Labels
Okay, I have several half-written posts in various places on labels and labeling, and nothing has gelled into postable form yet. Maybe this will be the one.
My basic hypothesis is that labeling actions is way more helpful than labeling people. People are complicated and do many things with ambiguous interpretations: if you've labeled them, you've going to bias your interpretation. Which is not to say you shouldn't let past events inform your interpretation as current events (the last 9 times she held out the football she pulled away at the last minute, but this time could be different), but that labeling a person tends to lead you to extrapolate beyond your data.* I find I actually get better at prediction through pattern matching when I stick to labeling actions, if that.
It also *really* pisses people off when you label them, or even give them a hint that they might have been labeled. Fears around being labeled incorrectly drive a lot of stupid, counterproductive behavior in humans.** I think that's what's going on in the The Creepy Debates. Feminists are saying "women have a right to be discomforted by an action" and MRAs are saying "it's hurtful to label me as a creep because I asked you out ineptly". And they are both correct, and the easy fix is to let people label actions as creepy without labeling the person as creepy. It's much easier to say "your actions are open to interpretation" than "who you are as a person is open to interpretation"
*A primer for non-statisticians: technically, any statistical model is only valid within the bounds of the x-axis values comprising your original data, because your model can't possibly know about complications outside it. For example, if you only measure the relationship between temperature and power use in the summer you will incorrectly conclude that lower temperature -> lower power use.
**I Thought It Was Just Me, by Brene Brown, is a really good book on this. It's labeled as a aid to learning how to identify places in your own life where you're acting suboptimal because of fear of labeling, or how you've labeled yourself, but it's primary value to me was making me a better listener.
My basic hypothesis is that labeling actions is way more helpful than labeling people. People are complicated and do many things with ambiguous interpretations: if you've labeled them, you've going to bias your interpretation. Which is not to say you shouldn't let past events inform your interpretation as current events (the last 9 times she held out the football she pulled away at the last minute, but this time could be different), but that labeling a person tends to lead you to extrapolate beyond your data.* I find I actually get better at prediction through pattern matching when I stick to labeling actions, if that.
It also *really* pisses people off when you label them, or even give them a hint that they might have been labeled. Fears around being labeled incorrectly drive a lot of stupid, counterproductive behavior in humans.** I think that's what's going on in the The Creepy Debates. Feminists are saying "women have a right to be discomforted by an action" and MRAs are saying "it's hurtful to label me as a creep because I asked you out ineptly". And they are both correct, and the easy fix is to let people label actions as creepy without labeling the person as creepy. It's much easier to say "your actions are open to interpretation" than "who you are as a person is open to interpretation"
*A primer for non-statisticians: technically, any statistical model is only valid within the bounds of the x-axis values comprising your original data, because your model can't possibly know about complications outside it. For example, if you only measure the relationship between temperature and power use in the summer you will incorrectly conclude that lower temperature -> lower power use.
**I Thought It Was Just Me, by Brene Brown, is a really good book on this. It's labeled as a aid to learning how to identify places in your own life where you're acting suboptimal because of fear of labeling, or how you've labeled yourself, but it's primary value to me was making me a better listener.