There is more talk about magic systems now than there has ever been. The phrase turns up in cover copy, in reviews, and in most arguments about the genre online, where readers speculate happily about limits and mechanics while the authors themselves stay rather tight-lipped.
Not all of them, though. Four of the great names have left a good deal on the record: laws, interviews, working notes, and in one case a chalkboard covered in equations. These are worth taking apart to see how they tick.
Start with J.K. Rowling’s plan for Harry Potter and the Order of the Phoenix. It survives as a hand-drawn grid, and the page that circulates covers chapters thirteen to twenty-four. It has a column for the chapter number, one for the month of the school year it falls in, one for the title, and one for the main plot of the chapter. Then it has six more columns, and they are these.
Prophecy. Cho/Ginny. D.A. O of P. Snape/Harry. Hagrid and Grawp.
Those are Rowling’s own abbreviations, and they name the book’s six running storylines: the prophecy about Harry, his romantic entanglement with Cho and Ginny, the secret student defence club called Dumbledore’s Army, the adult resistance group called the Order of the Phoenix, the feud between Harry and his teacher Snape, and the gamekeeper Hagrid’s attempt to hide his giant half-brother Grawp.
It reveals the inner workings of an author trying to stick the landing on a complicated, multi-threaded story. Yet at no point is magic mentioned. Surely, if this is how the story itself is planned out, the magic system must have received a similar treatment?
She had settled that part before she started
A grid with no magic in it invites an obvious conclusion, which is that Rowling never thought hard about how her magic worked, or perhaps subordinated it to the back of the creative process. This is not the case, and Rowling is on record saying exactly the opposite.
“The most important thing to decide when you’re creating a fantasy world is what the characters can’t do,” she said, in an interview published by the South West News Service (SWNS) on July 8, 2000.
Versions of that statement date all the way back to the beginning. To The Herald in 1998, she said, “I have to know in detail what magic can and cannot do in order to set the parameters and establish the stories’ internal logic.” And in a Barnes and Noble online chat the following year, “I spent a lot of time inventing the rules for the magical world so that I knew the limits of magic. Then I had to invent the different ways wizards could accomplish certain things.”
It’s the second statement that contains a key insight. The limits came first. The powers were worked out afterwards
Crucially, those limits cost the characters something, which is the test of any real limit. Food cannot be conjured from nothing, which is why three teenagers spend a substantial stretch of the final book hungry in a tent. Money cannot be conjured either. And the dead do not return, a rule she fixed in public in 1999 and then built the entire series on top of: “Magic cannot bring dead people back to life. Once you’re dead, you’re dead.”
So the grid is not the work of a writer who skipped the magic. It is the work of a writer who had finished with it. The rules were settled before the drafting began, and what still had to be held in the head, chapter by chapter, was who knew what. Only one of those two jobs needs a spreadsheet.
Which is probably why the misunderstanding persists. The spreadsheet is a photograph and it is two clicks away; the quotations are scattered across twenty-five years of interviews, a good number of them on a fan archive whose domain has since lapsed. The evidence that Rowling planned her magic is simply harder to come by than the evidence that she planned her plot.
That order of operations, limits first and powers afterwards, turns out to be the pattern rather than the exception. The writers who left the fullest accounts are the ones who make it clearest, and one of them wrote his down on purpose.
The only one who published the rules
Brandon Sanderson, who writes the Mistborn and Stormlight Archive novels and was handed The Wheel of Time to finish after Robert Jordan died, is the rare case of a working novelist who wrote his method down and put it where anyone can read it. His First Law runs: “An author’s ability to solve conflict with magic is DIRECTLY PROPORTIONAL to how well the reader understands said magic.”
Look at it closely, though, because it is not actually a rule about magic. Every term in it is about the reader. It sets no requirement that a system be consistent, detailed or physical. It says only that you may use magic to get your hero out of trouble exactly as far as you have already explained it, and not one inch further.
It says only that you may use magic to get your hero out of trouble exactly as far as you have already explained it, and not one inch further.
His worked example is Tolkien, and it is the best argument in the essay. Gandalf’s powers are never set out; we have no idea what he can and cannot do. That vagueness costs Tolkien nothing, because he almost never uses magic to resolve anything. Magic in Middle-earth creates problems rather than solving them. There is a reason Gandalf does not fly Frodo to Mount Doom, and it is not that eagles are unavailable.
The alternative is what Sanderson is actually warning against: the solution that arrives from nowhere, on the last page, by a means the reader was never told about. An author who has explained nothing and then rescues everybody with magic has not written a twist, but a cheat.
The second law is where the work is
If the First Law is about the reader, the Second is about the writer’s labour, and Sanderson reduces it to two words: Limitations > Powers.
He first printed the argument in the magazine Leading Edge in June 2011, and the formulation he uses is that what characters cannot do, or will not let themselves do, is generally more interesting than what they can. Rowling had said nearly the same thing twelve years earlier, which suggests this is less a discovery than the place every writer arrives at who takes the problem seriously. The sharp part is his warning against pretend limitations: a magic that requires some special implement the character is never actually without, at no real emotional or economic cost, has not been limited at all. It has been decorated. A real limitation ties the magic to the rest of the setting and makes life genuinely hard for the people using it.
The Third Law completes the set and is the least quoted: expand what you already have before you add something new.
He is careful about where the idea came from, too. He credits a book on writing by the novelist Orson Scott Card for the notion of setting down laws at all, and his model of a system done well is not a fantasy novel but Isaac Asimov’s robot stories: three laws, stated plainly at the outset, never violated and never added to, and several decades of fiction generated out of the friction between them.
The chalkboard, and what it is for
Patrick Rothfuss, whose The Name of the Wind and its sequel follow a gifted student through a university of magic, went further into the arithmetic than anyone else discussed here, and said so plainly: “Most of sympathy and sigaldry is pretty much thermodynamics.” Sympathy is the energy-transfer craft his students are taught; sigaldry is the rune-work taught alongside it.
He was not speaking loosely. “I literally have the math for a lot of these things,” he told the science-fiction magazine Reactor in 2016. “I have run the numbers about how much heat it takes for this and that, and accounting for slippage, or whatever. I can look at my chalkboard here and see all of the delta calculations for how much energy it takes to boil gold.”
A chalkboard of energy calculations for boiling gold, for a novel. What it buys is not realism, and Rothfuss says so himself in the same interview: “Once I explain that framework to you, if my characters are clever using the framework, then you can appreciate their cleverness at a different depth, and it’s very satisfying.”
That is the whole purpose. The sums are not there so the magic will be accurate. They are there so the reader can keep score. A character who escapes by a clever application of a rule you already understood is thrilling; the identical escape, by a rule introduced in the same paragraph, is nothing.
And Rothfuss proves he knows this is a choice rather than a virtue, because he built a second system and did the opposite with it. Naming, the deeper art in the same books, has no arithmetic behind it and is meant to produce wonder instead. One author, two systems, each tuned to a different effect. Hard and soft are settings on an instrument, not a ranking.
Frazer got there in 1890
Sympathy is not a coinage.
“Sympathetic magic” is a term coined by the Scottish anthropologist James Frazer in The Golden Bough, his vast 1890 comparative survey of myth and ritual, where it is the umbrella for two principles he found across a great many cultures. The Law of Similarity holds that like produces like, that an effect resembles its cause. The Law of Contact, or Contagion, holds that things which have once been in contact go on acting on each other after the contact ends.
Set that beside the books. A sympathetic binding needs a physical link, ideally a piece of the thing itself, which is Contagion. The binding works better the more closely the two objects resemble one another, and leaks energy as heat when they do not, which is Similarity with an efficiency coefficient bolted onto it. What looks like the most engineered magic in modern fantasy is Victorian anthropology with the losses calculated.
Le Guin started with the school
The oldest account of the four is also the plainest about method, and it belongs to Ursula Le Guin, whose wizard books, beginning with A Wizard of Earthsea in 1968, are among the few in the genre spoken of in the same breath as Tolkien’s. Writing in 1973, she set down the question she had started from:
“Wizards are usually elderly or ageless Gandalfs, quite rightly and archetypically. But what were they before they had white beards? How did they learn what is obviously an erudite and dangerous art? Are there colleges for young wizards?”
Her first move was not a power, a cost or a rule. It was an institution. She asked how the art is taught, and the book followed from the answer, which is why the wizards’ school on the island of Roke exists twenty-nine years before Hogwarts and thirty-nine before Rothfuss’s University.
Then consider what she actually built once she was inside it. In Earthsea, power over a thing is knowledge of its true name. Not a force you channel or a resource you spend: a fact you either have or do not have. The mechanic and the information are the same object. Every writer here is managing what the reader knows; Le Guin simply built a magic in which knowing is the power, and she got there first.
What the documents have in common
Four writers, four methods, and not one surviving working document that is a list of powers.
Sanderson’s laws are about how much the reader has been told and about what the magic is forbidden to do. Rothfuss’s mathematics exists so that cleverness can be scored. Rowling settled what magic could not do before she began, and the grid she then ruled up follows six threads of who knows what, with no line for magic in it anywhere. Le Guin’s system makes knowing the mechanism the mechanism.
The common failure looks like the opposite. The instinct, when inventing a magic, is to write the powers: what it does, how impressive it is, the names of the schools and the colours of the lights. That list is the easy half and it is close to worthless, because a reader cannot be impressed by capability alone. Capability with no stated limit is just the author promising to win.
What a magic system is, when it works, is a promise about what the book will not do to you. Nothing will arrive in the last chapter to save the hero that you were not shown early enough to have expected it. Every rule written down is a door the author has agreed to lock behind them.
Which is why the best notice this desk has taken of a magic system in the past year had nothing to do with its complexity. Reviewing a first novel in August, I found the thing worth praising was that its magic had a budget and could refuse. That is not a small compliment paid to a small book. On the evidence of the people who have actually explained their working, it is the entire craft.