Post hoc, ergo propter hocMarch 15, 2009 — Deacon Duncan
In my post on “Which God?“, I discussed cl’s hypothetical case of a man being decapitated, and having his head miraculously re-attach to his body, in the context of some Christian praying for him. To illustrate that this would not really be reliable evidence for God, I said:
Let’s say that during the hour the victim was decapitated, some saffron-robed monk wanders by and begins to pray, “Oh great Buddha, have mercy on this poor soul and heal him of his decapitation by your divine grace.” The rest of the story remains the same: after an hour, the head reattaches and the man walks away unharmed. Would this be evidence that Buddha is really God?
I think a lot of Buddhists would be fairly surprised if that were the case. But notice, the actual evidence of the miracle itself is no more specific than it ever was. The facts pertaining to the actual “recapitation” are exactly what they were before. All the prayer has really done is to create a context in which we might be prejudiced to prefer one superstitious attribution over another.
Commenter cl objects:
That’s incorrect. The facts are not exactly the same as before. In the first hypothetical scenario, we had no Buddhist praying, hence no reasonable grounds to connect the incident to a Buddhist prayer. In the second example, we have stronger evidence – the event occurred after a Buddhist prayer – providing us with a verifiable connection that strengthens preliminary justification for the possibility that Buddha performed this particular miracle.
Our friend cl has fallen into the fallacy known as post hoc, ergo propter hoc. “After this, therefore because of this” is an age-old tendency in human thinking that deceives a lot of people because it resembles scientific thinking in some respects. But it is still a fallacy nevertheless.
Let’s change the story again: this time, instead of a Buddhist monk praying, it’s a hippie eating a veggieburger. As soon as the hippie swallows the last bite of the burger, the head reattaches itself. Do we now have reasonable grounds for concluding that the act of eating a veggieburger has hitherto unsuspected miraculous powers? I really don’t think even cl would draw that conclusion. And yet it’s the same thing: we’re saying A happens, and then B happens, and therefore A must have caused B. In the first case, “A” is a Buddhist monk praying; in the second, it’s a hippie eating a veggieburger.
What tends to fool us is a malfunction of our psychological ability to recognize significance. When the Buddhist monk prays for the decapitation victim to be healed, we understand the meaning of what he is asking for. When the head “re-capitates,” we associate the meaning of the prayer with the description of the amazing event, and infer a cause and effect relationship that isn’t necessarily there.
Notice I’m not saying that I know the relationship is not there, or that we know Buddha didn’t heal the victim. I’m talking about the validity of the techniques we use to draw our conclusions. It’s not that the conclusion is necessarily false, it’s that the procedure used to reach the conclusion is necessarily fallacious and unreliable.
Let’s try another variation: this time, the veggieburger-eating hippie shows up again, and says, “Dudes, chill, it’s no big deal. As soon as I finish this veggieburger, that guy is going to come back to life.” He finishes his burger as before, and the miracle happens again. Now did eating the burger cause the miracle? We’ve got the verbal connection this time: the hippie said that the miracle would happen when he finished the burger. But did that mean that burger eating caused the miracle, or might it perchance have meant that the hippie was really Loki in disguise, and he just wanted to finish his food before he invoked his other, unrelated supernatural powers?
Verbal significance can mislead us into thinking that the post hoc fallacy is not really a fallacy. But it always is. The way to avoid the post hoc fallacy is to employ valid scientific techniques. The difference between a scientific explanation and a superstitious attribution is that the scientific explanation presents its mechanisms and initial conditions in sufficient detail that you can determine analytically what real-world, observable consequences would result from these conditions and mechanisms in operation. Superstition merely attributes things to supernatural causes via some kind of unpredictable magical poof.
So let’s try and do this scientifically. Let’s go back to the original variation, and suppose that we’re going to explore whether it is more scientific to conclude that Buddha intervened to reattach the man’s head to his neck and restore him to life. And let’s compare that to some other hypotheses, such as the hypothesis that the victim is some kind of über-highlander whose body has magical powers to heal itself after a brief interval, or the hypothesis that the “victim” was actually just a holographic projection of some kind, or that the witnesses were all exposed to a psychoactive gas that made them hyper-suggestible, and then prompted, via a concealed loudspeaker, with the details of the event they were to supposed to have witnessed.
Starting with the last hypothesis first, let’s predict what real-world consequences would result from the mind-altering gas. There might, for example, be some kind of residue, and there would certainly need to be some sort of container for delivering and releasing the gas. We could search the scene for evidence of such a mechanism, and the hidden loudspeaker, but we might not find anything. We would also expect that people outside the target area might have been affected by the drifting gas, so we could search for signs that other people were strangely affected, in diminishing degree the farther you got from the initial exposure zone, and generally drifting in the direction of the prevailing winds at the time. Also, a hypnotic voice can only supply so many details, so if we interview the witnesses individually, we ought to find unanimous agreement regarding the prompted details, but only random and inconsistent reports regarding any details that were left up to the imagination of the individual “witness.”
The holograph, meanwhile, would be testable in that there would be no visible signs of an actual decapitation (i.e. no bloodstains on the ground). We could test this hypothesis by seeing if there was any physical contact with the victim before or during the event. (Yeah, this is the easy one.)
The über-highlander hypothesis is a bit harder to test, at least ethically. You could try chopping off his head again, but it would really suck if he failed to pass. We could at least predict that his body would be different in some way from an ordinary human body, and give him a physical. An interview might also be revealing. This one would be hard to test, though, because there’s no real good reason for him to be magically immortal. The explanation itself has a lot of explaining to do, starting with the question of how a dead body can function well enough to heal itself, let alone move the head back to the neck without touching it.
But now we’re back to Buddha. What do we know about Buddha that will allow us to predict what consequences ought to result from praying to him? What observations have we made that allow us to draw reasonable inferences about likely outcomes? What mechanisms do we see operating in his actions that allow us to draw specific, testable conclusions about what consequences we would reasonably expect to see if those mechanisms operated in the real world?
We got nothing. Scientifically, we can’t get this one off the ground (and really, the same goes for über-highlander as well). Superstition doesn’t work forwards, it only works backwards. Once something happens, then you can assign the credit to your supernatural cause du jour. But going forwards, anything could be the “predictable” outcome. Buddha might choose to respond to the prayer by saying “No.” Maybe he wanted the guy dead. Maybe, in his mysterious will, the guy was better off being relieved of this world’s temptations, and having entry at last into the eternal bliss of nirvana.
Superstition is retroactive-only. It gives us no useful understanding of how things work in the real world, it only rationalizes the things we’ve already seen in terms of what we want to believe. So pardon me for being a stickler, but, especially in the area of apologetics and our reasons for believing, I think it’s vitally important to be clear and accurate in discerning the difference between valid observations and conclusions on the one hand, and superstitions and fallacies on the other.
The facts of the hypothetical miracle are the same in all of the cases. All I’m doing is changing the details of the context surrounding the miracle, in order to illustrate how our perception of the circumstances can mislead us into drawing fallacious and invalid conclusions. It’s certainly understandable that people would jump to the wrong conclusion under certain circumstances. But jumping to conclusions is still a fallacy, and does not tell us anything reliable about actual cause and effect.