• Monthly
  • $25
  • $50
  • $100
  • $other
  • use PayPal
KEEP COUNTERPUNCH AD FREE!

We are nearing the end. But if we don’t reach our modest goal, we will have to cut back on content and run advertisements (how annoying would that be?). So please, if you have not done so, chip in if you have the means.

FacebookTwitterRedditEmail

(Un)Willingness to be Puzzled

It has been established for some time that individuals with strong political beliefs tend to display greater confidence in them than individuals who have less strong beliefs. Despite both types of individuals holding specific, concretised views, there is a clear positive correlation between belief strength and belief confidence. Yet, until recently it was unclear whether this dynamic is due to an overconfidence bias or a failure of metacognition; meaning, does the correlation exist because individuals adopt a self-portrayal of unwavering confidence (a notion closely related to ‘virtue-signalling’), or does it exist because individuals, for whatever reason, lack an insight into the veracity of their beliefs and refuse to engage in a process of verification.

A recent study published in Current Biology   by Max Rollwage, Raymond J. Doland and Stephen M. Fleming at University College London explored this issue. They demonstrated via a low-level percetual task (abstracted away from any specific social or cultural issues) that individuals with radical political beliefs – both left- and right-wing – showed reduced insight into the correctness of their choices and less sensitivity to post-decision evidence. Hence, they displayed a generic resistance to revising their own mistakes, and were less willing to be corrected in the face of counter-evidence. The author’s use of a low-level perceptual task (estimating the number of dots on two images and estimating which one had more) helped them to control for effects of existing knowledge or motivational factors. Quite apart from broader socio-cultural considerations, of the kind invoked by postmodernist scholars and many on the woke Twitter left, the authors point to “a generic resistance to recognizing and revising incorrect beliefs as a potential driver of radicalization”.

It may be the case that individuals who violate Noam Chomsky’s central maxim (that one should always adopt a “willingness to be puzzled”) do so as a direct causal consequence of them exhibiting this metacognitive failure. It may also be the case that, for other individuals, the connection is less causal but more exploitative: That is, others may already be primed (by independent cognitive factors) to hold fast to their beliefs in the face of counter-evidence, and they merely buttress this process via their metacognitive failures. Following a core insight from the research of neurobiologist Karl Friston, brains of all types (and belonging to all different species) like to minimise surprise and maintain current representational states as much as possible, so as to reduce cognitive/neurophysiological ‘effort’. Metacognitive failure is therefore an efficient tool of Friston’s ‘free energy principle’ closely tied to surprise minimisation, which refers to how biological systems maintain their existing states.

I suspect that, although individuals will of course differ, the general trend will likely be that metacognitive failure arises first, not least because low-level perceptual competence is something well-developed and matured by childhood, long before (most) individuals have reached the age of reason and established concrete political views. Regardless, these findings should create a degree of concern. The lack of a willingness to be puzzled amongst radicals, from both the Left and Right, becomes even more troubling when we consider that it likely has an adaptational motivation, from the perspective of evolutionary psychology: It does not impact rates of survival and life quality on a local timescale whether or not one subjectively approves of the Keystone Pipeline, or of Jeremy Corbyn, or what one’s opinions are of transgender changing room policies, or of Hollywood diversity quotas, or of Milo Yiannopoulous’s fall from grace, and so serious pressure to ensure one that develops a well-reasoned stance is relatively absent.

Since rates of survival are very often not contingent on these things (even though they may be contingent on one’s actionsin relation to them, as when individuals decide to act on their beliefs via protest, non-violent civil disobedience, violent action…) then it is less important for individuals to consolidate their political beliefs in coherent ways than it is for them to consolidate their beliefs about other issues more directly impacting daily life. Other important cognitive factors, like whether one tends to be an intuitive or analytical thinker, likely play a role, with intuitive thinking being more susceptible to initially forming radical views – though this factor has little to say about what happens after the initial belief-formation, how they beliefs are maintained and evaluated, and so on. More broadly, no single factor will suffice here, and multivariate analyses will naturally be required to derive the numerous consequences of radical political belief formation. Nevertheless, metacognitive failures may well form a major part of this analysis.

One of the most interesting aspects of these findings is that they apply equally to Left and Right: Left-wingers cannot respond that individual-external factors (socioeconomic factors, etc) are to blame for this, but neither can right-wingers respond that certain individual-internal factors easily within the grasp of personal choice (e.g. a lack of appropriate self-development or personal responsibility) are to blame, given that metacognitive ability develops and stabilises through genetic factors. Perhaps activist meetings should therefore prioritise the teaching of metacognitive skills.

A range of survey data from the United States over the past decade has also confirmed that political disagreements tend to be largest among the most (formally) educated opposing partisans, exhibiting greater political polarisation than those who are less educated. Ben Tappin at Royal Holloway has recently collected similar data from a range of European countries, showing that for the topic of immigration the same polarisation dynamic emerges. A central hypothesis that emerges from this data, which is currently well-supported, is that human cognition appears to be biased to form beliefs that correctly signal to peers whose ‘side’ one is on, since a failure to signal can often lead to social ostracism. Since more cognitively sophisticated individuals can more readily reason about evidence pertaining to political issues, they are more likely to exhibit polarised beliefs. It is not difficult to think of ways to expand this important research project – nor is it difficult to think of the many ways its findings will be shaken off by the woke and hidebound.

Elliot Murphy teaches in the Division of Psychology and Language Sciences at University College, London.

FacebookTwitterRedditEmail