April 9, 2015

Suspicious Minds? Mentalizing, Religious Hypocrisy and Apostasy

Put simply, ‘Theory of Mind’ (ToM) and its associated near-cognates (mentalizing, mind reading, social cognition) refer to the socially indispensable human capacity to attribute mental states to others, thereby comprehending them as agents whose behaviour is driven by internal motivations. Generally this capacity is thought to arise in a predictable manner in the development of neurotypical individuals. Of course, this picture of ToM is not without its controversies: there is currently debate about whether, at least in more subtle aspects, it may actually be a culturally variable acquired skill; questions remain about how exactly it ‘works’ (i.e. through simulation, through the construction of naïve theories, or both); and the entire construct has been criticised as needing fractionation into more specific sub-mechanisms (Schaafsma et al, 2015). However, the main point here is to assess how the ToM construct has been applied in the cognitive science of religion (CSR), and in this respect it has been fertile indeed.

In his RSP interview Dr. Gervais gives us a clear and concise tour of many of the fundamental ways in which ToM is put to work in CSR (for a more in-depth treatment, see Gervais, 2013). The essential point is that we humans are so very primed to think in terms of agency that we overdetect it in (or overapply it to) our environments and this leads to the success of supernatural agent concepts which trigger the misattribution of mental states in ways that are intuitively compelling. ToM thus doesn’t ‘produce’ religious beliefs per se, but it does mould the forms they are likely to take; in the putative epidemiological struggle of concept against concept, we have a content bias to prefer those harnessing notions of agency. One branch of evidence for this comes in the form of Dr. Gervais’ own work, which suggests that there is a small but significant correlation between mentalizing fluency and willingness to entertain belief in supernatural agents (Norenzayan et al, 2012).

Content biases are not the only point of intersection between ToM and belief in supernatural agents, however. In some cases ToM itself may be purposefully manipulated through forms of practice to produce religiously ‘meaningful’ experiences. For example, Luhrmann’s ethnographic work describes the process whereby charismatic ‘Vineyard’ evangelicals painstakingly learn to ‘misrecognise’ some of their own cognitions as external thoughts channelled into their heads by Jesus, thereby ‘hearing His voice’ (Luhrman, 2012). Furthermore, as Gervais himself observes, content biases can explain why certain concepts are intuitively attention-grabbing, but not why people commit to the concepts they do (this is known as the Zeus problem – see Gervais & Henrich, 2010). In certain circumstances people may even commit to ‘concepts’ that cannot be grasped at all; Sperber’s largely ignored theory of the ‘guru effect’ – the tendency, visible in some religious contexts, to meta-represent recondite utterances as profoundly meaningful if they emanate from esteemed sources of authority – marks one interesting potential bias enabled by content that defies successful representation as opposed to content that ‘sticks in the mind’ (Sperber, 2010). More generally, context biases – namely biases to selectively attend to information based on features of its source – are also a factor of specific relevance to religious transmission. How many other people in our social circle hold the belief? Did we hear it from someone prestigious who is likely to be a source of fitness enhancing information? Did we hear it from someone we can trust – and how do we know they believe what they say they do?

One influential context bias proposed in the CSR literature is the CRED (Credibility Enhancing Display – Henrich, 2009): the idea that, due to the manipulative potential inherent in language, cultural learners have evolved the precautionary tendency to scrutinise cultural models for behavioural confirmation of commitment to stated beliefs. Accordingly, belief transmission, particularly in the case of empirically unverifiable beliefs, is strengthened when models ‘practice what they preach.’ Thus religious beliefs accompanied by costly actions the believer would not undertake if they did not believe what they said they did – painful or time-consuming rituals, charity, celibacy, martyrdom – will transmit more successfully than those that lack such trappings.

Suspecting another’s internal motivations of diverging from their stated intentions is a mentalizing operation if ever there was one. But if such a bias is exploited via CREDs to facilitate religious transmission, might there not also be scenarios in which similar capacities serve to actively undermine belief? Is irreligion aided simply by the absence of contextual cues to religiosity, or might there also be contextual cues to irreligion? As opposed to CREDs, my own research investigates ‘CRUDs’ – credibility undermining displays. In particular, I am interested in how displays by religious paragons which contradict expressed statements of belief may be uniquely corrosive to the religious certainty of believers. One does not need to look for long to locate examples of the connection between the attribution of insincerity to religious paragons and religious scepticism. New atheist forums are frequently aflame with outrage at perceived religious hypocrisy[i], and it often also features in atheist ‘conversion narratives’ (Wright et al, 2010). The steep and ongoing decline in Catholicism is often partially attributed to the clerical abuse scandals, and in particular the promulgation of such scandals in the media has been linked by sociologists to an acceleration in Irish secularisation since the early 1990s (Donnelly & Inglis, 2009). Indeed, modern methods of information exchange, through their reach and permanence, compound the problem of scandal for religion by tapping into the regulatory pan-human phenomenon of gossip: Mormonism, for example, seems to be currently experiencing a crisis of faith due to online revelations about Joseph Smith’s amorous adventures. Indeed, historically speaking, the credibility undermining display was effective enough to have been used as a form of counter-propaganda, at least in reported form; mediaeval anti-heresy tracts revelled in such rhetoric, describing heretics[ii] as ‘wolves in sheep’s clothing’ who preached chastity, compassion and asceticism while secretly indulging in orgies, rendering infants down into ceremonial black paste, and drinking toad excrement.

Perhaps part of the reason the relationship between mentalizing and suspicion of religious truth-claims has yet to come into focus in CSR may be due to an unspoken tendency to view ToM as an impartial would-be ‘mirror’ of sorts; though it errs and though its sensitivity may be extremely overtuned, we ultimately evolved the capacity in order to acquire optimally accurate representations of our own and others’ psychological motivations in order to facilitate cooperation. Much research in social psychology, however, would suggest that people can in fact be unkindly biased in the mental states they attribute to others versus those they attribute to themselves (Monin & Merritt, 2010). Moral failings are far more often seen as the results of malign intentions if performed by others, others’ pieties are often written off as the result of self-serving motivations, while individuals frequently overestimate the depth of their own moral commitment. It might be said that ToM seeks truth – insofar as it is useful for action. ToM is surely better seen as intertwined with and influenced by a range of other factors prioritising such phenomena as moral policing and deception-enhancing self-deception, frequently not so much an accurate gauge of others’ motivations as a cautiously (or opportunistically?) harsh one. Given these considerations, we might wonder about the relative potency of CREDs versus CRUDs. Such biases should mean that even a fairly insignificant act could trigger a CRUD warning; unlike with religiously bolstering displays, there is no ‘costliness’ barrier between an act of religious hypocrisy and its potential effects on belief. In fact, there may on the contrary be a heightened sensitivity to such transgressions.

Of course, there are many complexities to be teased apart here: Are some believers more prone to scepticism upon witnessing contradictory statement/behaviour pairings than others, and why might this be so? If CRUDs are so potent, then how do various religious traditions cope with them, and are some particularly vulnerable (see, for example, Wollschleger & Beach, 2011)? Might CRUDs affect theistic belief per se or only religious affiliation? And how does the issue of harm combine with religious hypocrisy in producing any putative effects on belief and/or affiliation (i.e. eating fish on a Friday versus abusing children)? It is possible that if religious scandals/hypocrisy can be a partial driver of religious decline, there may be at least two separable but intertwined psychological effects going on: CRUD-based socio-cognitive belief-scepticism on the one hand and institutional disaffiliation stemming from moral contempt on the other.


Donnely, S. & Inglis, T. (2009). “The Media and the Catholic Church in Ireland: Reporting Clerical Child Sex Abuse.” Journal of Contemporary Religion, 25:1, 1 – 19

Gervais, W. M. (2013). “Perceiving Minds and Gods: How Mind Perception Enables, Constrains and Is Triggered by Belief in Gods.” Perspectives in Psychological Science 8(4), 380 -394.

Gervais, W.M. & Henrich, J. (2010). “The Zeus Problem: Why Representational Content Biases Cannot Explain Faith in Gods.” Journal of Cognition and Culture, 10, 383 – 389

Henrich, J. (2009). “The evolution of costly displays, cooperation, and religion: Credibility enhancing displays and their implications for cultural evolution.” Evolution and Human Behaviour, 30, 244 – 260

Luhrmann, T. (2012). When God Talks Back: Understanding the American Evangelical Relationship with God. Vintage.

Monin, B. & Merritt, A. (2010). “Moral hypocrisy, moral inconsistency, and the struggle for moral integrity.” M. Mikulincer & P. Shaver (Eds.), The social psychology of morality: Exploring the causes of good and evil, Herzliya Series on Personality and Social Psychology, Vol. 3, American Psychological Association.

Norenzayan, A., Gervais, W. M., & Trzesniewski, K. H. (2012). “Mentalizing deficits constrain belief in a personal god.” PLoS ONE, 7, e36880.

Schaafsma, S., Pfaff, D., Spunt, R., & Adoplhs, R. (2015) “Deconstructing and Reconstructing Theory of Mind.” Trends in Cognitive Sciences, 19(2), 65-72

Sperber, D. (2010). Sperber, D. (2010). The Guru Effect. Review of Philosophy & Psychology, 1 (4), 583-592

Taylor, C. (2007). A Secular Age. Belknap Harvard

Wollschleger, J. & Beach, L. (2011). “A cucumber for a cow: a theoretical explanation of the causes and consequences of religious hypocrisy.” Rationality and Society, 23 (2), 155 – 174

Wright, B., Giovanelli, D., Dolan, E. & Edwards, M. (2011). “Explaining Deconversion from Christianity: A Study of Online Narratives.” Journal of Religion and Society, 13, 1-17

[i] Interestingly, this often includes both the hypocrisy of believers and also God’s own hypocrisy, i.e. theodicy.

[ii] And of course heretical movements have often been partially attributed to Church failings – simony, nepotism, corruption, venality and so on. I don’t assume here that the CRUD leads straight from orthodoxy to atheism by any means, but rather to scepticism about the expressed representation; historical and cultural context is key, and where theism is the inescapable idiom of the age, schism is the more likely outcome. The link to atheism becomes possible where it has come to exist as an option (i.e. Taylor, 2007).


Join the discussion

Your email address will not be published. Required fields are marked *