"She sent me a signal; just a subtle one, that someone not used to reading verbals would miss. Most people are terrible at reading verbals."

Verbal vs nonverbal communication

Until fairly recently, I had little ability to read nonverbal communication, which meant that I had little wiggle room on one of the principal axes of communication. But I still needed to communicate with people. So I looked for patterns in their behavior. Tried to remember explicit preferences they’d stated, notice when they responded positively or negatively to a thing, and infer the preferences revealed by the things they sought out or tried to avoid. I formed explicit mental models of how people behaved, which were generally modifications of my model of how people in general behave. This is the story of how I learned that this ability is not universal.

Occasionally I tried out the conventional advice to figure out what someone wants by putting yourself in their shoes, imagining what you’d want in their situation. This failed catastrophically. So I was stuck with explicitly modeling people.

While the feedback I get on this is wildly uneven, people have told me - and increasingly tell me - that I have good social skills in some domains, that I am able to be charming, even that I have high emotional intelligence. But my self-image is still that of someone who can’t do people things at all. And of course, I have the universal cognitive impairment known as the typical mind fallacy. So I assume that (since I’m “bad at social things,”) any skills anyone else that surprise me or seem beyond my capacities are simply layered on top of the ones I do have. After all, if I can do it, surely my more highly skilled friend can!

One of my friends has the unusual (in my experience) ability to act to directly lower my werewolf level, bypassing the need for much of the explicit trust-building. They read my body language and notice when I flinch or wince even slightly, and then ask what upset me. If I try to hide it with "Nah, it was nothing," sometimes they say, "I think there was something. What was it?" And then even though I squirm a bit in answering, it feels really good and like I'm being well cared for.

I had mentally rounded this off to them being better at reading social cues in general. So when we had this exchange in mid-July, I figured that by singling them out for a direct message, and proactively mentioning this problem, I was letting them know that I needed them. I figured that this was a clear request that, if they wanted to be around me and care for me as a friend, they initiate contact repeatedly and not take lack of responses as an unwillingness to follow up:

Me: Hey, FYI I'm going through what's probably a depressive period, so I won't be good at offering to hang out. Not sure how long this will last.

Friend: Hey, OK. Would a visit from me be helpful?

Me: Yes, though I'm OK tonight. It's not anything like an emergency. I'm generally happy to see you. Just warning you not to infer that I don't want to hang out, on the basis of me not offering.

Friend: I see. OK 🙂 I'd like to see you sometime 🙂 I was more checking to see of you preferred to be alone.

Me: Ah. Thank you for checking 🙂

Yeah, company's good, especially people I'm very comfortable with like you.

A few days ago I mentioned to this friend that I’d felt hurt during August that they hadn’t reached out very much. They told me that they’d figured I didn’t particularly need them, because I hadn’t consistently accepted their invitations, and hadn’t reciprocated with affirmative efforts to see them. They’d felt hurt and pushed away by this. This was shocking to me, because they’d been so good at reading my nonverbals, so I figured they were just a perceptive person (fundamental attribution error and halo effect!).

One hypothesis for why this happened is:

Reading verbals is hard.

The illusion of transparency is hard to beat. People who have ready access to nonverbal communication have no need to develop the patches I built around inferring preferences from others’ behavior and explicit reports, so they don’t - they read ongoing clues, body language, facial expressions, and vocal tone instead. People who are very similar to those around them don’t need to construct an explicit model of the generic person, either. Instead, they can just put themselves in the other person’s shoes.

My other hypothesis is:

Self-descriptions tend to be false

Let’s go back to my pattern of asking for things only once:

When I was a kid, I wanted to see the movie Toy Story. So I asked my mom if I could see it, and she said she’d think about it. I’m still waiting.

What was my model of communication, and how did it differ from the conventional model to produce this misunderstanding?

When I said that I wanted to see the movie Toy Story, I meant that I had considered the decision, taken into account the fact that there are many movies available and would be many more in the future, and decided that this one was a priority. I’d had the initial flash of wanting, noticed that it stayed around, and waited for a good opportunity to express it (a car ride in which there was no other conversation).

My mom assumed that I was expressing a momentary preference as soon as it came to mind. If it were persistent, then surely it would come to mind repeatedly, and I’d express it repeatedly.

By this hypothesis, most people don’t have a single well-fleshed-out model of themselves and their preferences. They have a bunch of partial models that make explicit verbal claims far beyond their evidence base - but the verbal claim isn’t meant to be a promise about future behavior. For example, when a partner told me “I hate boundaries,” she didn’t mean that I should consider myself absolutely prohibited from using them in any way because they were bad for her (which is how I took it) - just that she was experiencing some negative feelings towards them.

I think this explains why, when other people talk about being a bad person, it sounds like gibberish to me - and why it doesn’t to them. One of their many verbal framings for talking about themselves has assigned the label “bad” to one other part, to express a feeling they’re experiencing. It’s not that they reflectively endorse the claim that they, on the whole, are intrinsically bad.

Reading verbal communication: an example

Here’s a real-life conversation I had recently over text chat in which I did a reasonably good job drawing nonverbal inferences. For context, they had just accepted my invitation to hang out:

Acquaintance: I'm on meds right now so I'm significantly better at following up on social things. Still not great, though.

Me: Hmm

Would you like me to repeatedly ask you to do things even if you repeatedly turn me down or don't respond?

(Fine either way, just curious)

Acquaintance: No, that would make it significantly worse. My badness at following up comes from me fearing that the other person is more invested in social activity than me.

Me: (Just to be clear, I meant: if you turn down one thing, do you want me to invite you to another thing ever? Not repeated invitations to the same thing.)

Acquaintance: Ooh.

Yeah, that's different.

That sounds nice.

Me: OK, good. Glad I clarified 🙂

What happened here? I’d previously noticed a slightly surprising pattern of behavior. This new acquaintance (whom I hope to become friends with) had on occasion indicated enthusiasm for hanging out, other times turned down my invitations, but very rarely reciprocated. I entertained the hypotheses that:

1. They wanted to see me but for whatever reason lacked either the energy or confidence to initiate.
2. That they felt good about seeing me but not good enough to make it an affirmative priority.
3. They actually didn’t want to see me but were just being nice because they don’t like rejecting people overtly.

So when they brought up this pattern, I decided to gather some information. “How much do you like me?” isn’t a question I generally expect to get an honest or useful answer to, so I decided to ask directly how they’d feel about a behavior pattern that would help them under hypothesis 1, but harm them under hypothesis 3.

At first, their response seemed to me consistent with hypothesis 2 or 3 - but something about this interpretation rang false. I understood why repeated requests might not be good under those hypotheses. I even understood why that would make them less interested in responding at all to the average request. But what I didn’t understand was why they’d volunteer this information immediately after apparently enthusiastically accepting an invitation to hang out!

So I thought about how they could have read my question, that made their response make perfect sense. And then it was obvious. I’d meant to ask whether I should disregard declined invitations as evidence that I shouldn’t repeat the process. If they’d instead believed I was talking about badgering them about the same event, then it makes perfect sense that someone who initially wanted to hang out sometimes might start feeling guilty about not responding and want to hide from me.

So I gave an explicit verbal clarification - and it turned out that this hypothesis was correct!