Human communication as a fragile DWIM system

I’ve noticed a pattern about what gets me most excited about human connection: the feeling of being able to actually communicate with each other, the rare thrill of understanding and being understood.

That rareness comes in two pieces, maybe:

1) Depth; some things I usually just can’t/don’t get across at all most of the time

2) Ease; there will be things I usually can say to people, but it takes a dedicated, tiring effort

I’m trying to point mostly at the second one.  When communication goes really right, it feels like I can just say the words that come to mind, gesturing at some concept off in the world, and the other person just knows what I’m talking about.  I can tell they know (y’know, probably) because they say reply with words that shape the concept with additional detail in a way that would be surprising if they didn’t see what I was trying to gesture at.  They ask questions that jump right to the forefront of my own understanding, and that cut reality at what feel to me like its joints.  We can make up new bits of terminology rapidly, sometimes developing an entire new set of shared jargon just in the one conversation.

It doesn’t feel like I’m doing anything cleverer, communication-wise, in such situations; somehow, mysteriously, we just end up making a hugely unusual amount of sense to each other (it does seem to tend to be mutual).   It reminds me of Scott’s writing advice.

But often somehow later I find that this magic is gone, and I’m back to something like the normal degree of difficulty in communication with the person.  I tend to respond somewhat ungracefully to this, because my naive model always tends to be that the unusually good communication is an essentialist trait, that we’re somehow fitted well to be able to understand each other.  And in the same way that I didn’t know what I was doing so much better at first, I don’t feel as though I’m doing anything worse, later.

My current theory is that high-functioning inter-human communication is operating on something a lot like a “do what I mean” system (my understanding of the phrase is from chapter 13 of Bostrom’s Superintelligence).  The assumption is that when we speak, our concepts are probably never clear, to a first approximation – when we need to be clear, we don’t use normal human language at all, we use math and formal logic, and it takes a long time to say anything (like Entish, perhaps…).

In normal human language, we’re always relying (though it can be to a greater or lesser extent) on the goodwill of our listener.  We need them to assume we’re saying something interesting, or clever, and to search in their concept space for the idea that most reminds them of not only our words, but their model of us.  If someone we think of as a teacher says “do or do not, there is no try,” we nod along even if it doesn’t make any sense, and seek a meaning of the sentence that points in a helpful direction.  If someone we think of as annoying says the same words, we roll our eyes and feel frustrated at their lack of useful content.

And as I’ve discovered, it can even be the same person saying the same words, in different months, and I’ll hear them differently depending on how I feel about them at the time.

Does this have any useful ramifications?  It seems to predict that if I could re-frame my attitude towards someone I used to have great conversations with, I should be able to start having them again.  This would be very useful, if I could figure out how to do it – it doesn’t feel particularly tractable at the moment.

Alternate hypotheses:

1) Most of the untapped skill is actually in my mind, after all, and I just only use it when I’m really excited, which only happens when…?

2) Halo effect causes me to assume vague answers imply understanding rather than ignorance, so it’s just an illusion of transparency.

Leave a Reply

Your email address will not be published. Required fields are marked *