I want to say that I really like the Sazen → expansion format, and I like the explanation → ways you might feel → ways a request might look format even more.
1 to 4 and 6 to 9 I just straightforwardly agree with.
My issue with 5 should properly be its own blog post but the too-condensed version is something like, those cases where the other person is not also trying to converge on truth are common enough and important enough that I don’t blame someone for not starting from that assumption. Put another way, all of the other rules seem like they work even if the other person isn’t doing them, or at least fail gracefully. Following an unarticulated version of rule 5 has ever failed me badly before. I don’t know exactly what would falsify this claim and acknowledge that it might just be one or two highly salient-to-me examples, but if at the end of the conversation someone is going to smile at you and hand you a pen and ask you to sign something then I don’t think assuming your interlocutor is also aiming for convergence on truth is a good idea.
It is possible that kind of conversation is not covered under what you’re thinking of by a rational discourse.
With 10, I think I’m quibbling over phrasing. Holding to the absolute highest standard feels like it translates to never doing it, which is correct on the margin.
The most interesting disagreement I have is with 0. Years after adopting them, many of the changes I’ve made to how I communicate have become close to free. It is harder now to do the wrong thing, to make an ad hominem or to say I’m a hundred percent sure, in much the same way that I have to make an effort to stop myself from looking at my card in Hanabi (because when I draw a card and add it to my hand, the vast majority of card games have me look at that card!) or the way I have to make an effort to fall off a bicycle when my body knows how to balance and keep moving. I think the 0th guideline is a useful nudge in the right direction, but suspect that if I stick to these guidelines for the next five years, I will feel like it takes less energy to follow them than to break them.
I wholeheartedly agree that following 5 leaves you vulnerable to defection; the claim is that (especially within a subculture like LessWrong) the results of everybody hunting stag on this one are much better than the results of everyone choosing rabbit; you will once in a while get taken advantage of for an extra minute or two/a few more rounds of the back-and-forth, but the base rate of charity in the water supply goes WAY up and this has a bunch of positive downstream effects and is worth it on net in expectation (claim).
(This is elaborated on a good bit in the expansion of 5 if you haven’t read it and are curious. I’d love to be tagged in an objection post, b/c I’d probably engage substantially in the comments.)
10 should maybe be toned back a bit!
I strongly agree with your take on 0; this is hit pretty hard in the Sapir-Whorf piece from a couple of days ago (the thing feels effortful when it’s not reflecting your inner thought processes but using the language can update the inner thought processes, and speaking in a way that reflects the inner thought processes is subjectively ~0% extra effort). But if we’re wanting to gain new skill and not just stay as-good-at-discourse as we currently are, we’re each going to need to be nonzero trying on some axis.
I had read the expansions. We might be in practical agreement on 5. I would say if you’re debating in the comments of a Less Wrong thread, following 5 is positive expected value. You’ll avoid escalations that you’d otherwise fall into, and being defected against won’t cost you too much. It stands out to me because other rules guidelines (say, 2 and 8) I would be comfortable abiding holding myself to even if I knew my interlocutor wasn’t going to follow them. It’s when you’re having higher stakes discussions that leaving yourself open in good faith can go badly. (Edited when it was pointed out to me that they’re always referred to as guidelines, never rules. This is in fact a useful distinction I let blur.)
I agree that when starting out 0 is likely to require energy, like, more energy than it feels like it should to do something like this. “Expect good discourse to require energy until you become very used to it, then it should feel natural” is a weaker message but is how I am interpreting it. (I am trying and failing to find the sequences post about rules phrased as absolutes that aren’t actually absolutes, such that it stays your hand until need actually weighs you down.)
I will be sure to direct your attention to the objection post once I write it. It is partially written already and did not start life as an objection, but it does apply and will be finished. . . someday.
Since I haven’t said so yet, thank you for writing this and giving me a link to reference!
Er, this is maybe too nitpicky, but it’s pretty important to me that these are guidelines, not rules (with expansion on what a guideline means); I worked hard to make sure that the word “rules” appeared nowhere in the text outside of one quote in the appendix.
I want to say that I really like the Sazen → expansion format, and I like the explanation → ways you might feel → ways a request might look format even more.
1 to 4 and 6 to 9 I just straightforwardly agree with.
My issue with 5 should properly be its own blog post but the too-condensed version is something like, those cases where the other person is not also trying to converge on truth are common enough and important enough that I don’t blame someone for not starting from that assumption. Put another way, all of the other rules seem like they work even if the other person isn’t doing them, or at least fail gracefully. Following an unarticulated version of rule 5 has ever failed me badly before. I don’t know exactly what would falsify this claim and acknowledge that it might just be one or two highly salient-to-me examples, but if at the end of the conversation someone is going to smile at you and hand you a pen and ask you to sign something then I don’t think assuming your interlocutor is also aiming for convergence on truth is a good idea.
It is possible that kind of conversation is not covered under what you’re thinking of by a rational discourse.
With 10, I think I’m quibbling over phrasing. Holding to the absolute highest standard feels like it translates to never doing it, which is correct on the margin.
The most interesting disagreement I have is with 0. Years after adopting them, many of the changes I’ve made to how I communicate have become close to free. It is harder now to do the wrong thing, to make an ad hominem or to say I’m a hundred percent sure, in much the same way that I have to make an effort to stop myself from looking at my card in Hanabi (because when I draw a card and add it to my hand, the vast majority of card games have me look at that card!) or the way I have to make an effort to fall off a bicycle when my body knows how to balance and keep moving. I think the 0th guideline is a useful nudge in the right direction, but suspect that if I stick to these guidelines for the next five years, I will feel like it takes less energy to follow them than to break them.
I wholeheartedly agree that following 5 leaves you vulnerable to defection; the claim is that (especially within a subculture like LessWrong) the results of everybody hunting stag on this one are much better than the results of everyone choosing rabbit; you will once in a while get taken advantage of for an extra minute or two/a few more rounds of the back-and-forth, but the base rate of charity in the water supply goes WAY up and this has a bunch of positive downstream effects and is worth it on net in expectation (claim).
(This is elaborated on a good bit in the expansion of 5 if you haven’t read it and are curious. I’d love to be tagged in an objection post, b/c I’d probably engage substantially in the comments.)
10 should maybe be toned back a bit!
I strongly agree with your take on 0; this is hit pretty hard in the Sapir-Whorf piece from a couple of days ago (the thing feels effortful when it’s not reflecting your inner thought processes but using the language can update the inner thought processes, and speaking in a way that reflects the inner thought processes is subjectively ~0% extra effort). But if we’re wanting to gain new skill and not just stay as-good-at-discourse as we currently are, we’re each going to need to be nonzero trying on some axis.
I had read the expansions. We might be in practical agreement on 5. I would say if you’re debating in the comments of a Less Wrong thread, following 5 is positive expected value. You’ll avoid escalations that you’d otherwise fall into, and being defected against won’t cost you too much. It stands out to me because other
rulesguidelines (say, 2 and 8) I would be comfortableabidingholding myself to even if I knew my interlocutor wasn’t going to follow them. It’s when you’re having higher stakes discussions that leaving yourself open in good faith can go badly. (Edited when it was pointed out to me that they’re always referred to as guidelines, never rules. This is in fact a useful distinction I let blur.)I agree that when starting out 0 is likely to require energy, like, more energy than it feels like it should to do something like this. “Expect good discourse to require energy until you become very used to it, then it should feel natural” is a weaker message but is how I am interpreting it. (I am trying and failing to find the sequences post about rules phrased as absolutes that aren’t actually absolutes, such that it stays your hand until need actually weighs you down.)
I will be sure to direct your attention to the objection post once I write it. It is partially written already and did not start life as an objection, but it does apply and will be finished. . . someday.
Since I haven’t said so yet, thank you for writing this and giving me a link to reference!
Er, this is maybe too nitpicky, but it’s pretty important to me that these are guidelines, not rules (with expansion on what a guideline means); I worked hard to make sure that the word “rules” appeared nowhere in the text outside of one quote in the appendix.
You are correct and that is an important distinction I blurred in my own head, thank you.