From lierdumoa.tumblr.com:

To my friends on the [autism] spectrum, let me explain to you an unspoken social rule that possibly nobody has ever explained to you before

If a neurotypical asks you, “What game are you playing?” they’re not asking you to describe the game.

They’re asking you if they can play too.

If a neurotypical asks you, “What are you watching?” they’re not asking you to explain the plot of the movie/tv show to them.

They’re asking if they can watch it with you.

When neurotypicals ask you “What are you doing?”

What you think they’re asking: “Please explain to me what you are doing.”
What they’re actually asking: “Can I join you?”

Now here’s the really fucked up part. If you start explaining to them what you’re doing? They will interpret that as a rejection.

What you think you’re saying: [the answer to their question]
What they think you’re saying: This is an elite and exclusive activity for a level 5 friend and you are a level 1 acquaintance. You are not qualified to join me because you don’t know all this stuff. Go away.

This is why neurotypicals think you’re being cold and antisocial.

IT’S ALL A HORRIBLE MISCOMMUNICATION.

I don’t think this is always true – and when it is I would describe it as more of an open-ended attempt to start a fun conversation than a demand for participation – but I agree that it’s not just a straightforward request for information.

And there was some interesting discussion about this on Autistic Tumblr, which centered around: why would someone do this? Why can’t people just say what they mean?

And the best answer I saw – sorry, I can’t find it right now – explained that people were trying to spare their friends the burden of rejecting them. Say Alice is reading a book, and Bob asks “Hey, do you want to talk about that book?” Maybe Alice doesn’t want to talk about it. But the following conversation…

Bob: Hey, you want to talk to me about that book?
Alice: No

…sounds really rude. So by Bob saying his line, he’s putting a lot of subtle pressure on Alice to agree. Bob is a good person and he doesn’t want to do that. So instead he asks “Hey, what are you reading?”

Bob: Hey, what are you reading?
Alice: Not much. Just some random novel.
Bob: Oh, well, enjoy!

Or:

Bob: Hey, what are you reading?
Alice: Oh! It’s really interesting! It’s this book where Apollo 8 crashes into the celestial sphere surrounding the world, and suddenly everything reverts to kabbalistic Judaism…
Bob: Sounds neat! What happens next?

Here Alice either has an opportunity to signal that she wants to continue the conversation, or to reject Bob while maintaining plausible deniability that she’s doing that.

(The beauty of this theory is ruined only by the fact that half the time this happens in real life and I say “Just some random novel,” Bob actually answers “Oh! What kind of random novel?” and then I say “Oh, nothing really”, and Bob says “Come on! Something has to happen!” and then I start despairing that anything about social interaction can ever work at all. I don’t know. Maybe Bob is autistic.)

What I find interesting about “plausible deniability” explanations is that Bob has to operate as close to the border of “inscrutable confusingness” as possible without crossing it. He wants Alice to know he wants to talk to her, but he doesn’t want Alice to know that he knows she knows he wants to talk to her (I’m being very deliberate in putting the word “know” exactly three times there rather than just using a vague phrase like “common knowledge”). As long as Alice doesn’t know he knows she knows he wants to talk to her, Alice can give a non-answer, pretending that she believes Bob will believe that she just didn’t realize he wanted to talk to her.

And this sort of weird common-knowledge-denial-process only works if you’re skirting the border of incomprehensibility, hitting a sweet spot where you think the other person will understand, but it’s also just barely plausible that the other person wouldn’t understand. If you say something the other person would definitely understand, then the game is up. Given some sort of natural variation in how good people are at understanding cues, your best bet is to send a cue that will fail a small but non-zero percent of the time.

But if there are people who are unusually bad at understanding social cues, like autistic people, then any cue calibrated to be on the exact border of neurotypical understanding is likely to fail for them more often than not.

I don’t know how common this pattern is. Making requests seems like a pretty good example. Flirting seems to be centered upon this kind of thing. I’m not sure what else is involved, but I bet it’s a lot.

A while ago I quoted a paper by Lawson, Rees & Friston about predictive-processing-based hypotheses of autism. They said:

This provides a simple explanation for the pronounced social-communication difficulties in autism; given that other agents are arguably the most difficult things to predict. In the complex world of social interactions, the many-to-one mappings between causes and sensory input are dramatically increased and difficult to learn; especially if one cannot contextualize the prediction errors that drive that learning.

And I was really struck by the phrase “arguably the most difficult thing to predict”. Really? People are harder to predict than, I don’t know, the weather? Weird little flying bugs? Political trends? M. Night Shyamalan movies? And of all the things about people that should be hard to predict, ordinary conversations?

And I think part of the answer might be: ordinary conversations are hard to predict because they’re designed to be so. Conversation norms are anti-inductive. Like Douglas Adams’ conception of the universe, any time people start to understand them too well, they have to get replaced with something a little bit less comprehensible.