The Calvinball Defence

Another chapter in the pointless quasi-scientific investigation of intercessory prayer ends, the believers retreating to regroup and rethink, but never to let the results of their own studies actually affect what they think, except to confirm what they already believe according to that most anti-scientific throwback of human psychology: faith. NYT:

Bob Barth, the spiritual director of Silent Unity, the Missouri prayer ministry, said the findings would not affect the ministry’s mission.
“A person of faith would say that this study is interesting,” Mr. Barth said, “but we’ve been praying a long time and we’ve seen prayer work, we know it works, and the research on prayer and spirituality is just getting started.”

We haven’t seen it work, we don’t ‘know’ it works, we have indeed been praying for a long time, and the research bumbles on fruitlessly, wasting time and money which might better be spent actually treating people. At least the research is relatively cheap. But, assuming that money is thrown in the direction of prayer, it might be far more productive to look at the positive effects of prayer on the person praying, based as they are on solid cognitive mechanisms, the better to understand the actual processes involved, and therefore the better to separate them from the appeal to an imagined deity, in order to make them somehow accessible from a non-theistic world-view.

I entirely understand the impulse among some theists to demonstrate objectively the effectiveness of prayer. Partly it comes from a deeply-held conviction that prayer does indeed work (Barth’s assertion that ‘we’ve seen it work’ is genuine, but based on nothing other than anecdotal grounds); and if it works, why shouldn’t it be shown to work? Partly, I assume, it comes from some pretty powerful wishful thinking. Perhaps it also comes from a well-intentioned desire to reconcile science and religion for themselves, wishing as they do to see them as capable of co-existing.

Personally, I don’t see anything which ought to make religion exempt from scientific enquiry. Science isn’t a world-view; it’s a process by which we find out how things are. If prayer does work, then, should we want to, we can at least try to find out how and why: that’d be science. Theologies make claims all the time which depend on physical processes, and are therefore investigable as such: If the existence of a soul is claimed, then where does it exist? What form does it take? How does it endure after death? If transubstantiation is anything other than a metaphor, what processes are involved?

However, and it’s a big however, the appeal of a subset — perhaps a very large subset — of religious claims to the supernatural for their mechanism places them outside of what the scientific method can access. It’s one thing to claim that intercessory prayer works, for example — perhaps we don’t know how, perhaps we don’t care how, but it does work. In such a situation the claim is merely about the effect, which then makes it reachable by testing. (James Randi is dead right to always insist on testing whether some supernatural claim can be demonstrated, before spending as much as a moment considering how it might have been achieved.) It’s quite another thing, though, to claim that intercessory prayer works because of some intermediary deity.

The problem doesn’t arise if the claimed effect in fact shows up. Well, it complicates things: if intercessory prayer were shown to be real — allowing the rather large assumption that all other variables were successfully removed from the experimental method — then the appeal to a deity for the mechanism might obstruct the investigation. If you believe that a god is the solution, why look any further? (Besides, if intercessory prayer did work, and a god was the mechanism, how long would you spend looking for something more mechanistic?)

But when the claimed effect doesn’t, in fact, show up, as in this case, the fallback position of someone emotionally invested in a positive result can come to be what might be called the Calvinball Defence. It’s where you get to change the rules of the game any time things are going against you. Intercessory prayer not working? Well, perhaps god refuses to be tested in this way. From the St. Petersburg Times:

Religious leaders questioned whether the prayers were appropriately worded and whether those praying were really moved by the spirit.
“Prayer, particularly intercessory prayer, is asking God for something,” Plazewski said. “The reality is, “No’ is also an answer.”

Or maybe the people praying didn’t have enough faith. The St. Petersburg Times again:

The Rev. Abe Brown, pastor of the First Baptist Church of College Hill in Tampa, cited James 5:16. The Bible verse suggests that God only answers the prayers of the righteous.
Brown also said God answers prayers based upon believers’ faith.
“The one that’s sending up the prayer has to believe that God is going to answer,” Brown said.

Or maybe, I dunno, the prayers were just in the spiritual equivalent of a cellphone dead-zone:

Monsignor Laurence E. Higgins, pastor of St. Lawrence Catholic Church in Tampa, believes in the power of intercessory prayer.
“Maybe the people weren’t praying very hard,” Higgins said of the study. “I have no doubts that intercessory prayer works, (just) not all the time.”

These are, of course, spookily reminiscent of the weasely excuses given by charlatan ‘mediums’ like John Edward, Sylvia Browne and such, when their theatrical schtick is rumbled. These people live and breathe the Calvinball Defence.

In scientific terms, proposing a god as the mechanism for some process blows any claims of good methodology wide open. In the case of intercessory prayer, where — one assumes — the claimants propose that god is the mechanism by which the effect of prayer is transmitted (or not, depending on the deity’s whims and schedule that day) from the pray-er to the pray-ee, one can imagine a rather good analogy in which some new drug is to be tested. The doctors have designed a state-of-the-art double-blind trial. However, when the randomly-chosen drugs/placebos are to be given to the randomly-chosen and properly-controlled participants, the job of delivering them is delegated to some grad student in the lab. Perhaps he’s a little preoccupied, perhaps has a mischievous, capricious streak. Unbeknownst to the doctors, the grad student loses some of the drugs. Perhaps some of them get mixed up. Maybe he leaves them in a cab and, embarrassed about admitting his screw-up, replaces them with aspirins. So far as the doctors are concerned, the experiment is water-tight. But the mechanism is entirely unreliable.

A double-blind trial depends on neither the conductor nor the participant in the study knowing whether they’ve got the real drug or not. They’re both blind. However, if some study of intercessory prayer proposes a god as its mechanism, it follows that it’s also being proposed that the god knows which people are praying for which other people. It follows from that that the double-blindness of the trial no longer applies. It sounds facetious, but it’s not: you can’t have a double-blind trial in which one of the participants is omniscient and/or omnipotent. The only sense in which the double-blindness might be maintained, would be if the proposed god behaved in an entirely deterministic way to the prayer. That would entail that it wouldn’t in fact matter whether the god knew which people were being prayed for; the god would respond as predictably, as mechanistically, as without volition, as a tablet of ibuprofen. Obviously, neither the proponents nor the critics of such a study would make such a claim.

To be clear: obviously the failure of this intercessory prayer study says nothing about the existence of a god, nor about the effectiveness of prayer. One might conclude, though, having sharpened Occam’s Razor, that it’s reasonable to assume that intercessory prayer has no basis in reality until there’s a considerable weight of evidence to the contrary.

The Calvinball Defence only takes effect in the absence of a positive result. It’s entirely possible to obtain a positive result in a trial for which one can propose no workable mechanism (yet) — though of course that would make it rather harder to be sure that all of the independent variables had been nailed down. This intercessory prayer study might well have produced a positive result. What then? Well, lots and lots and lots more study. The ramifications of the positive result of a study of this type — assuming a sound experimental method, which is assuming a lot, but let’s go with it — would be vast; the study would require much replication before its conclusions were taken seriously, and that’s exactly as it should be. It didn’t produce a positive result, though. Hence, the Calvinball Defence.

I can’t help wondering, though, what the proponents of, and believers in, intercessory prayer would conclude from some hypothetical positive evidence for its effectiveness. What would it mean if someone who was (genuinely, honestly) prayed for had a better way of things than someone who wasn’t prayed for? What would it mean if someone who was prayed for by a larger group of people was more likely to survive serious illness, or to get better more quickly? If we assume that the god prayed to isn’t the non-volitional ibuprofen pill; if we assume — as we assume the prayer probably assumes — an omniscient, omnipotent god, making choices about how to act; then what should we conclude from a hypothetical situation in which prayer gives someone a better chance? The omniscient god would know exactly to what extent someone was deserving of their help — according to their own sense of morality — without the nagging from a congregation, however sincere. Would what’s left be a kind of popularity contest? Would god be like the switchboard for some American Idol-ish game show driven by public appeal?

I’m glad that there’s no evidence that intercessory prayer works, because I’m glad that it doesn’t, in fact, work.