A: The Harvard prayer study is a real challenge to my faith.
B: What, this?
A: Oh, I’d never read it, but thanks… This is such an odd way to run a study. The pastors themselves divided two groups and then prayed for one and not the other. But it’s hard to imagine a prayer that would be more heartless and invalid than one that actively refuses to pray for some group in order to see if they do worse.
B: Why “invalid”?
A: Prayer is a lifting of the heart to God. I don’t see how that’s happening if your actions are saying “I’m gonna actively ignore this group of people and see if that harms them”.
B: Okay. But I doubt there’s any way to make this work.
A: Not even if you simply asked questions about whether the patients themselves were praying, or believed that others were praying for them?
B: No, because I don’t see any way to establish what it would mean for the prayer to work. Say a statistically significant group of persons who prayed to get better got better. Is this evidence that God responds to prayer or that prayer has a placebo effect?
A: But what if those who prayed actually did worse or no better?
B: Who knows? Maybe we take that as evidence that prayer doesn’t work. But all this would mean is that it’s determined a priori that we won’t find evidence of the effectiveness of what a believer calls prayer.
A: So if you’re trying to validate or invalidate what a believer calls prayer, this experiment isn’t going to do it. The validation hypothesis is undefined.
B: Right. I’m not sure whether it would be underdetermined in every attempt to experiment with prayer, but anything close to this experiment would be. I suppose that they wanted to have “third party” prayer to avoid the placebo effect, but this demands that those who pray level a de facto curse on the control group. Who would expect a prayer like that to make a difference?
A: You wonder how the command not to put the Lord to the test fits into this.
B: Right. The atheist take will be “yeah, don’t ever check whether what you’re doing is working! That’s a real strategy for success!”
A: I can imagine Boghossian saying that.
B: He’s who I had in mind. It seems like he’d be giving his opponents a more charitable reading if he discussed miracle claims in Catholicism: the Tilma, the blood of Januarius, the healings at Lourdes, or even the various intercessions that are offered as evidence in canonization. These are actually times when the Church claims where looking for evidence that there was divine intervention does not involve putting the Lord to the test.
A: It seems like you’re giving “you shall not put the Lord to the test” a logical sense – as soon as you are testing the Lord, you adopt a viewpoint that requires there can be no meaning to “passing the test”.
B: This might be a fault in all rationalist schemes. If your faith is founded on any rational test you can only believe it as far as reason can go. But faith demands assent to the unlikely and the unprecedented, while reason has to have a bias against these things. The closest reason can get to faith is to say “yeah, we can’t rule that out” or maybe “yeah, those sources were convinced this faith claim was true.” But assent to anything requires more than “I guess I can’t rule it out”.