Politics in Nature Neuroscience

ast week, it was some poll by an obscure advertising company named "Most Think Country on Wrong Track" --also known as Ipsos--which made the dubious claim (backed up by some dumb rhetoric by Pat Schroeder ) that liberals read more books than conservatives. This week, another inflammatory conservative-bashing article has turned up. Not in New Republic, Slate or Weekly World News as you might expect, but in the scientific journal Nature Neuroscience.

The article, "Neurocognitive Correlates of Liberalism and Conservatism" [1], has been widely publicized, and widely misunderstood, across the Internet. While the article itself is of limited significance, it's important to scientists because it drags the field of cognitive neuroscience a little closer to the morass of political psychology.

Front cover of this week's Nature Neuroscience as portrayed on the Internet

If you're from Russia or China, you may be familiar with political psychology. In those countries, those who disagreed with the party in power, if they weren't shot, were sent to insane asylums and re-education camps. Political psychology is the mixture of science and pseudo-science that justified putting them there. Here in America, practitioners of political psychology have been mostly limited to such activities as posthumously psychoanalyzing Adolf Hitler and Josef Stalin, to find out where it all went wrong. From these inauspicious beginnings, political psychology has only one direction to go: up.

The Nature Neuroscience article purports to demonstrate a brain difference between American-style "liberals" (i.e., left-wingers) and "conservatives" (i.e., right-wingers). The difference was revealed by a behavioral test (which might be called the 'Dubya test'), in which subjects were presented either a "W" or an "M" on a computer screen. The subjects were supposed to press a key only when an "M" was presented. Since most of the stimuli were M's, a W was unexpected and subjects often erroneously pressed a key after seeing a W. EEG electrodes measured the evoked response during the error, and a score called an "error-related negativity" or ERN was computed. The ERN was smaller in conservatives than for liberals at a significance level of p<0.001. This means that one part of the brain called the anterior cingulate cortex was more active in liberals than in conservatives when they made an error.

The anterior cingulate cortex is a very small region buried deep inside the brain, on the corpus callosum, which is a bundle of fibers that connects the left and right hemispheres. It is involved in various types of low-level processing, such as emotion, reward anticipation and error-detection tasks [2]. It also generates the ERN, which is a sort of feedback signal that occurs when the expected input differs from the perceived input [3]. In most people, the size of the ERN correlates with the severity of the error. The anterior cingulate cortex then generates an emotional or "affective" response. Reduced activity in the anterior cingulate cortex correlates with depression, while increased activity is taken to be an indicator of emotional awareness.

Randomly-generated numbers with r value of 0.32

This part of the brain has long been a political battleground. Many spurious claims of racial and gender superiority, based on small differences in the size or activity of the corpus callosum, have been made over the past 100 years. None of these studies claiming a correlation with IQ or any other index of mental ability has ever held up.

According to the authors of the Nature Neuroscience study, liberals scored slightly better than conservatives on the Dubya test, although the r value was only 0.30, which means it was barely significant (p<0.05). Since there were 43 subjects, that means it was a very weak correlation indeed, and it is very likely that this finding will likewise not hold up in future studies that are done more carefully.

Just to show that a correlation with an r value of 0.3 is not much to get excited about, I generated 43 random numbers and changed four of them by hand until the coefficient of correlation (r value) was 0.3. The results are shown in the graph at left. The r value of these data is actually 0.32, slightly higher than in the paper. Scientists, of course, would be aware of this; the average layman hearing about this study in the press might not be. Maybe the data behind this statistic were omitted to save space. But they probably looked something like this.

What the findings mean

The paper is what scientists call a "one-graph paper", where a single piece of data is shown. Both Science and Nature publish many such one-graph papers. Such papers tend to be superficial and tend to be published only if they fit the prejudices of the editors. This explains the barrage of global warming papers in these two magazines (although Nature has cut back on global warming papers of late, perhaps an indication that the editors of Nature are starting to become a little more skeptical).

If the findings are true, it undermines a small body of prior research that attempted to relate the anterior cingulate cortex to pathological states such as schizophrenia and so-called attention deficit hyperactivity disorder (ADHD). By definition, conservatives are normal, since they comprise over 50% of the population (except possibly at liberal arts colleges, where liberals comprise 99.9999% of the population). Using this minor paper to claim, as some people undoubtedly will be tempted to do, that liberals are "smarter" than conservatives (or vice-versa), or that liberals are "more empathetic" would be unwarranted. There is no suggestion of a difference in intelligence between liberals and conservatives in the article. Its main significance is to undermine previous work on ADHD.

Of course, that won't stop former congressmen from Colorado, or the vast armies of rationality-challenged people on the Internet, who enjoy taking juvenile swipes at their political opponents. It's scarcely an exaggeration to say that if laws against hate speech were enacted, it would instantly wipe out half the Internet. (Laws against porn would wipe out the other half.)

More seriously, the infusion of politics into neuroscience also undermines the confidence of lay people in the medical research establishment. We have seen how political footballs like global warming have undermined the credibility of climatologists. As scientists, we should resist the politicization of neuroscience.

The authors themselves interpret the results to mean that liberals have greater conflict-related neural activity when inhibition of a response is required. The implication is that liberals are better at restraining themselves from making a mistake. However, although this view is clearly held by the authors, the results presented do not justify this conclusion.

Problems in the study

The biggest problem in this study is that the subject population is not described, other than to say that there were 43 subjects, all right-handed, and 63% female. Are these college students or what? Are they randomized for age, social background, race, and intelligence? The report does not say.

Scientifically, it's a minor result because performance on a simple motor task tells us very little about the psychological makeup of the subjects. It could be, for example, that the liberals had been spending all their time playing video games, and therefore inadvertently training for this task, while the conservatives were tired from working at something called a "job" in a place called "the real world."

Even if the finding is real, its meaning is obscure. As I see it, the main value of this study is that it gives me a chance to crack jokes about liberals. Example: liberals undoubtedly have better arm-stretching and grasping skills because of all the practice they get hugging trees. Their memory skills are probably better too, because of all the practice they get memorizing nonsense syllables at Jane Fonda seminars. (Had enough?)

The choice of articles cited in the references also betrays the lack of impartiality of the researchers. For example, the authors say, "The study of personality variables that accompany differences in political opinions goes back more than fifty years." This is true, although the authors would probably have been better off not mentioning it. They cite The Authoritarian Personality, which was a seriously flawed book by Adorno et al. published right after World War II that attempted (and failed miserably) to draw connections between the so-called authoritarian personality type and Nazism. Another reference is to an equally flawed 2003 paper by Jost et al., [4] who is also a co-author on this paper. In the 2003 paper, Jost claimed, implausibly, that the communist dictators Josef Stalin, Fidel Castro, and Mao Zedong were all "conservatives." Conservatives are still chuckling over that one.

Presumably, the 2007 study dropped most of the left-wing totalitarian dictators from the "conservative" group (although the paper doesn't say so; come to think of it, we haven't seen much of Fidel Castro lately).

J. T. Jost, in particular, appears to be a committed left-wing ideologue, as the titles of his papers show. In his past work, he consistently (and erroneously) defines conservatism as "resistance to change." Two titles are: "Are needs to manage uncertainty and threat associated with political conservatism or ideological extremity?"[5] and "Moral outrage mediates the dampening effect of system justification on support for redistributive social policies." [6].

The latter paper consisted mainly of hand-wringing about "how and why people tolerate ongoing social and economic inequality." (The answer, not surprisingly, is "system-justifying ideology.")

One conservative-bashing paper might have been a bizarre side-project. But a career of such papers gives the impression of someone trying to use science as a vehicle to promote his ideology. Because of this, some people may wonder: how many different measures did the authors have to test before they found one on which liberals performed better than conservatives--and would they have reported the results if they had found one that favored conservatives? That this question can even be asked demonstrates the corrosive effect of bringing politics into neuroscience.

The lead author, David Amodio, is another committed liberal who specialized in attempting to find and quantify something called "race bias" which is supposedly revealed by our eyeblink responses [7]. That avenue of research appears so far not to have been wildly successful. On the other hand, the senior author, C.M. Yee, appears to be a more serious scientist who is mainly interested in studying event-related potentials in a variety of contexts.

Where do liberals and conservatives come from

There is some evidence that people become political liberals because they perceive the government to be an extension of their parents. That's why college-age people tend to be more liberal than older people. For young people, who cannot yet survive well on their own, the dependence on parents and their financial resources is still strong. Yearning for a paternalistic government in many ways represents a transference of parental dependency onto a strong, Contrary to popular belief, political conservatism may actually be a measure of a person's willingness to take personal risks--to go without a net, so to speak. Liberals, in contrast, look to a strong central government as a source of guidance and financial security. all-embracing central authority that provides guidance and minimizes risk. The ideal government for liberals is one that takes care of people's financial and health-care needs at the expense of individual freedom, just as a parent would do.

The most extreme examples of this were in the totalitarian communist states of the 20th century, where the government dictated every aspect of speech and behavior. In Communist China, for example, a disrespectful comment about communism would land a person in a re-education camp, where they were pressured to publicly recant and apologize. (Sound familiar?) As the power of socialism increases, a socialist state automatically becomes more and more totalitarian [8].

The theory that liberals are seeking a substitute parental authority figure may explain two other common characteristics of liberals: their oft-mentioned herd mentality, where young people imitate those with higher social status in order to be fashionable or "cool"; and their intense interest in sex and gender issues such as homosexuality and women's rights. It also may explain other features of liberal groups that have been mentioned by many political commentators: their appeal to emotion instead of reason ("For the children!"), the undercurrent of rebelliousness ("Speak truth to power"), and their frequent use of personal attacks instead of argument. These are all characteristics of college-age people who are still developing their skills at argumentation, but are also common to older liberals.

As people gain more experience, self-confidence and courage, most of them develop an increased ability and willingness to take personal risks. Along with that comes an increased sense of personal identity, and a correspondingly heightened valuation of individualism and independence. To these people, big government is often simply an unwanted intrusion that interferes with their desire for freedom. These individuals are what in this country are called "conservatives." From this, one can predict that risk-takers (such as businessmen) would tend to be more politically conservative, while those in professions where there is little personal or professional risk (such as tenured professors) would tend to be more liberal. Women also tend to be more risk-averse [9], and accordingly are more likely to be liberal, while soldiers are more often politically conservative.

However, political philosophy can be fluid, and a person's political beliefs may change many times over a lifetime. That fact alone is strong evidence against a neurophysiological basis for ideological choice. Changes in personal political beliefs can be brought about by reason and argument, but often they are forced upon the person by a global political crisis. For example, a person may realize that their culture is in danger, and adopt an ideology that expresses nationalistic values; or they may decide that their culture is evil, and adopt a nihilistic ideology. There are undoubtedly also elements of groupthink that predispose the formation of extremist ideologies on both sides of the spectrum. There are also undoubtedly many different kinds of conservatives (although liberals, I am told, all march in lock-step).

Conclusion

The Amodio paper, and others like it, are motivated by a desire to medicalize political dissent. If these authors succeed in convincing their colleagues that political "conservatism" is some sort of brain dysfunction, the next logical step is to devise a treatment for it. Anyone familiar with the history of the Stalinist state in the 20th century, regardless of their party affiliation, should find that a chilling thought.

References

[1] Amodio DM, Jost JT, Master SL, Yee CM. Neurocognitive correlates of liberalism and conservatism. Nat Neurosci. 2007 Sep 9.
[2] J Comp Neurol. 2005 Dec 5;493(1):154-66. Neural mechanisms of autonomic, affective, and cognitive integration. Critchley HD.
[3] Rushworth MF, Behrens TE, Rudebeck PH, Walton ME. Contrasting roles for cingulate and orbitofrontal cortex in decisions and social behaviour. Trends Cogn Sci. 2007 Apr;11(4):168-76.
[4] Jost JT, Glaser J, Kruglanski AW, Sulloway FJ. Political conservatism as motivated social cognition. Psychol Bull. 2003 May;129(3):339-75.
[5] Psychol Sci. 2007 Mar;18(3):267-74. Moral outrage mediates the dampening effect of system justification on support for redistributive social policies. Wakslak CJ, Jost JT, Tyler TR, Chen ES.
[6] Jost JT, Napier JL, Thorisdottir H, Gosling SD, Palfai TP, Ostafin B. Are needs to manage uncertainty and threat associated with political conservatism or ideological extremity? Pers Soc Psychol Bull. 2007 Jul;33(7):989-1007.
[7] Amodio DM, Harmon-Jones E, Devine PG. Individual differences in the activation and control of affective race bias as assessed by startle eyeblink response and self-report. J Pers Soc Psychol. 2003 Apr;84(4):738-53.
[8] The Road to Serfdom, F. A. Hayek University of Chicago Press, 1944.
[9] Rosen AB, Tsai JS, Downs SM. Variations in risk attitude across race, gender, and education. Med Decis Making. 2003 Nov-Dec;23(6):511-7.