At the 2006 World Fantasy Convention, there was one panel titled “The God or the Machine?,” which addressed the division (or non-division) of magic or science. It was, hands-down, the best panel I went to that weekend, because it got me thinking, and left me with useful thoughts. I like entertaining panels as much as the next person, but this kind’s even better.
Let me start with the things that I don’t think usefully distinguish magic and science from one another. (Top of the list is Frazier’s approach, where you’ve got magic when you’re a primitive society, religion when you get a little more advanced, and science when you reach the top. But enough about nineteenth-century armchair anthropology.)
I don’t think it’s useful to say that science works within the laws of nature, while magic violates them. Whose laws? What nature? This view takes modern, rationalist Western science as the default, which is problematic not just on our own planet (where there are plenty of people with other opinions) but in invented worlds, where the laws of nature may be whatever the author pleases. “The supernatural” isn’t a word I particularly like; if it exists, how is it not a part of nature, in the non-environmentalist sense? If it doesn’t exist, then doesn’t “the supernatural” really mean “the fake”? Bleh. Sure, magic may violate the laws of scientific nature, but you could just as easily say science violates the laws of magical nature. A dead-end, to my way of thinking.
Then there’s the idea that magic operates by/is a manifestation of will. While there’s some truth to this, I can poke two holes in it. First, a lot of magic systems require more than just will; even David Eddings’ Belgariad is based on the Will and the Word. Usually you need to do something. Second, isn’t there an element of will involved in science, too? “It is by will alone I set my mind in motion,” Piter de Vries says in Dune, but Mentats are human computers more than magicians. I’m also reminded of Apollo 13, when Jim Lovell says, “From now on, we live in a world where man has walked on the moon. And it’s not a miracle; we just decided to go.” Sure, they had to do more than make the decision; they had to build things and develop technologies and work out mathematical equations. But so do magicians, much of the time. It may not take as many people, as much time and money and experimentation as the space program did, but both of them are based on an element of deciding you want to do something, and then doing what you have to in order to make it happen.
So what are you left with, at this point? Most of the time, we make the distinction based on trappings. If you chalk a circle on the floor, burn herbs, chant arcane mantras, et cetera, then you’re doing magic. If you take measurements and draw graphs and solve equations, then you’re doing science. Or we distinguish them by their effects: demon-summoning and fireball-throwing are magic, while genetic engineering and lasers are science. But I think we can agree that this is a pretty sloppy way to separate the two.
I never had a good answer to the question until Ted Chiang made a comment, during the WFC panel, that lit the proverbial light-bulb over my head. He was talking about alchemy, which is a classic case of fuzzy distinction between magic and science; it has elements of both, and sort of slipped from one to the other over the course of centuries. What he pointed out was the idea, once common in alchemy (but lost by the time alchemy turned into chemistry), that the process of alchemical transformation was also a transformation of the alchemist, that the creation of the Philosopher’s Stone was also a process of spiritual refinement.
I thought through the real-world magical systems I have any familiarity with, since I’m of the opinion that any division between magic and science ought to hold true in our reality, not just made-up ones. Generally speaking, they include some element — call it spiritual, call it moral, call it personal — that influences the act based on the actor. Who is performing the steps matters, not just based on their knowledge (i.e. whether they know what to do and do it correctly), but based on some more intangible quality. People are born with magical talent. People undergo spiritual training to acquire magical talent. People can only work magic if their hearts are pure (or foul). Et cetera.
Science, on the other hand, will work for anybody who knows what they’re doing and has the right equipment.
In simpler terms, it’s a matter of scientific principles, like objectivity and replicability. The results I get from an experiment should be the same as the results anyone else gets, provided they repeat my conditions. If you remove the personal element, making the procedure something anyone can do, then you have science, not magic. Even if it doesn’t obey the laws of science as we know them, it’s imaginary or invented science, not magic. Some parts of alchemy didn’t work in the slightest, but that didn’t stop them from being scientific in their approach. And you could write a very passable world where they do work.
There are still some fuzzy border cases; there always are. The western ceremonial tradition of magic (or magick) has been suggested to me as one such, as some of its practitioners promoted the idea that it was replicable by anyone with the right knowledge. It’s instructive to note, though, that this tradition got its start around time that science as we think of it was really starting to come together. Magic performed via contracts with other entities might be another borderline example, though in that case, I would stop to consider how it’s performed. If the contract is, for example, a simple matter of barter (I give this demon my soul, he does stuff for me), then the magic is not in the contract itself, but rather in the feats the demon performs, which are a consequence of its nature, and therefore not scientific. If, on the other hand, the person must control the demon by force of will or some similar quality, then we’re again looking at magic, this time residing in the person herself.
I tend to be utilitarian when it comes to theoretical constructs. For me, the test of an idea is whether or not it clarifies things for me that were muddy before. And in this case, it does. I’ve always had an odd relationship to China Mièville’s Bas-Lag novels; theoretically they’re fantasy, and he says things in various places about thaumaturgical energy and the like, but it never felt quite right to me. I like my fantasy, my magic, to have a numinous quality — but lacking a way to clarify what I meant about “the numinous,” it was hard for me to say how and why I found it absent in that setting. Looking at it in this light, I can see exactly what I was missing. When parts of Armada are mining ore that they refine to produce that thaumaturgical energy, when the process can be automated and industrialized, divorced from the people involved, then you can call it by magical terms all you like, but it feels like science to me, not magic.
(Whether or not that means I think his novels are SF instead of fantasy is a complicated question, and one for another post.)
Other people might have equally useful ways of dividing the two; any system of this sort falls squarely into the “your mileage may vary” category. But personally, I find this one clearer, simpler, and more watertight than most.