, Tuschman’s book attempts a feat that those of us monitoring the emerging science of politics have long been waiting for—explaining the now well-documented psychological, biological, and genetic differences between liberals and conservatives with reference to human evolution and the differential strategies of mate choice and resource allocation that have been forced on us by the pressures of surviving and reproducing on a quite dangerous planet. It may or may not stand the test of time, but it certainly forces the issue.
That said, Mooney seems to want this to "force the issue." I've documented that above well enough, I think. It's part of his history.

Beyond that, "liberal" and "conservative" in the US exclude libertarians. Arguably, Mooney's reasoning does in other countries as well. Germany's Free Democrats, and possibly Britain's Liberal Democrats, would reject being put in either half of Mooney's dichotomy. 

And, that's not all. On to a few other points, with briefer discussion.

Another part of his history is to want to wrong-foot conservatives on this issue. Specifically on some subpoints, like this:
Being defensive, risk aversive, hierarchical, and tribal makes sense when the threats around you are very real and immediate.
So, our second point? Liberals are arguably just as tribal as conservatives. The "social justice warrior" movement, especially as illustrated by a group like Atheism Plusers, is Example No. 1 of this, with phrases like "mansplaining" and online actions like the Twitter add-on of the Block Bot, all while blaming only "the other," often very broadly defined, for the problems and actions involved. Since Mooney has recently moved the online location of the Point of Inquiry podcast because of said tribalism, he knows this full well. He refuses to discuss or admit this, but it's true. Indeed, what I call his "wrong-footing" is itself a tribal action.

In their book "Blindspot: Hidden Biases of Good People," authors Mahzarin Banaji and Tony Greenwald show, with social-psychology research, how self-professed liberals can exhibit, unconsciously, "tribal" behavior toward people of other racial groups, the opposite sex, the elderly and people of different sexual orientation.

Related to that? Again,the Big Five factors don't totally translate within the developed world. Even less do "conservative" and "liberal. In the non-WEIRD portions of our world, say, a New Guinean tribal chief, he'd scratch his head over this discussion.

Third? "Conservative" and "liberal" aren't "conservative" and "liberal."

Let's get back to this quote:
Being defensive, risk aversive, hierarchical, and tribal makes sense when the threats around you are very real and immediate.
Behavioral psychologists like Daniel Kahnemann and Amos Tversky have shown that we're ALL risk-aversive in many ways, like fearing a small financial loss more than salivating over a possible big gain.

Mooney knows this, too, I'm sure, even if he doesn't want to tie it in with his discussions.

Fourth? As I've indicated above, and on previous posts about Mooney, and occasionally about personality psychology in general, the Big Five is NOT the be-all and end-all of personal psychology study. I'm afraid that a number of people, including Mooney, obviously, are making it into a new MMPI, though. In reality, approaching vs. avoiding temperaments, people driven by intrinsic vs. extrinsic motivation (which partially overlaps Jung's old introvert and extravert) and other takes on personality reveal things the Big Five doesn't.

And fifth? Along the lines of David Hume, is does not equal ought. People do change political and social allegiances, just like religious affiliation.

In short, while it's not quite Pop Ev Psych, it's in the neighborhood. And, while it's not the worst example of scientism I've seen, it is an example.

On personality profiles and related areas of psychology, Mooney either doesn't know that much about all the science issues that are at stake, or else he's cherry-picking. I'm sure the latter is part of it, based on his previous writing on political differences. But, I don't think that's all of it.

He's good on things like climate change, but on psychology in general and personality psychology in particular, I think there are some understanding deficits, too.

And, the more he writes on this, the worse he gets. I've written a lot about him on this issue; click the "pulling a Chris Mooney" tag (not totally about him, but primarily so) for more. There's no other way to put it, other than to say that "worse" is, in my opinion, a tribalism-based "worse."

No comments: