On making fictional miracles seem plausible

post by Yair Halberstadt (yair-halberstadt) · 2021-05-31T11:32:59.114Z · LW · GW · 4 comments

Consider the following two plot overviews of a book.

a)

The dark lord attacks Goodia.

At first Goodia holds out, but eventually the dark forces break through and take over the land, torturing and enslaving it's inhabitants. 

The Goodian forces retreat to their mountain fortress where they are able to survive for years by growing mountain crops. 

Eventually a traitor burns the food supplies, and the defenders begin to starve. 

The Goodian king, now old and ailing, challenges the dark lord to a duel, the winner who will take over both kingdoms. The dark lord agrees.

Just as he is about to kill the Goodian king, a meteorite flies out the blue and strikes the dark lord, killing him instantly. Everyone lives happily ever after.

b)

A comet is flying through space.

The dark lord attacks Goodia.

The comet nears the sun.

At first Goodia holds out, but eventually the dark forces break through and take over the land, torturing and enslaving it's inhabitants.

Bits of the commit begin to break off, and stream behind it in a tail.

The Goodian forces retreat to their mountain fortress where they are able to survive for years by growing mountain crops.

We focus on one of these meteors, zooming through space.

Eventually a traitor burns the food supplies, and the defenders begin to starve.

This particular meteorite is not flying off into the endless void of space, but by a million to one chance is heading for a planet.

The Goodian king, now old and ailing, challenges the dark lord to a duel, the winner who will take over both kingdoms. The dark lord agrees.

As we get closer we see the planet is blue and green, with streaks of white.

Just as he is about to kill the Goodian king, a meteorite flies out the blue and strikes the dark lord, killing him instantly. Everyone lives happily ever after.

Now these are clearly both awful stories, but the first seems significantly more awful. I think I could almost get away with writing the second, but I wouldn't stand a chance with the first. The climax is a miracle happens and the dark lord is killed? What a waste of a build up!

What is curious is that the second story adds exactly 0 information. We know how meteorites occur, and we could have extrapolated everything I wrote about the comet simply by knowing that the dark lord was struck by one. So why does the second seem so much better a story than the first?

It's plausible that this is literature specific. We like there to be a build up to the resolution of a story, so that an astute reader can see it coming and guess what happens. The second story gives us that, but the first doesn't. If that's the case, then there's not much this can tell us about how people think in general.

The second is that this is an instance of a more general bias: if we can see the mechanisms of something, then we assume that it's much less unlikely. Everything has a cause so there's nothing left to explain, hence it must be plausible. I think HPMOR touches on this: 

"Ah, that reminds me, I have a question of my own." The Defense Professor was now giving Harry an intent look. "What gave me away at the last, in the corridor outside these chambers?"

Harry put aside other emotions to weigh up the cost and benefit of answering honestly, came to the conclusion that the Defense Professor was giving away far more information than he was getting (why?) and that it was best not to give the appearance of reticence. "The main thing," Harry said, "was that it was too improbable that everyone had arrived in Dumbledore's corridor at the same time. I tried running with the hypothesis that everyone who arrived had to be coordinated, including you."

"But I had said that I was following Snape," the Defense Professor said. "Was that not plausible?"

"It was, but..." Harry said. "Um. The laws governing what constitutes a good explanation don't talk about plausible excuses you hear afterward. They talk about the probabilities we assign in advance. That's why science makes people do advance predictions, instead of trusting explanations people come up with afterward. And I wouldn't have predicted in advance for you to follow Snape and show up like that. Even if I'd known in advance that you could put a trace on Snape's wand, I wouldn't have expected you to do it and follow him just then. Since your explanation didn't make me feel like I would have predicted the outcome in advance, it remained an improbability.

Hearing how something unlikely occurred seems like it explains how that unlikely thing happened, but it doesn't. Of course every unlikely thing that does end up occurring will have some mechanism which caused it to happen! A true explanation needs to show that your original priors for how likely it was to occur were incorrect, and the event was actually much likelier than you predicted.

I'm interested which of these two mechanisms you think is the more likely explanation for why the second story seems better (or even if it does seem better)? My guess is it's a combination of the two. The evidence for the first is that if I gave that description of the comet after the dark lord get's killed by the comet, the story would still seem pretty awful despite the meteorite being equally well explained. The evidence for the second is that if I were to replace the comet with a giant pink elephant flying through space, and the meteorite with one of the elephants toenails coming loose the story would still seem pretty pathetic, despite there being equally as much buildup, and the result being equally predictable.

4 comments

Comments sorted by top scores.