The Spotlight

Author: Alicorn. Link to original: http://lesswrong.com/lw/1za/the_spotlight/ (English).
Tags: lesswrong, Rationality Submitted by Remlin 14.01.2013. Public material.

Translations of this material:

into Russian: Луч света. Translation complete.
Submitted for translation by Remlin 14.01.2013 Published 5 years, 5 months ago.

Text

Inspecting thoughts is easier and more accurate if they aren't in your head. Look at them in another form from the outside, like they belonged to someone else.

You may find your understanding of this post significantly improved if you read the fourth story from Seven Shiny Stories.

One problem with introspection is that the conclusions you draw about your thoughts are themselves thoughts. Thoughts, of course, can change or disappear before you can extract information about yourself from them. If a flash of unreasonable anger crosses my mind, this might stick around long enough to make me lash out, but then vanish before I discover how unreasonable it was. If thoughts weren't slippery like this, luminosity wouldn't be much of a project. So of course, if you're serious about luminosity, you need a way to pin down your thoughts into a concrete format that will hold still.

You have to pry your thoughts out of your brain.

Writing is the obvious way to do this - for me, anyway. You don't have to publicize what you extract, so it doesn't have to be aesthetic or skillful, just serviceable for your own reference. The key is to get it down in a form that you can look at without having to continue to introspect. Whether this means sketching or scribing or singing, dump your brain out into the environment and have a peek. It's easy to fool yourself into thinking that a given idea makes sense; it's harder to fool someone else. Writing down an idea automatically engages the mechanisms we use to communicate to others, helping you hold your self-analysis to a higher standard.

To turn your thoughts into non-thoughts, use labels to represent them. Put them in reference classes, so that you can notice when the same quale, habit of inference, or thread of cognition repeats. That way, you can detect patterns: "Hey, the last time I felt like this, I said something I really regretted; I'd better watch it." If you can tell when something has happened twice, you can tell when it hasn't - and new moods or dispositions are potentially very important. They mean that you or something around you has changed, and that could be a valuable resource or a tricky hazard.

Your labels can map onto traditional terms or not - if you want to call the feeling of having just dropped your ice cream on the sidewalk "blortrath", no one will stop you. (It can be useful, later when you're trying to share your conclusions about yourself with others, to have a vocabulary of emotion that overlaps significantly with theirs; but you can always set up an idiolect-to-dialect dictionary later.) I do recommend identifying labeled items as being more or less similar to each other (e.g. annoyance is more like fury than it is like glee) and having a way to account for that in your symbolism. Similarities like that will make it more obvious how you can generalize strategies from one thing to another.

Especially if you don't think in words, you might find it challenging to turn your thoughts into something in the world that represents them. Maybe, for instance, you think in pictures but aren't at all good at drawing. This is one of the steps in luminosity that I think is potentially dispensible, so if you honestly cannot think of any way to jot down the dance of your mind for later inspection, you can just work on thinking very carefully such that if something were to be out of place the next time you came back to your thought, you'd notice it. I do recommend spending at least five to ten minutes trying to write, diagram, draw, mutter, or interpretive-dance your mental activity before you give it up as untenable for you, however.

Once you have produced a visible or audible translation of your thoughts, analyze it the way you would if someone else had written it. (Except inasmuch as it's in a code that's uniquely understandable to you and you shouldn't pretend to do cryptanalysis on it.) What would you think of the person described if you didn't know anything else? How would you explain these thoughts? What threads of reasoning seem to run in the background from one belief to another, or from a perception to a belief, or from a desire to an intention? What do you expect this person to do next? What's your next best guess after that? And: what more do you want to know? If you met the person described, how could you satisfy your curiosity without relying on the bias-laden answer you'd get in response to a verbal inquiry? Try it now - in a comment under this post, if you like: note what you're thinking, as much of it as you can grab and get down. Turn on the anti-kibitzer and pretend someone else said it: what must be going on in the mind behind this writing?