The only claims magicians make that science could falsify are those that are about the material world, and those that imply a change in that world that an observer can see from the outside. This makes it all rather sticky, because if you take the idea of magic as a change in possible worlds in configuration space, then there’s nowhere to stand to observe any change that’s made. The causal chain will always point to coincidence, no matter what, because the movement of the index across configuration space is narratively coherent.
But we do have some tools that could hint that magic might work.
- If I claim that tarot cards predict the future, and I draw a tarot card to predict my day every day for a year, we can hypothesize that certain cards will more likely come up more often, since my life has certain themes that recur in any given short time. For example, I do tend to get the Hierophant rather a lot, because I’m a teacher. If we observed that statistical clustering over a long period of time, then we could say that tarot cards are not random. And yes, this has been done as I explain in my book on the Lenormand, and not by me. Jane English wrote an article published in Wheel of the Tarot where she describes such an experiment and its very impressive results.
- You could conduct an experiment to shift reality in very small ways and analyze those shifts statistically. For example, you could make up a matrix like this: “an X wearing Y will do Z.” Then you make a list: “X” could be “man, woman, boy, girl.” “Y” could be “black pants, blue pants, red pants,” and so on. “Z” could be “pick their nose in front of me, eat a slice of pizza in public, dance for no reason in public,” and so on. Now you could set up a program to select two instantiations at random. One of them it shows you, and one of it is hidden from you. The one it shows you, you do magic for. The hidden one, of course, you ignore. You do this for years, one each couple of weeks to give them time to manifest. If magic works, you’ll have more experiences like those you enchanted for, and not as many of the experiences described by the hidden permutations. I haven’t set this up in this double-blind way, but I’ve done some informal experimentation along these lines and found it impressively significant. The only problem is that it does kind of give me pause to deform my life in this weird way, because now women won’t stop singing in public and guys in black shorts keep striking up conversations with me.
- You could set up a series of goals over time, of varying difficulty (i.e., distance from the index in configuration space, which would show up looking like probability from a fixed indexical perspective). This is pretty much what every magician just does, and then you look back over your magical journal and count up how often you succeeded. You could even quantify it, and maybe someday I’ll write something about how to do that, but not here because it’s even more boring than usual. The point is, you can then quantify the success rate. If you use some mildly sophisticated techniques to estimate probability before doing the magic, you can compare the success rate to the probability and determine if the delta between the two is significant. I’ve done this informally, on the back of napkins and the like, and for me it’s quite significant. But this is hedged all around with dangers, such as confirmation bias and sharpshooter fallacy and so on. So my napkin-stats are scientific worthless, even if I find them personally convincing.
Every method of scientifically studying magic runs into the same problem, though: it just won’t convince a scientist, because it won’t end up in a peer reviewed journal. Jane English’s excellent article ended up in an obscure book of essays on the tarot (a very good obscure book — I highly recommend it). Many of the peer reviewed articles debunking psychic phenomena begin from the assumption that it must work according to the same material laws as physics, not that it might be a completely different magisterium. Having decided to look for something that they know isn’t there, they crow that they didn’t find it. It reminds me of the guy who lost his keys and was looking under a street lamp. He didn’t lose them there, but the light was so much better . . .