I’m still working on the “how to design a fantasy religion” thing, but got a bit bogged down. Just this evening, however, I was re-reading John Kim’s Breaking Out of Scientific Magic Systems, and it sparked a couple of ideas.
This is all unrefined musings, of course....
John proposes five central principles that are worth considering, in the sense that he thinks (and I agree) that they are worth overturning. They are:
1. Magic is a known system and thus non-mysteriousI’m not going to summarize in much detail, but to remind you:
2. Magic is a force separate from Nature
3. Magic happens as spells from deliberate users
4. Magic obeys conservation of (magical) energy
5. Magic works regardless of morality, ethics, or other intangibles
#1 says that magic obeys rules, known to the players at least and quite possibly to the characters. These rules are generally mechanistic.
#2 is best summarized by John’s reference to “detect magic” spells: if magic isn’t operative, things work “normally.”
#3 says that people who do magic do it on purpose and know that they are doing it.
#4 says that magic is a kind of energy that works on a pool or battery principle.
#5 is pretty obvious from the title.
What struck me on re-reading was the constant reference to “myth” and to ethnographic or anthropological sources on magic. I want to note that these aren’t quite the same thing: what we usually think of as myth (Greco-Roman, Norse, etc.) does not arise directly from the kinds of cultures traditional anthropologists studied (Polynesians, Australian aborigines, African tribes, etc.). Nevertheless the point is well taken: in none of these do we find something so mechanistic as FRPGs usually propose as normative to magic.
First of all, let’s think about these principles in terms of the history of scholarship on magic.
#1, the “magical laws,” really comes most directly from Sir James Frazer’s The Golden Bough, although he had predecessors (especially Edward Tylor). Frazer said that magic had two basic laws: the Law of Similarity (similarity can entail causal relations) and the Law of Contact (once in contact, always in contact). For example, if I make a doll that looks like John Kim, I can stab it and he’ll feel pain (Similarity); if I make a doll and include his hair, the same thing will be true (Contact). And these are commonly mixed.
Interestingly, Frazer was not so simplistic about this as are RPGs. He said that such laws were absolutely not known to the practitioners; they were implicit but not explicit. As I put it to my students, simplifying a fair bit, if these laws were absolute and known, they would be obviously stupid. If I have a big piece of chalk, and I break it into two halves, the pieces are quite obviously both similar and previously in contact. Now if I do something to one piece, and nothing happens to the other piece, it’s obvious that these laws don’t work as stated. Frazer’s point, really, is that this sort of thinking—what I’ve just done—depends on my having abstracted a mechanistic and scientific-style law. Once that happens, it’s obvious that the law can and should be tested, and when that leads to failure I discard the law. The “savages” don’t do this, not because they’re stupid (Frazer doesn’t think they’re stupid) but because they do not think in higher-order abstractions, and thus they do not abstract mechanistic laws from the vast series of magical practices.
In a game situation, we can say that such a structure as proposed by type #1 could be read one of two ways: either it requires a radical disparity between the way the players think and the way their characters think, or it requires that magic in RPGs be utterly unlike magic in the real world of culture, because (as John says) it makes magic into a series of scientific-style principles.
#2, the separation from nature, has a fascinating analogue in the work of Bronislaw Malinowski. Malinowski proposed that magic is exactly what John describes: it is an extension of practical desires beyond the ability to satisfy them practically; its function is not actually to achieve the goal as such, but rather to produce psychological catharsis with respect to the desire in question. In other words, the natives have practical knowledge (science) that they employ quite rationally and effectively to achieve what they want: if they want to go sailing on the deep sea, they know how to build quite sophisticated boats for this, and they know how to navigate by the stars, and so on. But this practical knowledge has limits that do not encompass the totality of the risks and concerns in this action: a bad storm may overwhelm the boat however well-sailed. When the risks and stakes are sufficient, the natives perform magical actions symbolic of the results desired, and this gives them confidence and allows them to attack their practical work with greater surety.
I always use the example of the elevator. You’re late for an important appointment (an interview or whatever), and there are too many stairs, and you just simply have to get there, but there is nothing at all you can do. The elevator is presumably on its way, because you have pushed the button (the rational, scientific thing to do), but it’s not coming fast enough. So you push it a lot more times. This makes you feel better, because you’re doing something, even if that something isn’t practically speaking going to achieve the desired end. You could, of course, hop up and down on one leg and make chicken noises, but since this has no symbolic relation to the desired end you aren’t likely to do this; instead you do something that is (a) symbolically related to the desired end, and (b) socially approved—nobody will give you a hard time if you push the button a lot, and in fact people will correctly interpret your act as meaning that you are in a hurry.
The thing is, Malinowski insisted (though he didn’t do so as strongly as he might have) that the natives do not make the same distinction. It’s not that the natives do the practical stuff and then step back and say, “Hmm, that’s not enough, better add some magic.” For them, it’s all of a piece: all these things are normal and necessary to achieving the desired end. The anthropologist, standing outside the culture, can make a functional distinction: these acts have different results as empirically observable. But the natives do not distinguish in this way; rather, they distinguish on the basis of their own meaning-systems, which are extremely helpful but not determinate with respect to understanding the functions of these actions.
So here we get the same result as with point #1. To impose this rule requires that either (a) the characters think quite differently than the players do, or (b) the characters mean something by “magic” that is radically unlike what we modern people can actually observe in cultures that use magic.
#3, the deliberate spell-casting, is rather more complex. Here I think John is over-generalizing somewhat. A great many cultures do have such a thing as a professional magician, and these people really do perform magic quite deliberately as a matter of ritual. On the other hand, as he says, mythological characters very often do not act this way: they just sort of do things, or it turns out they have certain powers for no adequately explained reason (or because someone’s grandmother the sparrow taught it, or whatever).
What we’re touching on here is something deeper, and something I think underlies a great deal of RPG thinking about magic. What do we have in mind when we say “magic”? It’s all very well to point to Tolkien, but Tolkien is in turn pointing to a limited corpus of mythological material, and it’s not clear that what he has in mind is simulation (in the broad sense) anyway. If we say that we want “myth,” which myths? If we have in mind cultures that actually use magic—and these are rarely represented directly in actual myths, because that would make the myths oddly reflexive—then we’re talking about something quite different.
To take a relatively extreme example, consider the disparity between what Australian aboriginal peoples do (or did) and what they say their ancestors living in the dream-time did. The latter provide a precedent for the former, but the disjuncture between “then” and “now” is terribly important; indeed, Durkheim and everyone coming thereafter (Eliade, Lévi-Strauss, even Campbell in his vague way) generally recognizes that this disjuncture is quite often foundational for the whole local conception of the sacred, with all that this entails—including what magic is. To blur the distinction, as most RPGs do, means that you can’t have magic in the way it occurs in tribal societies, because you have annulled its theoretical foundation.
This takes us to #4, conservation, which John rightly relates to the Polynesian “mana.” I should note that the Polynesian usage is still hotly contested among specialists. But RPG uses of this term aren’t getting it from reading up on Polynesia; they’re getting it from Marcel Mauss and Émile Durkheim, who postulated mana as a general principle.
The idea is that mana is an abstract principle of power that extends beyond what we might loosely call normality (which is not the same as nature). By this logic, mana is discerned by its use—not the other way around. If someone does something powerful or extraordinary, it must be that he has mana, and this will affect how he is treated in the future. With this reputation, he may attempt to capitalize in a broader sense; this is not unlike the Arabic conception of baraka. In North America, the terms wakan, orenda, and the like are thought by some (including Mauss and Durkheim) to be conceptually cognate. And because of this fact that a lot of cultures seem to have a notion like this, Mauss and Durkheim postulated that such a notion is in fact the underlying minimal notion without which religion and magic could not exist.
There are any number of destructive criticisms of this argument, many of them powerful on one grounds or another. Perhaps the strongest point is that if you read the data carefully, you find that all these notions are not at all cognate, nor does any one of them mean what is stated here. So the idea that this is a minimal prerequisite fails because nobody actually seems to have that minimum. But I’d like to note that my pal Lévi-Strauss proposed, in 1950, an interpretation of these conceptions that I think is theoretically speaking accurate; it also gets at why these ideas are really not workable in RPGs in the way they usually appear.
In short, Lévi-Strauss says that when language first appeared, i.e. when people first came up with the vague, unconscious notion of using arbitrary symbols to communicate with one another about things in the world, this was earth-shattering. In essence it meant that the whole world was quite suddenly filled with potential meaning, because suddenly you could think about things as symbols. But just because things could be symbolized and thus understood in human terms did not mean that people already knew how to understand all these things; even now we don’t know how to do this with everything, in any culture. So there was a kind of gap: you had things that were filled with meaning, but you didn’t know what they meant. And Lévi-Strauss says that things like that had mana (wakan, orenda, etc.). In other words, these terms are ways of saying, “Yes, it’s a big deal, and it’s powerful and important, and we have no way of grappling with its qualities or nature, nor of classifying it; at the same time we refuse simply to bow in awe and frustration before this difficulty, so we’ll give it a temporary holding-term and get back to it later.”
As it happens, I think he’s wrong, but he’s on to something. What he has explained fairly well is why terms like mana and wakan and orenda have this strange not-quite-meaningful quality in their native usages, and perhaps more than anything his theory is supported by the ascription of these terms to white explorers with guns and so on. What he hasn’t explained, however, is why such terms should also be associated with extra-normal power—even, quite often, extra-natural power. The fact that we don’t understand them doesn’t entail this, unless (going back to point #2 in John’s list) we say that magic (=mana) refers to what is exterior to nature.
What he’s getting at, and at this level he’s right in my estimation, is that these terms refer to powers exterior to classification, and thus they are both powerful and threatening because they expose the instability and weakness of classification itself, by showing its limits. In short, magic (if we interpret it this way) is powerful because it shows that science (read here as knowledge in a classificatory sense) is limited and narrow, which means that we really don’t know what we think we know because magic could potentially undermine things. Thus we label such powers and notions “magic” as a way to avoid dealing with them, and thus we protect the system of knowledge upon which we stand.
Now this interpretation works great for lots of things—but it works very badly for mythical acts often called “magical” and it works stunningly badly in the context of most RPGs. But the latter, I’d say, is no bad thing: we’ve already seen that RPG magic systems are usually founded on scientific and even scientistic assumptions, as John rightly notes throughout the article (that’s the point, right?), and now we see that the whole construction of magic in RPGs fits a general pattern within scientific and other rigidly classifying systems.
As to #5, the relation of magic to ethics and so on, I have little to say except that John is dead right. Magic, pretty much however defined, is never independent of such variables. Mauss rightly noted that magic often marks outsiders and marginals, precisely because they are such: they are in a problematic social, structural, moral, ethical, legal state, and therefore they have magical power—and vice-versa. And I think on the whole everyone has accepted that insofar as “magic” can be used as a substantive category in ethnographic or historical work, it certainly carries this complex structure most of the time.
- - - - - - - - - -
So what have we learned, Dorothy?
I don’t really know. From my point of view, the first question you have to decide when designing a “magic” system is what you want to model or, on the other hand, what you mean by “magic.”
I remember having an argument with a player who may be reading this entry, a few years back. I said that the system was eminently flexible, and asked what he wanted. He wanted to know what the structure of the system was, so he could figure out what he could do. I said he could have more or less whatever he wanted. He insisted that, without limits, nothing could be said. And I think we were both right.
From a design standpoint, you cannot really create an entirely mutable and open-ended system and expect it to work, because the players will never have any idea what can and cannot be done with it. In the real world, people know what can be done by magical means (in whatever sense that may or may not be taken locally or ascribed by outsiders) because they know the myths and they have heard the stories about that guy over there and besides they learned some stuff from their grandmothers. All of this discourse—and I mean that technically—provides a solid and quite rigid framework in which to work, even when the work pushes the boundaries. Some things are obviously stupid or unworkable; others are perfectly plausible, even obvious, and if they don’t work this time there must be some additional reason for it. But your players will never have this complete knowledge and sense of the obvious because they do not live the system.
If you want to model something, I suggest not modeling myth—with apologies to John. Myth seems to me always to have some sort of parallel discourse running, usually not available to us, and without this we really don’t know what’s going on. In a game, the only parallel discourse would be that of the players, creating a weird sort of reflexivity that I think is not at all what John means about myth. Myth generally stands behind, below, and so on with respect to a living, dynamic culture; myth itself is created at a distance: things don’t happen in myth because of any direct and ordinary sense of how people and animals and the world actually behave. Myth is about the tellers, at base, and in an RPG the only tellers would be the players—thus the reflexivity.
To my mind the thing to model is the sort of culture that uses and tells myth—and performs magic. And that’s most cultures, really, including ours (note the powerful revivals going on over the last 40 years or so, which have had continuous precedents of one sort or another running back as far as you care to mention).
A last note: nobody has yet succeeded in defining “magic” satisfactorily as a general substantive category. Consequently one form of argument or explanation is simply out of court: that which says that magic is “obviously” anything. It isn’t. There are some common agreements in one field or another, but these days most anthropologists won’t touch “magic” with the proverbial 10' pole. I disagree with this approach, but that’s an issue for another day. So if we’re going to develop magic systems for our games, they have to be driven entirely by their own internal concerns and constraints; there isn’t a right or wrong, and not just in the usual mealy-mouthed “whatever makes your game happy” sense.