Highlights and Shadows

Author: Alicorn. Link to original: http://lesswrong.com/lw/209/highlights_and_shadows/ (English).
Tags: lesswrong, Rationality Submitted by Remlin 15.01.2013. Public material.

Translations of this material:

into Russian: Блики и тени. Translation complete.
Submitted for translation by Remlin 15.01.2013 Published 4 years, 11 months ago.


Part of a good luminosity endeavor is to decide what parts of yourself you do and don't like.

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

As you uncover and understand new things about yourself, you might find that you like some of them, but don't like others. While one would hope that you'd be generally pleased with yourself, it's a rare arrogance or a rarer saintliness that would enable unlimited approval. Fortunately, as promised in post two, luminosity can let you determine what you'd like to change as well as what's already present.

But what to change?

An important step in the luminosity project is to sort your thoughts and feelings not only by type, correlation, strength, etc, but also by endorsement. You endorse those thoughts that you like, find representative of your favorite traits, prefer to see carried into action, and wish to keep intact (at least for the duration of their useful lives). By contrast, you repudiate those thoughts that you dislike, consider indicative of negative characteristics, want to keep inefficacious, and desire to modify or be rid of entirely.

Deciding which is which might not be trivial. You might need to sift through several orders of desire before finally figuring out whether you want to want cake, or like liking sleep, or prefer your preference for preferentism. A good place to start is with your macro-level goals and theoretical commitments (e.g., when this preference is efficacious, does it serve your Life Purpose™, directly or indirectly? if you have firm metaethical notions of right and wrong, is this tendency you have uncovered in yourself one that impels you to do right things?).

As a second pass, you can work with the information you collected when you correlated your ABCs. How does an evaluated desire makes you feel when satisfied or unsatisfied? Does it cripple you when unsatisfied or improve your performance when satisfied? Are you reliably in a position to satisfy it? If you can't typically satisfy it, would it be easier to change the desire or to change the circumstances that prevent its satisfaction? However, this is a second step. You need to know what affect and behavior are preferable to you before you can judge desires (and other mental activity) relative to what they yield in those departments, and judging affect and behavior is itself an exercise in endorsement and repudiation.

Knowing what you like and don't like about your mind is a fine thing. Once you have that information, you can put it to direct use immediately - I find it useful to tag many of my expressions of emotion with the words "endorsed" or "non-endorsed". That way, the people around me can use that categorization rather than having to either assume I approve of everything I feel, or layer their own projections of endorsement on top of me. Either would be unreliable and cause people to have poor models of me: I have not yet managed to excise my every unwanted trait, and my patterns of endorsement do not typically map on to the ones that the people around me have or expect me to have.

Additionally, once you know what you like and don't like about your mind, you can begin to make progress in increasing the ratio of liked to unliked characteristics. People often make haphazard lurches towards trying to be "better people", but when "better" means "lines up more closely with vaguely defined commonsense intuitions about morality", this is not the sort of goal we're at all good at pursuing. Specific projects like being generous or more mindful are a step closer, but the greatest marginal benefit in self-revision comes of figuring out what comes in advance of behaving in a non-endorsed way and heading it off at the pass. (More on this in "Lampshading".) The odds are low that your brain's patterns align closely with conventional virtues well enough for them to be useful targets. It's a better plan to identify what's already present, then endorse or repudiate these pre-sliced thoughts and work on them as they appear instead of sweeping together an unnatural category.