menu

Atheists should pray (in randomized, controlled trials)

Atheists should pray (in randomized, controlled trials) October 28, 2014

Screen Shot 2014-10-27 at 10.09.33 PM

Over at Cafe, Jess Whittlestone summarizes a paper on whether atheists should pray by Oxford philosopher Tim Mawson. The argument isn’t about expressing gratitude, co-opting religious rituals to cultivate mindfulness, or taking part in some metaphysically spooky form of positive thinking. Instead, Mawson thinks prayer is an important way to collect evidence about whether or not God exists. He writes:

If you think that there’s a non-negligible chance that God exists, and you think that whether or not he exists is important, then (a few caveats/technicalities aside), you should pray to God and ask Him to help you out.

The basic logic essentially mirrors that of a scientific experiment—if God show himself to people who answer prayers, then we can get evidence about whether or not God exists by praying and seeing what happens. This works both ways—if God answers prayers, than religious experiences serve as evidence that God exists while silence serves as evidence he doesn’t. Mawson explains:

If a theist prays and gets no apparent reply, that should decrease his or her confidence in God’s existence. It’s precisely because this sort of prayer experiment is open to all kinds of outcomes that theists, agnostics and atheists should all engage in it.

On the whole, I think this argument has some force to it. While I think it’s worth taking seriously, I have a few reservations.

Just looking out into the world, I can see many cases of mutually-incompatible religious experiences. That is, a lot of people pray and report religious experiences that can’t all be true. I have good friends who have turned to the book of John in times of struggle, prayed to God like it directed, and had a deeply intense personal religious experience. Others have the same feeling reading the Quran.

To me, this calls into serious question how much weight we should put into this kind of evidence, and Mawson seems to recognize its limitations. Whittlestone summarizes:

Additionally, the point of the experiment Mawson is proposing – seeing what happens if you try praying to God – isn’t about changing your belief in a black-and-white way, but rather a matter of seeing if your rational confidence in atheism changes.

I’m concerned, though, that if I were to have a deeply intense personal religious experience, it’d be difficult to treat it in anything other than a black-and-white way. I think a religious experience would give me a lot more confidence than I should have, and that worries me.

Unless I have good reasons from the get-go to discount the other religious experiences that are incompatible with the tradition I’m going to test out, an experiment like this would make me extremely nervous.

Mawson seems to preempt this point, but I’m not satisfied. Whittlestone summarizes again:

In science we do experiments all the time, and there’s always some possibility that doing the experiment will make us more likely to see a misleading effect – but this isn’t enough to make the experiment not worth doing. If it were, we wouldn’t be able to do any experiments – which would mean, among other things, no life-saving drugs or medical treatments. 

I think we have really good reason to suppose the false-positive rate for prayer is relatively high, since many different religious traditions report mutually incompatible religious experiences. A test isn’t useful if it gives the wrong answer too often, especially if the results of that test make you less likely to take disconfirming evidence seriously in the future.

Once we start treating prayer as a scientific way to test religious hypotheses, though, we could get around these issues by designing a real experiment. There should be no reason at all to constrain our explorations of prayer to personal experience, just like we shouldn’t have to test aspirin ourselves to learn if it works. Instead, we can use experimental manipulations and controlled trials. Taking a sample of atheists and randomly assigning them to receive daily 5-minute prayer in either one of two religious traditions would be far more effective than testing prayer out myself. If we wanted to rule out any potential effect of the act of prayer or contemplation, or even the existence of a pluralistic God that responds to all major world religions, we could have a daily 5-minute meditation session or a prayer to a God that was made up on the spot to serve as further controls.

While it might not make it past University ethics boards, I think the results of an experiment like that would be much more informative way to experiment with prayer than the kind Mawson suggests. Until then, I’m not sure how helpful prayer can be for an atheist to learn whether or not God exists.


Browse Our Archives