(If you missed the previous parts, you can start here. Go on, I’ll wait for you. ETA: This post actually stands on its own quite well. Feel free to read this one first then go back, I think you won’t be confused.)

This is a long one, but I think important. There are two large parts of making an argument: that of the logic and mechanics of the argument itself, and that of the rhetoric of making the argument. The logic of an argument is vast: as anyone who has known a small child will understand, one can ask “why?” indefinitely and continue to produce answers right down to the bedrock of “Because I said so.” Even an argument about what to have for lunch can, in principle, be elaborated upon in an exhaustive manner.

In general, that does not happen: in addition to the logic of an argument, there is the rhetoric of it. That is, there is a practice of deciding what to talk about and how. This is in many ways a study of one’s audience, which is something that a writer of fiction ought to be thinking about anyway.

Today I’m going to talk about a subject out of the heart of rhetoric: the level at which the reader agrees with the author about what’s settled or proven.

This is an old question. Quintilian‘s Institutes of Oratory (Book 3, Chapter 6) goes into the question of what he termed status, or standing. The concept is much older than that, though; Quintilian felt the need to point out that the term pre-existed Hermagoras of Temnos, who was mucking about in the first century BC. The Greek term stasis is usually used, and it basically represents the point of argument: where do the arguers stand in common?

Picture your hard-boiled detective, Hardwick Steele, pickled in strawberry martinis and waving his phaser. He bursts through the door and stares down his suspect. Unsteadily pointing his finger, he growls, “I figured it out! You’re the one who murdered Humperdinck Tragismus IV!”

The suspect, orthodontist Bettina Overbyte, gasps and responds…

Well, Dr. Overbyte has options. Let’s say that her response is an indignant, “I didn’t kill anybody, and nobody saw me!” What we have here is a stasis of conjecture: whether the thing (Overbyte killing Tragismus) exists or is true. Murder requires a killing, she did not kill anyone, therefore Steele’s statement is false. She denies the statement entirely, and the argument proceeds with Steele producing evidence to back up his claim, and she attacking his evidence (indeed, she’s pre-emptively attacking his argument by claiming that he cannot have eyewitness evidence).

Maybe somebody did see her, though, that little ginger kid with the beady eyes. So instead her response is, “It was an accident! I was just driving my car back and forth in the middle of the street (as one does). I had no idea he was under the wheels!” What we have here is a stasis of definition: She implicitly or explicitly cedes the conjecture (she killed him) but argues that said killing cannot be defined as murder. She says that it was an accident, and accidental deaths are not, by definition, murder, therefore her act of killing Tragismus was not murder, and so Steele’s statement is false.

Maybe that ginger kid saw her open the door, kick the corpse, and drive back and forth a few times. So instead she says, “He was on his way to burn down the opera house! Can you imagine? All those sopranos? Everyone in the city would have gone deaf! I just had to stop him.” Overbyte is ceding here the conjecture (she killed him) and the definition (that act amounted to murder) and the stasis is of quality: it was murder, sure, and so Steele’s statement is technically correct, fine, but he says it like it’s a bad thing. And it really isn’t… The argument proceeds according to whether it was justified (or whether it merits the punishment Steele implies).

Note that if she cedes conjecture (she killed him), definition (totally murdered, LOL) and quality (probably not very nice) then there is no argument. “It’s a fair cop,” she says as she’s led away. (Or, “You’ll never take me alive!” as she pulls a kukri knife from the credenza, whatever a credenza is) Depending on the kind of argument you’re having, there’s also a stasis of policy — what ought to be done about it –but I’m going to set that aside.

In general, when you reply to an argument, you want to make your strongest reply. Never cede something, implicitly or not, that you can successfully challenge. Why not always challenge the conjecture? Well, if your defense of it is only very weak, then you may be better of ceding that and mounting a strong defense at definition. It is often a bad idea in ordinary arguments to mount multiple arguments at different stases. If you slide down the bannister, as it were, without waiting for your opponent to start an argument as each stasis is resolved (“I didn’t kill him, and anyway it was an accident and even if it wasn’t, it’s a good thing he’s dead!”) then you look desperate and shifty and smelly.

There is one last stasis: if instead of all those, she replies, “What are you doing in the ladies’ room? Get out!” then we are instead at a stasis of procedure: she and he cannot have this argument in this venue, it is not the proper place. It is in theory a matter of principle. Not every argument could take place just anywhere. This is why we have courts, to prevent arguments in venues where those arguments could turn to fights, or where one arguer has some unfair advantage. Argument of procedure, in my opinion, is one of the foundations of civilization. Yet arguing procedure is risky: it can look like you’re grasping at straws or stalling. After all, if you could argue conjecture, you would at any opportunity, right?

So, that’s amusing and interesting and all, but how do we apply this to making an argument in fiction?

When you argue with the reader, you need to have some sense of how much your audience is willing to concede. And that means you need to have a sense of your audience, and of your market.

Writers of science fiction have these sorts of arguments with their audience all the time. Different readers of different technical bents will argue with their science: “People can’t travel faster than the speed of light, and even if they could, that hyperspace garbage wouldn’t work, and even if it did work, they wouldn’t use it like that.” (Dunno what argument of procedure would be. “Why are you submitting this story to ‘Swordchucks and Biceps Monthly’?” maybe.) That is, some of your potential readers just won’t accept anything other than science fact (as they see it), and you’ll lose them with this story. Many will accept faster-than-light travel, but will balk at being asked to accept a method for doing so that they think is ridiculous. And still more will even accept the ridiculous as long as you handwave (it has the quality of “unimportant”) and don’t rub their noses in it or use it in a dumb way. How can you tell where the sweet spot is? Be conservative in estimating where your stasis is. Know your market. Editors are smart and canny: Read the books or short stories they accept.

I think we can break down the stases a bit more explicitly than that. Unless you’re writing for a specific person, you’re playing the numbers. Some will be at different stases than others, and you’ll need to argue at multiple levels, but it’s still a story and there’s a limit to how much argumentation it will hold. I’d say that there’s a rule of thumb. When you get to the point where the bulk of your audience will balk: that’s where you start explaining yourself. When you get to the point where the bulk of the audience will think you’ve made your point so get on with it: that’s where you stop explaining yourself.

Stasis of conjecture is where the readers and the writer cannot agree on the facts of the world. Writers of science fiction, fantasy, and horror stumble on stasis of conjecture most often when their books are picked up by readers who are alien to the genres. “Magic doesn’t exist!” “You can’t travel faster than light!” “Why is Mr. Bennett talking about zombies? What’s that a metaphor for?” There’s a whole article to be written on this subject — and we are fortunate that Jo Walton has in fact written it. I’m not sure what to do about this except to choose your market well, and be careful with cross-genre pieces. Sometimes the story can be accepted anyway as an extended metaphor: that is, the reader won’t agree about the facts of the world, so you take a step back and agree: “You’re right, this isn’t real. When I talk about zombies, mentally substitute ‘Justin Bieber fans’.” I think that this is, to some extent, what those authors are doing, who insist that their dystopian novels aren’t science fiction.

There’s a similar issue with books like the Left Behind series. (No, no link. Just… no.) Stasis of conjecture comes strongly into play in the theology behind the series. People who already believe that particular theology will accept the fact of the Rapture with no problem. People who do not will balk at it, or take it as an extended discussion of how people with strong religious convictions ought not own pets. Because the market of the audience is intended to consist entirely of believers, I’m betting that little effort is made in persuading the reader to accept the facts.

When I said earlier that Ayn Rand lost me at the end of Atlas Shrugged when the world collapsed on cue, it was because I disagreed with her on the facts: “That wouldn’t actually happen,” I insisted. I would not concede the conjecture. But that’s not to say that she blithely assumed that I would agree: her market was wider than just people who already agreed with her (though it did assume some sympathy, I think). She anticipated this disagreement at this stasis, and argued at length that this would happen, and why. She hinted at first, then her characters said straight-out that it would happen, and why it would happen. If I had accepted everything the characters said, then I would have conceded that that’s what would have happened.

What if I accept the facts of the world, but disagree about what they mean: not concede the definition? In Agatha Christie’s The Murder of Roger Ackroyd, Hercule Poirot investigates the murder of a local man. He does his thing with the little grey cells, and causes the murderer to confess. The crime is described in great detail, and is built up very much like an argument: the claim is that this person killed Ackroyd, here the evidence, this is why that evidence proves that claim, &c. Pierre Bayard, a French professor of literature, wrote a book entitled Who Killed Roger Ackroyd? (I have a copy laying around somewhere, but can’t find it. It’s an excellent read, though.) In it, he argues that there is an alternative — and more convincing — interpretation of all the facts at hand, including the confession. Bayard also wrote a book with the modest title, Sherlock Holmes was Wrong: Re-opening the Case of the Hound of the Baskervilles.

Larry Niven, after the publication of Ringworld, had a number of fans from MIT who went to a conference with him and chanted “The Ringworld is unstable! The Ringworld is unstable!” The age of the Ringworld combined with lack of certain details implied stability, and they had proved that it was not. They didn’t disagree with the fact of its existence, but they said it wasn’t what he said it was. In response, Niven conceded the point and got a sequel out of it, The Ringworld Engineers. (I’m pretty sure that this is stasis of definition rather than quality. It feels more like “Yeah, I killed him, but it wasn’t murder” than “Yeah I murdered him, and it’s a good thing, too.” But I think it’s open to interpretation. What do you think?)

Finally, there is difference over what the facts of the world mean. I always wanted the witch to win in Hansel and Gretel. All right, I’m a bad person, but look: two obnoxious children piss of their parents so much that they get left in the woods to die. They find a house. Do they offer to do chores for food or lodging? No! They just start in with the vandalism. And this nice old witch takes them in, gives them food and some discipline, and in return they deceive and then eat her. W.T.F. What we have here is an argument of quality: I accept what happened, I accept what it’s called, but I’ll be damned if I consider it a happy ending. Pfui!

Quite often the author has a very good idea of where the reader will balk, and describes the world with that in mind. As I said, Ayn Rand anticipated where she thought the reader would refuse to concede the point, and argued there. She also knew that some readers would refuse to concede different points (and that once convinced at one level, argument could move to the next) and so the book had to argue at several different stases: Once everyone’s convinced that the events of the book made sense, they can get onto the question of whether John Galt was a savior of humanity, or just a genocidal prick? This may happen, by the way, over multiple readings or through the course of one reading. If your arguments require multiple readings to satisfy someone going through all three stases, you may lose your audience. (The subject of re-reading deserves its own post, but I doubt I’ll get to it. If you have any thoughts, please share!)

Whenever you come across a book with an unreliable narrator, the author is setting up an argument over conjecture — usually for the sake of tension, but often to make a philosophical point. In this case, the author knows that in order to establish the quality of something as good or bad, it first must be proven to be true and as-advertised. The unreliable narrator (who, as I’ll explain later, lacks ethos) must make the case that he is telling the truth.

Arguments of definition are extremely common in murder mysteries. The characters treat them as though they are arguments of fact, but by the time the murderer is named, the facts of the matter are not in question: the reader has seen them and knows them to be true, but merely has not made all the logical leaps. The argument between reader and author is whether these true things add up to this definition. The same can be said more broadly for any villain who is unmasked, or for a Chosen One. Even when the world takes sides, it often does so indirectly at the fact level, and relies on the reader deducing, “Well, the rays of light coming from this person’s head… and everyone falling to their knees at the sight of him… holy shit, he’s emitting ionizing radiation!”

This is probably a good place to stop. It’s going to be a while before my next post on the series — I’ll be at Boskone! — but this is a long enough one that it’ll take time to digest. There’s a lot more to be said on the subject of stasis, of course, and if you’re interested I have several good links. I originally got the Quintilian link from Paralepsis, who has a good treatment of the subject, geared toward political argument.  There’s also the Purdue OWL treatment of stasis theory.