Friday, April 08, 2011

Liberals Are Conflicted and Conservatives Are Afraid

This sums up the basic conclusion of a new study on political orientation and brain structure by Ryota Kanai, Tom Feilden, Colin Firth and Geraint Rees in the journal Current Biology. Yes, that Colin Firth...

Colin Firth's Speech during the 2011 Academy Awards. Firth won Best Actor for The King's Speech.

Why are Colin Firth and Tom Feilden, both listed with BBC Radio 4 affiliations, authors on this paper? Let's go back to Tuesday, 28 December 2010 and two pieces that appeared on the BBC website.
Politics: Brain or background?

Science correspondent Tom Feilden: "What started out as a bit of fun has turned into quite a significant piece of science."

Scientific research commissioned by this programme on behalf of our guest editor, Colin Firth, has shown a strong correlation between the structure of a person's brain and their political views.
You can also listen to a brief audio clip of Feilden discussing the study at the link above. Firth actually commissioned Professor Geraint Rees at University College London to obtain structural MRI scans from two diametrically opposed politicians: conservative MP Alan Duncan (a member of the Conservative Party) and liberal MP Stephen Pound (a member of the Labour Party).

Feilden then asks a question that is unanswerable from studying brain structure in adults: "Are political beliefs learnt, the product of experiences in our environment, or 'hard wired' in the brain?" Since a comparison of n=1 liberal versus n=1 conservative is not scientifically valid, Rees went back to a database of MRI scans from UCL students and asked these participants about their political beliefs. Feilden then discussed the results before the paper had been formally submitted for publication [according to the journal website, the paper was received by Current Biology on 11 January, 2011]. Briefly, he said that the gray matter of the anterior cingulate cortex was thicker among the liberal or left wing participants while the amygdala was much larger in those who identified as conservative or right wing.

"But is it cause and effect?" asks an interviewer. Rightfully so. Correlation does not equal causation. Then there's the claim that the structural brain variation means the political differences are "hard wired". The observed anatomical differences prove no such thing. Any experience will change the brain in some way, and repeated patterns of behavior, whether it's learning to juggle or voting conservative due to a certain set of core beliefs, can alter the brain. Nonetheless, we have the following headline:
Are political beliefs hard-wired?

Tom Feilden| 08:10 UK time, Tuesday, 28 December 2010

"Give me the child until he's seven and I'll give you the man."
It's clear from their motto that the Jesuits are firmly in the acquired camp when it comes to whether our political beliefs and values are learned or hard wired from birth: the product of experience rather than genetics.
But is that true?
...along with the eventual admission:
Although the results do show that political belief is reflected in the physical structure of the brain it's not clear which comes first. Whether the structure of the brain shapes political belief or political belief leads to the differential development of brain structure.
All right, that was a media stunt, you say -- but how about the peer reviewed paper (Kanai et al., 2011)?

A total of 90 healthy middle-class to upper-class participants (mean age = 23.5 yrs) underwent MRI scanning and [later?] filled out a very brief questionnaire on their political views:
Participants were asked to indicate their political orientation on a five-point scale of very liberal (1), liberal (2), middle-of-the-road (3), conservative (4), and very conservative (5). ... Because none of the participants reported the scale corresponding to very conservative, the analyses were conducted using the scales of 1, 2, 3, and 4.
If I'm not mistaken, no special effort was made to recruit very conservative participants, because the study was conceived after the MRIs were obtained.

As reported by Feilden, being liberal was associated with a larger anterior cingulate whereas being conservative was associated with a larger right amygdala1 (see Figure 1 below).

Figure 1 (Kanai et al., 2011). Individual Differences in Political Attitudes and Brain Structure. (A) Regions of the anterior cingulate where gray matter volume showed a correlation with political attitudes are shown overlaid on a T1-weighted MRI... A statistical threshold of p < 0.05, corrected for multiple comparisons, is used for display purposes. The correlation (left) between political attitudes and gray matter volume (right) averaged across the region of interest (error bars represent 1 standard error of the mean, and the displayed correlation and p values refer to the statistical parametric map presented on the right) is shown. (B) The right amygdala also showed a significant negative correlation between political attitudes and gray matter volume. Display conventions and warnings about overinterpreting the correlational plot (left) are identical to those for (A).

The results were based on measurements of gray matter density in these two specific structures. How were they chosen? First, the anterior cingulate was selected based on the finding of Amodio et al. (2007) that...
...the amplitude of event-related potentials reflecting neural activity associated with conflict monitoring in the anterior cingulate cortex (ACC) is greater for liberals compared to conservatives. Thus, stronger liberalism is associated with increased sensitivity to cues for altering a habitual response pattern and with brain activity in anterior cingulate cortex.
I had issues with this interpretation of the Amodio et al. study in 2007, which I will summarize here. One problem was attributing the observed results to political viewpoint and not to other factors. The study used EEG recordings, specifically event-related potentials. The ERP brain waves reflect electrophysiological activity recorded remotely from the scalp. While it's great for determining the temporal parameters of neural activity, it's not so great at determining where the activity is located in the brain.

One brain wave of interest was the error-related negativity (ERN), recorded at the time that people make mistakes in a task:
The response-locked error-related negativity (ERN), which peaks at approximately 50 ms following an incorrect behavioral response, reflects conflict between a habitual tendency (for example, the Go response) and an alternative response (for example, to inhibit behavior in response to a No-Go stimulus).
However, it's not at all clear that ERN reflects conflict-monitoring (Carbonnell & Falkenstein, 2006). Thus, based on a smaller-sized ERN in conservatives, one cannot conclude that they are "less responsive to conflict." In fact, if one wants to apply the logic of conflict monitoring to political viewpoint, one could say that conservatives might be more freaked out by ambiguity and conflict, since it violates their simplistic world view. Although liberals did indeed show larger ERN waves than conservatives when making mistakes, so do individuals with clinical diagnoses such as obsessive-compulsive disorder (Gehring et al., 2000) or major depressive disorder (Chiu & Deldin, 2007). So we can't dismiss the possibility that the liberals might have been more depressed or obsessive compulsive than the conservatives.

Back to the present day, the specific ACC region of interest (ROI) was chosen from an fMRI study [not the EEG study, which has lower spatial resolution] entitled, "A midline dissociation between error-processing and response-conflict monitoring" (Garavan et al., 2003). Garavan and colleagues found that different locations in the medial frontal cortex responded preferentially to error processing (the ACC ROI selected by Kanai et al., 2011) and to conflict monitoring (a more posterior and dorsal region called the pre-SMA). However, Kanai and colleagues concluded:
Thus, it is conceivable that individuals with a larger ACC have a higher capacity to tolerate uncertainty and conflicts, allowing them to accept more liberal views. Such speculations provide a basis for theorizing about the psychological constructs (and their neural substrates) underlying political attitudes. However, it should be noted that every brain region, including those identified here, invariably participates in multiple psychological processes. It is therefore not possible to unambiguously infer from involvement of a particular brain area that a particular psychological process must be involved.
But it seems to me that if you want to say "a higher capacity for conflict correlates with a larger X brain region", based on Garavan et al. you'd have to choose the pre-SMA and not the ACC.

Although Kanai et al. did offer caveats, other issues with their paper were raised by Professor Martha Farah in a ScienceNOW article, Does Your Brain Bleed Red, White, and Blue?
It's an appealing story and a topic worth investigating, says cognitive neuroscientist Martha Farah of the University of Pennsylvania. But there's plenty of reason to be cautious, she says. For one, it's not clear what a bigger amygdala—or a bigger anything in the brain—actually means in terms of brain function and behavior. The research, she says, is unclear and often contradictory on this point.

Another problem is that most brain regions have multiple functions, Farah says: "Who says fear is the only function of the amygdala?" She notes that this brain region also responds to sexually arousing images and pictures of happy faces, and one recent study found a correlation between amygdala volume and the size of people's social networks. Likewise, the anterior cingulate cortex has been implicated in a long list of cognitive functions. By picking and choosing from the previous studies, "they're indulging in a bit of just-so storytelling," Farah says.
After all this criticism, I have to point out an impressive aspect of the paper, and that is the replication of results in an independent group of 28 participants. In the end, I don't doubt that there are differences between the brains of liberal and conservative people. But how they got that way, and what it means, are questions for further investigation.


1 Why only the right amygdala and not the left? The authors didn't provide an explanation.

Further Reading

Additional posts on the "political brain" by The Neurocritic:
The Error of Prognosticating Political View by Brain Wave

Liberals Are Neurotic and Conservatives Are Antisocial

David Amodio Responds

Conservatives Are Neurotic and Liberals Are Antisocial


Amodio DM, Jost JT, Master SL, Yee CM. (2007). Neurocognitive correlates of liberalism and conservatism. Nature Neurosci. 10:1246-7.

Carbonnell L, Falkenstein M. (2006). Does the error negativity reflect the degree of response conflict? Brain Res. 1095:124-30.

Chiu PH, Deldin PJ. (2007). Neural evidence for enhanced error detection in major depressive disorder. Am J Psychiatry 164:608-16.

Gehring WJ, Himle J, Nisenson LG. (2000). Action-monitoring dysfunction in obsessive-compulsive disorder. Psychol Sci. 11:1-6.

Garavan H, Ross TJ, Kaufman J, Stein EA. (2003). A midline dissociation between error-processing and response-conflict monitoring. Neuroimage 20:1132-9.

Ryota Kanai, Tom Feilden, Colin Firth, Geraint Rees (2011). Political Orientations Are Correlated with Brain Structure in Young Adults Current Biology : PMID: 21474316

Liberals Are Conflicted and Conservatives Are Afraid?

Senate Majority Leader Harry Reid (D-Nev.), left, and House Speaker John Boehner.
Harry Reid, John Boehner can't agree on why Congress can't agree to avert government shutdown

Subscribe to Post Comments [Atom]


At April 09, 2011 4:24 AM, Anonymous Anonymous said...

In "The Invisible Gorilla", Christopher Chabris and Daniel Simons call this "Brain porn".

At April 09, 2011 5:34 AM, Blogger Nikolaj Lykke Nielsen said...

Thanks for a very informative and clear headed blog. Is it possible to get you to look at this article?
Moral Concepts Set Decision Strategies to Abstract Values

"Persons with a predominant collectivistic (altruistic) value system applied a “balancing and weighing” strategy, recruiting brain regions of rostral inferior and intraparietal, and midcingulate and frontal cortex. Conversely, subjects with mainly individualistic (egocentric) value preferences applied a “fight-and-flight” strategy by recruiting the left amygdala."

At April 09, 2011 10:53 AM, Blogger The Neurocritic said...

Nik - Thanks for the link to the paper, I hadn't seen it before. The experimental design and classification scheme are a bit more complicated. I'll have to take a closer look before I offer an opinion.

At April 09, 2011 2:30 PM, Anonymous Emmy said...

Despite the positive correlation, I find these studies a bit puzzling. Right away, why not use the voter registration as a measure (e.g. Registered Democrats vs. Republicans)? I'd think the personality scale would be too broad a category.

Also confusing: which part correlates to religion; absolutist thinking; reactionary thinking due to wealth or paranoia....wouldn't it be better to compare on a more narrow scale, such as those who vote a certain way on immigration, taxes, or the like?

At April 10, 2011 2:48 PM, Anonymous Anonymous said...

People who report scientific studies in major newspapers should put your blog on their required reading list. If I had a dime for every time the media overstated the conclusions of a neuroscience study, I wouldn't have to write grant proposals.

At April 10, 2011 6:51 PM, Anonymous The Neuro Times said...

I'm skeptical. I recently described this material as nihilistic individualism's insidious impact upon commonsense.

At April 11, 2011 4:22 AM, Anonymous Anonymous said...

As for the ERN reduction found in Amodio et al. (2007), there seems to be a confounding with the error rates (Olvet & Hajcak, 2008. Clin. Psychol. Rev. p.1345).

At April 11, 2011 6:15 AM, Blogger Liz said...

I'm sure I've heard of other psychological (not sure about neurological) studies which go towards proving conservatives are fraidy cats. They've been reporting them on Alternet for years before R4 mixed itself in. (Don't put off actors from funding research if they want to!) And doesn't that all mesh in with the theories of that George Lakoff guy?! And aren't all these dictator types basically just an extreme version of your paranoid alpha male?!

At April 11, 2011 10:01 AM, Blogger The Neurocritic said...

Emmy - The Current Biology study was done in the UK. Party affiliation might be too coarse a classification, anyway.

Anonymous April 10, 2011 2:48 PM - Thanks! And that would be a great alternative to grant writing...

The Neuro Times - Studies in the new field of cultural neuroscience are starting to look at differences between participants from collectivist vs. individualistic cultures.

Anonymous April 11, 2011 4:22 AM - I'll have to check out that article. Amodio et al. didn't report the actual error rates for liberals vs. conservatives but did see a correlation between liberalism and accuracy, which would result in larger ERN for the liberals. But they also said:

"This association suggests that a more conservative orientation is related to greater persistence in a habitual response pattern, despite signals that this response pattern should change (for example, on No-Go trials). This behavioral finding is consistent with the relationship that we observed between political orientation and neurocognitive sensitivity to response conflict. However, a partial correlation analysis revealed that the relation between political attitudes and the ERN remained strong after covarying behavioral accuracy (r(40) = 0.53, P < 0.001), suggesting that liberalism (versus conservatism) is associated with greater neurocognitive sensitivity to cognitive conflict, beyond what was observed from behavioral performance alone."

Liz - I agree with you that actors should fund more neuro/psych research. We now have two Oscar-winning neuroscientists!

At May 07, 2011 7:39 AM, Blogger R. J. Welland said...

Thanks for reporting on this study. With such uninspiring correlation coefficients (both less than .25), I'm surprised that the article was published at all. As a journal reviewer & researcher, I'm sure that I would've recommended rejection based on the very low correlation coefficients, alone!

At January 30, 2012 8:29 AM, Blogger MT said...

R.J Welland: the correlation rate, even if small, is pretty much significant. And in this kind of studies I would not even discard a .23 correlation as trivial. Indeed, I would expect Amygala volume as correlated with much more other features than political ideology and sharing more tha 20% can represent a very intriguing result. Said that, I think tat more than the results per se, the explanation should be more cautious. Indeed, I especially agree with two objections: the one reprted by the author from Martha Farah and the one pointed out by Emmy. I think that Kanai et al. should have ruled out if there may have been other confounding variables (as social network size) which can be correlated both to amygdala volume and (who knows?) to conservatorism alike. Also, as stressed by Emmy, political orientation may be a too vague phenomenon with underlying processes that can mediate the relationship between the former and amygdala volume. For instance this relationship may be mediated by attitudes toward immigration or, in general, what can be called social fear, as suggested even by a study by Oxley and colleagues published in 2008 on Science. That study showed autonomic responses (which are determined by amygdala reactivity) to fearful stimuli and attitudes toward socially protective policies.


Post a Comment

<< Home

eXTReMe Tracker