Rigging a Thought Experiment
Tell the story you set out to tell; don't switch courses
Several months ago, I read Golden Enclaves, the third novel in Naomi Novik's Scholomance fantasy trilogy. As I wrote in my earlier blog post about it, I enjoyed the book, but I was sorry that Novik had simplified the ethical conflict in that book by making the antagonists turn out to be more clearly evil than they'd appeared in the previous two books. I kept chewing on that thought, and I realized that it's one example of a broader problem I've seen in too many books: if the world the author set up is a thought experiment, they rig the thought experiment.
Then I remembered I'd originally come across something close to that phrase on another blog. It's from a Scott Alexander essay on dystopian fiction, from his old blog which's only on the Internet Archive these days, quoting John Rawls:
Anyone familiar with the Straw Vulcan trope - the idea that anyone who's good at science or analytical thought must speak in a monotone all the time, condemn music and humor and love as "illogical", and suggest improving efficiency 28% by killing puppies since they have no productive function - will recognize dystopian literature as basically Straw Vulcanism as applied to cultures rather than individuals...
There's a much more active antipathy not just for logical people, but for logic itself; a feeling that anything which has been logically "optimized" is unclean, has necessarily lost whatever elements make it pure and good and human...
It was (appropriately enough) in a paper on John Rawls that I first read the phrase "rigging a thought experiment". And that's exactly what's going on here.
Alexander uses it to condemn a lot of modern dystopian fiction for setting up an interesting thought experiment; a culture which has made interesting tradeoffs that gain it some things but lose other things. But then, instead of exploring these tradeoffs, the author decides to get their readers to clearly oppose that culture. (After all, it's marketed as a dystopia!) So, the author decides to further destroy a separate, attractive aspect of life even though that's not necessarily part of the tradeoff.
For example, he complains how, in Burgess's A Clockwork Orange, the technology to condition criminals to be nonviolent - which could otherwise spark an interesting ethical debate - also keeps our protagonist from enjoying classical music. This makes it clear it's evil, and short-circuits the debate.
This's a good point in itself. But I'll take it further. Rigging your thought experiment isn't just a risk with dystopias.
Also several months ago, I read Hench by Natalie Zina Walschots. It's a story responding to (or perhaps parodying?) the superhero genre. Our protagonist, a henchwoman for supervillains, takes to cataloging all the damage superheroes cause to bystanders while fighting supervillains. This could be a challenging exploration of morality, but the "heroes" here are heroes in name only. We never see the superheroes do anything good aside from thwarting supervillains, and - after one scene at the beginning - we rarely see the supervillains do anything unambiguously bad. In theory, surely the villains too cause collateral damage... but we don't even hear it referenced in the book. What we do see is the villain giving our protagonist a good work environment and friends and rescuing her when she's kidnapped by the heroes.
In that universe, if what we see onscreen is representative, I agree with our protagonist that superheroes are a bane to society. In fact, I'd say they have the labels reversed. There are reasons the original superhero comics show our heroes doing things like rescuing cats from trees... but if anyone in this book were doing that, I'd say it would be the villains. This pushes out whatever other commentary Walschots is meaning to give. She's not exploring actual heroes' damage to bystanders, or the challenges of working with actual villains. She's exploring this specific situation where the labels are reversed. People (like me) who wanted to see the exploration of morality it sounded like we'd originally be getting are left disappointed.
I'm not merely saying that the settings here have flaws or are portrayed unrealistically. There're plenty of other flaws that can come up in portraying settings, such as the Harry Potter series which suffered growing pains from trying to build an epic fantasy series on top of the first volume's whimsical school story. The issue is when the writer breaks the rules he's put in his story explicitly or implicitly.
For example, if a writer's writing about a chess game between two people who care about chess and are good at it, he should have both players play as best as he can have them play. If he has one player start without a queen, or if he suddenly has one player make an illegal move and get away with it, he's essentially rigging the game. If he has one player start playing stupidly, he's done the same thing (like Hench where the supposed superheroes aren't representing their side of the question). If he suddenly has them start playing checkers instead of chess, that's rigging things on another level of magnitude (perhaps analogous to Golden Enclaves where we don't get an answer to the initial messy moral situation).
But if the author has them playing checkers from the start, or shows one player as bad at chess, that's not a problem (or at least, not this problem). For example, take Melissa Scott's Roads of Heaven science-fiction series. From the beginning, she shows one major medievalesque country as being unrealistically and offensively sexist beyond any actual medieval country. I personally consider that a flaw in the setting (one of several), but it's not rigging the thought experiment because she never pretended it was anything besides what it is.
Rigging a thought experiment doesn't just mean revealing your setting to be something other than what it looked like. Brandon Sanderson, in the Mistborn fantasy trilogy, loves pulling out the carpet from under his readers. Sometimes I personally didn't like that because I liked the carpet, but he's definitely dropped hints earlier that things are otherwise than they seem. A reader who didn't pick up on those hints could definitely accuse him of rigging the thought experiment by robbing us of the overthrowing-the-evil-overlord story things looked like at the beginning. But personally, I think he did give us enough hints that wasn't what we were getting.
Or, take The Giver. Scott Alexander mentions it in his blog post as a bad example, arguing that its message leads us to "wanting to keep society frozen" and rejecting utopian schemes to alleviate problems. The dystopian Community in The Giver does take its philosophy Sameness far enough to remove color and emotions, which aren't necessarily linked to its central goal of removing conflict. However, that's not shown to us as a bait-and-switch. The thought experiment isn't rigged, because it's never been pretended to be something other than what it is. We see from the beginning that the dystopian Community does these dystopian things, before we even hear any more of its philosophy. We could imagine another novel about a similar community that doesn't remove color and emotions, but that would be a different premise.
Still, you could argue that The Giver still rigs its thought experiment by showing us, late in the novel, that the Community also commits infanticide. But even if that's the case, as I said in my previous post on the book, I scarcely think that rigging is necessary - the book shows us more than enough dull banal Sameness to answer the experiment anyway. Perhaps the infanticide was added to wake up any young readers who've overlooked the rest.
Or, I can see a good argument that the infanticide is a natural outgrowth of the same philosophy the Community demonstrates in the rest of the story. In that case, it wouldn't be rigging the thought experiment because it's naturally part of the experiment. The chess game would be played out by the same characters according to the rules; it would be the same moral conflict and juxtaposition from beginning to end.
My examples so far have been from science fiction and fantasy books. In part, this's because they're the novels I read most - I'll hopefully write a post sometime about why. But also, I think this flaw really does show up there most often, because they're the books that do the most worldbuilding. If you're writing something set in the real world, then you aren't going to be able to invent superheroes that actually don't do any good, or reformation methods that actually also destroy your love of classical music. Either they don't exist, or they fit into the world in a non-random way that your readers know or could know.
But sometimes, stories set in the real world can rig thought experiments too. For example, take the first two of Chesterton's "Father Brown" detective stories. The premise of the series is that this Roman Catholic priest named Father Brown can often solve mysteries ahead of other detectives and the police, through his knowledge of human behavior and the human heart. In the first short story, we're introduced to Father Brown juxtaposed with the atheist policeman Valentin. Understandably, Valentin is frustrated when Father Brown solves the mystery before him, and the story sets up their two perspectives against each other. But, Chesterton abruptly twists this personal conflict to a quick ending in the second story, where Valentin himself commits a crime supposedly motivated by his atheism. Chesterton unrealistically short-circuited the conflict, rigging the thought-experiment he'd set up by juxtaposing Valentin and Father Brown.
Or, even in philosophy, I've read that Rawls himself - who came up with this phrase - has been accused of rigging one of his central philosophical thought experiments by implicitly assuming that the people in his thought-experiments are maximally risk-averse, and many actual people would make a different choice. But then, rigging thought experiments isn't a new thing in philosophy. I'm sure that dinner parties with Socrates didn't actually go as nicely as Plato wrote them in his Dialogues.
Why do authors do this?
Sometimes, I'm sure, they do it because they want to solve matters more quickly. I can definitely see a lot of points where Novik sped up the plot of Golden Enclaves so she could close the plot of her series in that book. I wouldn't be surprised if Novik also made the enclaves more blatantly horrible to enable that. I expect that Peter Jackson simplified the moral conflicts in the Lord of the Rings movies (versus how they were in the books) for the same reason, such as how he portrayed Denethor as wanting to surrender rather than literally killing himself before seeing Sauron's triumph. (Though, to his credit, Jackson never rigged his thought-experiments specifically.)
This's a legitimate reason. But, it still impoverishes the story. If an author makes an antagonist act against his character to speed up resolving the plot, that's a flaw in the story. Similarly, cheating in a thought experiment is also a flaw in the story.
Or, I expect, authors sometimes do this because they want the moral conflict to actually appear to be solved. I think this's related to what Alister MacQuarrie pointed out in 2019 about how "The heroic rebel on screen is often very evasive about the principles behind their actions... What beliefs the non-ideological hero does have are often vaguely defined." Any actual beliefs the hero has - any opinions on, say, Burgess's initial tradeoff between reforming criminals and personal autonomy - would make the readers uncomfortably "confront serious and difficult questions... we do not want to ask."
If the protagonist is forced to seriously wrestle with the central dilemma, the central tradeoff over personal autonomy or bystander harm from superheroes or risk-aversion in philosophy - well, that's a difficult topic. Readers might not want to think about it. The author might not want to think of it either, or might be concerned about readers' reaction to a book that squarely faces it. If the protagonist is forced to actually make a decision on it in the narrative, without being free to use something else like classical music or the superheroes not actually being heroic as a distraction - then the protagonist has taken a position on a debatable topic. Readers might disagree with the protagonist. They might decide that the antagonist (whether Novik's enclaves, or Chesterton's Valentin) has a more valid moral position. Authors generally don't want the readers to end up disagreeing with the author's position there.
So instead, the author makes the antagonist unrelatedly horrible to close that possibility off. But we're robbed of the interesting questions and moral exploration we could've had. The author's taken a shortcut past the good story we thought we were getting. And, we feel we're also robbed of the interesting characters we thought we were getting.
One way out of this problem is to not pretend to settle the moral issue. Even if your protagonist makes a moral judgment on a difficult question, you don't need to portray him or her as clearly correct. If they continue wrestling with the question and visibly caring about their uncertainty, that could even help win readers' sympathy.
For example, take Dorothy Sayers' Lord Peter Wimsey, the protagonist of her detective series. The point that won my sympathy for him in the second novel, Clouds of Witness, is his visible emotional conflict at having to dig into his relatives' secrets while investigating a murder in their family circle, and then his arguing with himself about whether to reveal his brother's obviously-relevant secret when he finally finds it. He comes to a decision, but he keeps agonizing over whether it's the right decision. Even in the last book in the series (Busman's Honeymoon), we see him go through an emotional breakdown at the ghastliness of his digging into so many people's private affairs.
I have a definite opinion on whether Wimsey is making the right choice - I'm sure most of Sayers' readers did and do - but it wins our sympathy that he doesn't let himself rest in clear certainty. It shows he's noticing; it shows he cares. Showing this is a strength of Sayers'. I've read she was the first detective writer to give her detective a multidimensional personality and relevant emotions; she's definitely done it more than her predecessors. More writers today, even in other genres, could learn from it.
I've got a post coming up soon about sequels, and one of the things I'll be talking about there is how readers can be disappointed at the sequel's not matching up to the first story - whether because it's disappointing in quality, or because it's good but it's a different sort of story. But even more important than that is for the ending of a story to match up to the beginning and middle. The chess game between good players should end as a chess game between good players, not with one bad player, nor as a checkers game. The ethical study of superheroes' collateral damage should end as a story about superheroes' collateral damage, not as a story about damage from some people wrongly called by the name "superheroes."
Otherwise, like Scott in the quote at the beginning, the reader will be disappointed. The protagonist of Golden Enclaves was disillusioned when she learned how the enclaves were being blatantly immoral, and that was appropriate inside the story. But what wasn't good was for me, the reader, to also be disappointed. The story should honestly tell the story it sets out to tell. The thought experiment should be an honest, un-rigged experiment.
Overegging the pudding is perhaps the most common flaw of fiction which tries to examine ethical issues.
For example, the past week I saw _Avatar 2_; the big moral theme of it is a 'save the whales' theme, about how evil the humans are for hunting space-whales, even if the humans get something important out of it (instead of lamp oil and perfume ingredients, _Avatar_ says they extract a literal cure for human aging). OK, fair enough, certainly this is an issue which is still live, as whale hunting is still done by a few countries and tribal groups, and more broadly about the ethics of meat-eating, it is worth examining... except then _Avatar_ goes on to describe how the space-whales in question are not just intelligent animals which are a major part of the ecosystem, but in fact, *superhumanly intelligent*, possessed of vastly greater emotional sensitivity than mere humans, and spend their leisure time pursuing superhuman mathematics and art we can't understand. At which point Cameron has overegged the pudding: no one will endorse killing space-whales who have more moral personhood than we do, as intended, but by rigging the thought experiment so badly, it now has no more relevance to the intended analogy than the question 'is murdering humans bad?' is relevant to whether Alaskans should be allowed to hunt the occasional bowhead whale. Yes, it's bad to kill humans or space-whales, but they are so different from real whales that Cameron's multi-hundred-million-dollar thought experiment is of zero interest.
This also happens almost every single time consequentialism is critiqued in popular fiction: the villain is always consequentialist, arguing that reaching the ends is too important to quibble about the means... and then in the end, the heroic heroes demonstrate that the means were not necessary to reach the ends at all (and therefore could not have been justified by the ends), thereby not refuting, or even critiquing, consequentialism at all. (You might hope that they would salvage such terminally stupid plots by at least throwing in a rule-consequentialism moral - 'you shouldn't be hasty about using bad means because you may be mistaken about them being necessary' - but they never do, they are entirely un-self-aware of how they are trying to have their cake & eat it too.)
Your point is my argument against using my fiction to argue for my political views — it's too easy to use author control to rig the game. I don't know if you have read my _Salamander_, but the antagonist (not the villain, who shows up later) is an intelligent and, on the whole, admirable person acting for defensible motives. One of the themes that appeared in the course of writing it was the issue of the end justifying the means — because there is a sense in which it does, as the antagonist makes explicit early on in an exchange with the protagonist.