Systems of Magic, and a request

[rising_moon]

Recently I've read a few excellent fantasy novels which were written around believable, consistent, and reasonable systems of magic. Believable magic is one of the elements that will sell me on a writer. I've enjoyed The Abhorsen Trilogy, by Garth Nix, and, most recently, The Name of the Wind, by Patrick Rothfuss

Comments

[hissilliness]

I started out quite disliking Sanderson's essay. Eventually I had to admit that this is just because I'm an opposing partisan. His taxonomy, on consideration, strikes me as sound, but my tastes run in the other direction. Over-explicated magical systems (like WoT) end up reading to me like RPG sourcebooks with some characters stuck in for color. I'd much rather read something like Susanna Clark, which, over and over, gave me the shivery sense that is so much of what I come to fantasy for.

[rising_moon]

this is just because I'm an opposing partisan

And this is one of the reasons I love your commentary! :D

Your review of Last Call described the chief fallout from the Rules hangup: half a book's worth of Law, then half a book's worth of the writer shoehorning the plot into the Law.

Then there's China Mieville, whom you know I adore, who simply makes shit up as he goes along, and you realize that his rules are beside the point.

Susanna Clark, eh? Where should I start?

[hissilliness]

Clark has two books out. Jonathan Strange & Mr. Norrell is her novel, and The Ladies of Grace Adieu is short stories set in the same world.

[rising_moon]

Oh, duh. Thank you. I'm sorry to say that I started Jonathan Strange and then The Ladies, feeling like I should enjoy them, since all my friends did, but I didn't quite gel with either one. I never really liked Mr. Norrell enough to get over his being painted as unlikeable... or something. Maybe you can convince me to pick up Strange again.

[dilletante]

huh, in general i find tim powers does a good job of preserving a sense of wonder with his rules-- i can't always predict what's going to work for his characters but usually it feels like it fits with the rest of the magic in retrospect.

[rising_moon]

Now, I really enjoyed Last Call, particularly for stunts like the lead's imaginative approach to attending the title event. :) The hero/maiden/mother myth scene at the lake, though, kind of felt shoehorned in to fill out the rest of the archetype.

I'll return it with cookies. :)

[dilletante]

i went into the essay with your comments in mind, but had trouble figuring out what viewpoint he presents that one could be partisan about. :) but i guess that's because i just read the essay as being about the taxonomy and especially about his rule of thumb (you can only use magic to solve problems to the extent that the readers understand how magic works). do you disagree with his rule of thumb? (i think it's an excellent one).

i find myself thinking about it in terms of mystery novels: you can totally end your mystery with "oh, it turned out the butler had an evil twin, ha ha" but some readers will feel cheated. similarly, coming up with a new magic rule that solves everything strikes me the same way.

[hissilliness]

Samuel Delany suggests in a couple places that whenever we set up ostensably equal dichotomies, there's always an implicit hierarchy.

The only people I ever hear talk about Hard vs Soft SF are people who love the former, and usually show a bit of contempt for the latter. Sanderson goes out of his way to assert that he believes both the approaches he describes to be equally valid, but, to be blunt, I don't believe him.

I do think his proposed rule is worth musing on, I think it is nowhere near as hard-and-fast as he asserts. Exhibit A here would once again be Jonathan Strange, which is way to the Soft end if I understand the essay at all, and does not fall neatly into the problems-sloved-with-magic/problems-solved-not-with-magic dichotomy.

More generally, yeah, if you're setting up a puzzle game like a mystery novel, explicit rules are necessary. Not all fantasy novels are, or should be, puzzle games.

My New Theory

[hissilliness]

This has just popped into my head, I need to mull it a bit more before I'm convinced I believe it:

the how-not-to-do-it examples in your and Sanderson's descriptions have two salient qualities: arbitrariness and effortlessness. The former, I think, is a matter of tone more than anything else. Magic needs to feel consistent, it does not need to necessarily be consistent.

The deus-ex-machina climaxes that Sanderson warns against are really more about that effortlessness. An example: in WoT, which, as I said, is generally too Hard for my tastes, there's a wicked important climactic bit where the main protaganist attempts a massive alteration of the very nature of magic itself. It is a very Soft moment--there had been no prior indication that human effort could produce any such effect.

However, Jordan emphasizes what a huge amount of work, skill, suffering, and risk it requires, and that's what makes it not feel like a cheat.

Conclusion: magic has to cost somehow. I suspect that this is why so much of the magic in Harry Potter felt