Some weeks ago I attended a lecture by Daniel Dennett at the Oxford Union on religion. As expected, it was a lively presentation that predicted the demise of religion. However, one matter that started me thinking was how Dennett concluded his lecture: he ended by pondering what we might do with all the redundant places of worship once his prophecy was fulfilled. His suggestion was that they might satisfy a secular purpose, as places where the community might come together to address the novel challenges of the modern world. I started me thinking as I wondered whether a belief in religion might be better than atheism for attaining this, or any other, goal. Some, such as Jonathan Haidt (The Righteous Mind (2012)) have suggested that religion is a particularly effective force for bringing people together.

I would like to ask a broader question, which is whether religion is better than atheism for attaining any particular normative goal. The reason for this is that confining the question to which is best for promoting cohesion begs the question as to why cohesion is important. To attempt to avoid this problem, one could pose the question more broadly: given any chosen normative goal, is religion or atheism more condusive to attaining it?

I should probably add that I am an atheist myself, a great fan of Dennett, and very sceptical of religion. As such, I would suggest that I do not have an axe to grind, or at least the type of axe that Dennett worries about (Breaking the Spell (2006) p 32). Nonetheless, I struggle with the reasons behind the proposition that atheism is better than religion for attaining normative goals.

One consideration could be that religion causes people to believe things that are not true. Richard Dawkins, another of the ‘Four Atheists of the Apocalypse’, points out that the ‘beneficial effects in no way boost the truth value of religion’s claims.’ (The God Delusion (2007) p 194). Dawkins then quotes George Bernard Shaw: ‘The fact that a believer is happier than a sceptic is not more to the point than the fact that a drunken man is happier than a sober one.’ But I’m not convinced: it seems possible that a believer could be more likely to attain [insert your chosen normative goal here] than a sceptic, even if he believed things that were not true.

Believing things that are true rather than things that are not true could have some value. But the topics on which beliefs diverge seem quite peripheral: how much difference does it actually make to a person’s behaviour whether they believe that the life was created by a god rather than by a process of evolution by natural selection? Equally a belief in tenets of religion that cannot be true (talking snakes, virgin births, resurrection, etc) may have little actual impact on how people live their lives.

By contrast, it is difficult to exclude the possibility that untrue beliefs have a function that is not linked to their truth. Perhaps what is important is not whether religious beliefs are true, but that they are specific to that religion. That is, religious beliefs act as a ‘badge of identity’ for that religion that is difficult to fake. It would be easy for adherants to that religion to learn the weird and wonderful tenets of that religion, but difficult for outsiders. Thus beliefs could be the means of working out whether a person is safe to trust. If this was their function, they would necessarily have to be arbitrary so that they could not be worked out through logic.

This uncertainty highlights the fact that we simply do not know enough about religion or atheism to weigh up their effects.

Dennett’s solution to this, propounded particularly in Breaking the Spell (2006), is to recommend investigation. In so recommending, he notes the concern (echoed in the title of the book) that so doing would undermine the advantages that religion provides. He concedes that no one knows the answer to this (p 15) but insists that such questions can only be answered in the light of all the evidence:

‘Eventually, we must arrive at questions about ultimate values, and no factual investigation could answer them. Instead, we can do no better than to sit down and reason together, a political process of mutual persuasion and education that we can try to conduct in good faith. But in order to do that we have to know what we are choosing between, and we need to have a clear account of the reasons that can be offered for and against the different visions of the participants.’ (Dennett (2006) p 14.)

The difficulty with this approach is that I don’t think it answers to the concern. If we don’t know what religion does, it’s hard to be certain that investigating it will not undermine its effects. The well recognised defensiveness of believers, combined with the fact that few scientists are religious (Dawkins (2007) pp 125 – 128) suggest that this is not a fanciful worry.

Though I am sympathetic to Dennett’s proposal, it doesn’t seem to follow as a matter of logic. To me it seems like a leap of faith.