Shariff replies – the prosocial effects of religion

The recent review in Science on the prosocial effects of religion (Religious situations, but not beliefs, help foster trust) left me with a couple of questions. So I wrote to Azim Shariff, one of the two co-authors, and put them to him. Here’s my questions and his response:

Q. The first one is about the causal mechanism of religious priming in generating prosocial behaviour. In your review, you link this to the ‘unseen observer’ effect – the idea that people behave better if they think they are being watched. However, if this was the case then surely you would not expect secular primes to work as well as religious primes, and religious primes to work as well in non-believers as in believers. Perhaps the mechanism is more straightforward – perhaps simply subliminally reminding people that they ought to be good has the effect of prompting better behaviour?

A. With regards to the mechanism acting behind the priming-generosity effect, the issue that you raise is one that we’ve thought about quite a bit. First, there’s no necessary reason to think that both primes worked through exactly the same mechanism.Though that study in particular cannot distinguish between the two mechanisms that you suggest, the reason we’ve discussed the watcher hypothesis with reference to the religious primes is because it is consistent with a lot of the other studies that try to address the same phenomenon. We mention a few of these in the Science paper. The second reason is that suggesting that religion just prompts people to think that they ought to be good kinda begs the question. There needs to be some mechanism that explains the link between religion and good behavior that was found. Our suggestion, which we attempt to support in the Science paper, is that the development of these large, omniscient and morally involved deities was that mechanism. That’s also why in the original 2007 paper that emerged from those studies, we gave the religion prime finding so much primacy over that of the secular prime. To me, it’s not that surprising that reminding people of these institutions that are meant to enforce prosocial behaviour, makes them prosocial. With the religious priming, there’s another step to it, and that implicit connection between religion and good behavior is, I think, or particular interest.

Q. The second relates to the connection between prosocial behaviour and costly signalling. If religion increases trust not as a result of beliefs but rather as a consequence of environmental prompts, then how do behaviours linked to beliefs indicate trustworthiness?

A. The costly signalling stuff, I admit, is probably the least supported aspect of the theory. The reason we included it at all was because it is a necessary part of the argument (it should exist), and it enjoys some support the one investigation that Sosis and Bressler published about how the number of costly requirements predicted the longevity of religious communes. But more research needs to be done on that topic before we can claim to fully understand it. With reference to your specific question, my own theory is that the costly signaling works in two ways. In one case, there are a number of greetings and outward signals of religiosity that people use to both activate people’s religious prosocial tendencies, and to confirm that they are among those who should benefit from them. Thus, the (potential) cooperation partner acts as the environmental prompt. There some research to support this with charity solicitation work done in Israel where the researchers varied whether or not the solicitor was wearing a Kippah. The second way I think costly signalling works is, if there is enough saturation of your religion among your community, cooperation occurs whenever religion is activated (typically quite frequently), but non-religious people, that is, those who are visibly not costly signalling, become red flagged. So everyone is assumed to be a godfearing cooperation partner until proven otherwise. This is an idea that we’re starting to look at more in the lab – the atheist stigma. It seems that it is indeed driven more by moral distrust, than anything else. I do eventually want to look more into how religious people tend to ‘sniff out’ these freethinking freeriders. There’s a whole bunch of interesting anecdotes on the topic, so I think it makes for an exciting new avenue of research.