It’s almost like a moral decision, but not because no one will find out
Peep Show’s Jez on deciding whether to
have sex with his best friend’s fiancé’s mum.
Unlike the feckless Jez quoted above*, most people recognise that the immorality or wrongness of an act is not determined by whether we get caught or not. (And, conversely, a good deed doesn’t only gain moral value if others are around pat us on the back and say “Well done!”.) Yet people nonetheless do behave differently when their actions are publicly observable. In fact, so sensitive are we to social scrutiny that even subtle cues that we’re being watched can change our behaviour. In one study, simply putting a picture of a pair of eyes above an ‘honesty box’ used to collect money for milk in a university coffee room increased donations by almost three times compared with putting up a picture of flowers (Figure 1)1.
Findings like this have led to the idea that belief in God might promote moral behaviour in the same way. This ‘supernatural monitoring hypothesis’ suggests that to the extent that belief in God induces people to behave more honestly or fairly — and studies show it can2 — it does so through the same cognitive mechanisms by which social scrutiny enhances moral behaviour. According to this idea, it’s not so much that religious belief inculcates a robust sense of moral rectitude that believers adhere to because they recognise the inherent rightness of this moral code; rather, believing in God is akin to feeling like there’s a CCTV camera constantly focused on you that will catch you out when you’re bad, and get you punished accordingly.
The supernatural monitoring hypothesis doesn’t portray religious morality in the purest light: doing the right thing because you feel it’s the right thing is quite different from doing the right thing because someone’s looking over your shoulder, be that God or your neighbour. Yet irrespective of why people do good and refrain from doing bad, any mechanism that encouraged both would provide community benefits. There might be less cheating, more trust, and greater scope for working together for common goals as part of a larger moral community. One obvious mechanism for a social species like ours is punishment by peers who have found out that we’ve been violating moral norms — free-riding on the efforts of others or stealing from our neighbours — and experimental studies confirm that we are indeed punitive in just this way3. The importance of reputation and standing have given rise to a social psychology keenly attuned to how we’re seen by others, and what others see of us. When we’re watched, we’re better.
It’s not that much of a stretch to suppose that culturally created religious ideologies tap into the very same aspects of the social mind, and to achieve similar effects. And what better an invention than the idea of an omnipresent, omniscient, and moralising God to provide a 24/7 monitoring service, which even the most vigilant community could not achieve? Yoke this together with threats of damnation and eternal punishment for violations of divine commandments, and you could have a virulent set of moral memes on your hand: if belief in a moralising, punitive deity promoted altruism, group cohesion, and success in large-scale cooperative endeavours — from building houses to waging wars — then it could easily become widespread. And as survey of 186 societies found, belief in a moralising God is indeed correlated with measures of group cohesion and size4.
And this brings us to a paper in press at the Journal of Experimental Social Psychology5 by Will Gervais and Ara Norenzayan, in which they put the supernatural monitoring hypothesis to the empirical test. In three experiments Gervais and Norenzayan use various means of priming people to think about God, and look at whether this has similar effects to being under regular public scrutiny.
In the first experiment, Gervais and Norenzayan used a simple priming task to activate God concepts. This involved participants reading a list of 13 adjectives — such as ‘loving’ and ‘distant’ — and then rating them according to how well they apply to God. (This was called the God Prime condition.) Next, Gervais and Norenzayan wanted to gauge whether people felt a sense of being watched or under the microscope. So they had participants complete a Situational Self-Awareness Scale, which included three items that measure public self-awareness by asking respondents to rate agreement with the following statements: “Right now I am self-conscious about the way I look”, “Right now I am concerned about what other people think of me”, and “Right now, I am concerned about the way I present myself”. Responses to these questions are known to be sensitive to cues of social surveillance, such as the presence of a video camera, and so provide a means to assess the sense of being monitored and under scrutiny. Finally, participants completed the Intrinsic Religiosity Scale to assess how strongly they believe in God, after which the study sample was divided into High and Low Believer groups for later analysis.
Along with the God Prime condition, other participants were assigned to two others: a People Prime condition, in which participants rated the 13 adjectives on the extent to which they capture how other people view them; and a Control Prime condition, in which the adjectives were rated on their perceived frequency of use in everyday speech. Results across the three conditions (God, People and Control Prime), with participants broken down into High and Low Believers, were then compared.
The logic of this experiment is straightforward: if, relative to the control condition, the effect of the People Prime on public self-awareness is similar to that of God primes, then that would suggest that both primes have the same kind of psychological effects. And this is just what Gervais and Norenzayan found — at least among High Believers, for whom both People and God primes increased public self-awareness relative to control primes. So while watchful eyes make a difference for High Believers, perhaps surprisingly it didn’t matter whether they were God’s or Gary’s.
But you’re probably asking, “What about the Low Believers?”. Curiously, God primes reduced public self-awareness compared with control primes among this subset of people (which would naturally include out-and-out non-believers, who made up at least 16% of the total sample used in this experiment). And compared with controls, People primes didn’t make a significant difference either way for Low Believers. It should also be noted, however, that the Low Believer group showed higher public self-awareness in the control condition than High Believers, pointing to a difference in starting points between the two groups that could have some implications for where they end up after priming (Figure 2). It’s not clear what’s going on here, and I return to this issue below.
This first set of experiments used a relatively explicit priming task, in that the adjective-rating task asked people to reflect on the applicability of certain words in describing God. So in a second set of experiments, Gervais and Norenzayan used a more implicit priming task, one that involved unscrambling ten sets of five words in which one had to be dropped in order to form a four-word sentence. For the God Prime condition, five of the word clusters contained words commonly associated with God (spirit, divine, prophet, sacred and God). (Although the mention of God, as well as the other words, could be construed as being equally explicit as in the first experiment, debriefing questionnaires showed that no participants picked up on the fact that these were religious primes.) After completing this implicit priming task, participants completed the same Situational Self-Awareness Scale used before, and religiosity was assessed with a ‘yes’ or ‘no’ answer to “Do you believe in God?”. Results from the God Prime condition were compared with a control condition in which the scrambled words did not have religious connotations.
In this experiment, the implicit God primes increased public self-awareness relative to control primes. Yet unlike the first experiment, this effect that was not moderated by participants’ belief in God: so both believers and non-believers became comparably publicly self-aware after the God primes, compared with controls.
In a third experiment, Gervais and Norenzayan set out to replicate the preceding results using the same implicit priming tasks but employing a different method to assess the effect of God primes. So instead of looking at public self-awareness, they looked at whether people would be more likely to describe themselves in socially desirable ways, using the Marlowe-Crowne Social Desirability Scale. This requires participants to say whether certain statements apply to them or not — things like “I am sometimes irritated by people who ask favours of me”, or “No matter who I’m talking to, I’m always a good listener”. Both of these are socially desirable, but are unlikely to be completely true of most mortals. As such, the degree to which people tend to give socially desirable but delusional answers provides a measure of how concerned they are to give a positive presentation of their public self. After going through these tasks, participants then rated their belief in God on a scale from 0 (God definitely does not exist) to 100 (God definitely exists).
Just as God primes increased public self-awareness in the earlier studies, so too they increased socially desirable self-ascriptions relative to controls in this third round of experiments. However, unlike the second experiment (but like the first), this effect was moderated by how strongly participants believed in God: that is, God primes only led to more socially desirable responding among High Believers. And again like the first experiment, there was the unexplained result that Low Believers in the control condition showed higher levels of socially desirable responding that High Believers. In other words, Low Believers responded roughly the same after both God and control primes, and this level of socially desirable responding fell about halfway between the low levels seen with High Believers in the Control condition and the high level shown by Highs after God primes (Figure 3).
So what do these results add up to? It’s a bit of a mixed bag. Across three experiments, using both explicit and implicit primes, and using two separate measures of sensitivity to social surveillance, High Believers were found to respond to thinking about God in much the same way as everyone responds to social scrutiny. This supports the supernatural monitoring hypothesis specifically, and the more general idea that religion draws on the same kinds of cognitive machinery that are in place to deal with the secular problems of social living — you don’t need to posit specialised ‘God modules’ to explain why and how religious beliefs can affect mind and behaviour.
Some of the other results are a bit more confusing. Why is the effect of thinking about God on public self-awareness, and the tendency to describe oneself in socially desirable ways, absent or reversed among those with weaker beliefs — at least in two of the three experiments reported here? If this is a genuine phenomenon, what are the differences between the truly devout, tepid believers, and full-blown atheists that make a cognitive difference? And what to make of the fact that non-believers/low believers tended to show higher levels of public self-awareness and to present themselves in more socially desirable ways than believers in the control conditions? Could this be something to do with the way different categories of believer were constructed in these studies (sometimes using a simple ‘yes/no’ question about belief in God, at other times using a sliding scale or more sophisticated questionnaire)?
These are still early days in the cognitive science of religion. It would make life easier if there was a neat and conclusive story to tell here, but there’s no reason to expect that this is how things will pan out, and even less that initial forays in this area will be free from ambiguity. The more important point is to come up with good questions, of the sort addressed by Gervais and Norenzayan. The answers will be revised, clarified, and extended by others over time — and following this evolving story is where the fun’s at.
1. Bateson, M., Nettle, D. & Roberts, G. Cues of being watched enhance cooperation in a real-world setting. Biology Letters 2, 412–414 (2006).
2. Shariff, A. F. & Norenzayan, A. God is watching you: Priming God concepts increases prosocial behavior in an anonymous economic game. Psychological Science 18, 803–809 (2007).
3. Fehr, E. & Gatcher, S. Altruistic punishment in humans. Nature 415, 137–140 (2002).
4. Roes, F. L. & Raymond, M. Belief in moralizing gods. Evolution and Human Behavior
24, 126–135 (2003).
5. Gervais, W. M. & Norenzayan, A. Like a camera in the sky? Thinking about God increases public self-awareness and socially desirable responding. Journal of Experimental Social Psychology doi: 10.1016/j.jesp.2011.09.006
*You can watch the relevant scene here, starting at 08:47 — go on, it’s funny: