I started writing this in response to a thread about “safety”, but it got long enough to warrant breaking out into its own thing.
I think it’s important to people to not be attacked physically, mentally, or socially. I have a terminal preference over this, but also think it’s instrumental towards truth-seeking activities too. In other words, I want people to actually be safe.
I think that when people feel unsafe and have defensive reactions, this makes their ability to think and converse much worse. It can push discussion from truth-seeking exchange to social war.
Here I think mr-hire has a point: if you don’t address people’s “needs” overtly, they’ll start trying to get them covertly, e.g. trying to win arguments for the sake of protecting their reputation rather than trying to get to the truth. Doing things like writing hasty scathing replies rather slow, carefully considered ones (*raises hand*), and worse, feeling righteous anger while doing so. Having thoughts like “the only reason my interlocutor could think X is because they are obtuse due to their biases” rather than “maybe they have I point I don’t fully realize” (*raises hand*).
I want to avoid people being harmed and also feeling like they won’t be harmed (but in a truth-tracking way: if you’re likely to be attacked, you should believe it). I also think that protective measures are extremely risky themselves for truth-seeking. There is a legitimate fear here a) people can use the protections to silence things they don’t like hearing, b) it may be onerous and stifle honest expression to have to constrain one’s speech, c) fear of being accused of harming others stifles expression of true ideas, d) these protections will get invoked in all kind of political games.
I think the above a real dangers. I also think it’s dangerous to have no protections against people being harmed, especially if they’re not even allowed to object to be harmed. In such an arrangement, it becomes too easy to abuse the “truth-seeking free speech” protections to socially attack and harm people while claiming impunity. Some of it’s truth-seeking ability lost to becoming partly a vicious social arena.
I present the Monkey-Shield Allegory (from an unpublished post of mine):
Take a bunch of clever monkeys who like to fight with each other (perhaps they throw rocks). You want to create peace between them, so you issue them each with a nice metal shield which is good at blocking rocks. Fantastic! You return the next day, and you find that the monkeys are hitting each with the mental shields (turns out if you whack someone with a shield, their shield doesn’t block all the force of the blow and it’s even worse than fighting with rocks).
I find it really non-obvious what the established norms and enforced policies should be. I have guesses, including a proposed set of norms which are being debated in semi-private and should be shared more broadly soon.Separate from the question I have somewhat more confidence in the following points and what they imply for individual.
1. You should care about other people and their interests. Their feelings are 1) real and valuable, and 2) often real information about important states of the world for their wellbeing. Compassion is a virtue.
Even if you are entirely selfish, understanding and caring about other people is instrumentally advantageous for your own interests and for the pursuit of truth.
2. Even failing 1, you should try hard to avoid harming people (i.e. attacking them) and only do so when you really mean to. It’s not worth it to accidentally do it if you don’t mean to.
3. I suspect many people of possessing deep drives to always be playing monkey-political games, and these cause them to want to win points against each other however they can. Ways to do that include being aggressive, insulting people, etc, baiting them, and all the standard behaviors people engage in online forums.
These drives are anti-cooperative, anti-truth, and zero-sum. I basically think they should be inhibited and instead people should cultivate compassion and ability to connect.
I think people acting in these harmful ways often claim their behaviors are fine by attributing to some more defensible cause. I think there are defensible reasons for some behaviors, but I get really suspicious when someone consistently behaves in a way that doesn’t further their stated aims.
People getting defensive are often correctly perceiving that they are being attacked by others. This makes me sympathetic to many cases of people being triggered.
4. Beyond giving up on the monkey-games, I think that being considerate and collaborative (including the meta-collaborative within a Combat culture) costs relatively little most of the time. There might be some upfront costs to change one’s habits and learn to be sensitive, but long run the value of learning them pays off many times over in terms of being able to have productive discussions where no one is getting defensive + plus that seems intrinsically better for people to be having a good time. Pleasant discussions provoke more pleasant discussions, etc.
* I am not utterly confident in the correctness of 4. Perhaps my brain devotes more cycles to being considerate and collaborative than I realize (as this slowly ramped up over the years) and it costs me real attention that could go directly to object-level thoughts. Despite the heavy costs, maybe it is just better to not worry about what’s going on in other people’s minds and not expend effort optimizing for it. I should spend more time trying to judge this.
5. It is good to not harm people, but it also good to build one’s resilience and “learn to handle one’s feelings.” That is just plainly an epistemically virtuous thing to do. One ought to learn how to become less often and also how to operate sanely and productively while defensive. Putting all responsibility onto others for your psychological state is damn risky. Also 1) people who are legitimately nasty sometimes still have stuff worth listening to, you don’t want to give up on that; 2) sometimes it won’t be the extraneous monkey-attack stuff that is upsetting, and instead the core topic—you want to be able to talk about that, 3) misunderstandings arise easily and it’s easy to feel attacked when you aren’t being, some hardiness to protection again misunderstandings rapidly spiralling into defensiveness and demonthreads.
6. When discussing topics online, in-text, and with people you don’t know, it’s very easy to miscalibrated on intentions and the meaning behind words (*raises hand*). It’s easy for their to be perceived attacks even when no attacks are intended (this is likely the result of a calibrated prior on the prevalence of social attacks).
a. For this reason, it’s worth being a little patient and forgiving. Some people talk a bit sarcastically to everyone (which is maybe bad), but it’s not really intended as an attack on you. Or perhaps they were plainly critical, but they were just trying to help.
b. When you are speaking, it’s worth a little extra effort to signal that you’re friendly and don’t mean to attack. Maybe you already know that and couldn’t imagine otherwise, but a stranger doesn’t. What counts as an honest signal of friendly intent is anti-inductive, if we declare to be something simple, the ill-intentioned by imitate it by rote, go about their business, and the signal will lose all power to indicate the friendliness. But there are lots of cheap ways to indicate you’re not attacking, that you have “good will”. I think they’re worth it.
In established relationships where the prior has become high that you are not attacking, less and less effort needs to be expended on signalling your friendly intent and you can get talk plainly, directly, and even a bit hostilly (in a countersignalling way). This is what my ideal Combat culture looks like, but it relies of having a prior and common knowledge established of friendliness. I don’t think it works to just “declare it by fiat.”
I’ve encountered push back when attempting to 6b. I’ll derive two potential objections (which may not be completely faithful to those originally raised):
Objection 1: No one should be coerced into having to signal friendliness/maintain someone else’s status/generally worry about what impact their saying true things will have. Making them worry about it impedes the ability to say true things which is straightforwardly good.
Response: I’m trying to coerce anyone into doing this. I’m trying to make the case you should want to do this of your own accord. That this is good and worth it and in fact results in more truth generation than otherwise. It’s a good return of investment. There might be an additional fear that if I promote this as virtuous behavior, it might have the same truth-impeding effects as it if was policy. I’m not sure, I have to think about that last point more.
Objection 2: If I have to signal friendly intent when I don’t mean it, I’d be lying.
Response: Then don’t signal friendly intent. I definitely don’t want anyone to pretend or go through the motions. However I do think you should probably be trying to have honestly friendly intent. I expect conversations with friendly intent to be considerable better than those without (this is something of a crux for me here), so if you don’t have it towards someone, that’s real unfortunate, and I am pessimistic about the exchange. Barring exceptional circumstances, I generally don’t want to talk to people who do not have friendly intent/desire to collaborate (even just at the meta-level) towards me.
What do I mean by friendly intent? I mean that you don’t have goals to attack, win, or coerce. It’s an exchange intended for the benefit of both parties where you’re not the side acting in a hostile way. I’m not pretending to discuss a topic with you when actually I think you’re an idiot and want to demonstrate it to everyone, etc., I’m not trying to get an emotional reaction for my own entertainment, I’m not just trying to win with rhetoric rather than actually expose my beliefs and cruxes, if I’m criticizing, I’m not just trying to destroy you, etc. As above, many times this is missing and it’s worth trying to signal its presence.
If it’s absent, i.e. you actually want to remove someone from the community or think everyone should disassociate from them, that’s sometimes very necessary. In that case, you don’t have friendly intent and that’s good and proper. Most of the time though (as I will argue), you should have friendly intent and should be able to honestly signal it. Probably I should elaborate and clarify further on my notion of friendly intent.
There are related notions to friendly intent like good faith, questions like “respect your conversation partner”, think you might update based on what they say, etc. I haven’t discussed them, but should.
Some Thoughts on Communal Discourse Norms
I started writing this in response to a thread about “safety”, but it got long enough to warrant breaking out into its own thing.
I think it’s important to people to not be attacked physically, mentally, or socially. I have a terminal preference over this, but also think it’s instrumental towards truth-seeking activities too. In other words, I want people to actually be safe.
I think that when people feel unsafe and have defensive reactions, this makes their ability to think and converse much worse. It can push discussion from truth-seeking exchange to social war.
Here I think mr-hire has a point: if you don’t address people’s “needs” overtly, they’ll start trying to get them covertly, e.g. trying to win arguments for the sake of protecting their reputation rather than trying to get to the truth. Doing things like writing hasty scathing replies rather slow, carefully considered ones (*raises hand*), and worse, feeling righteous anger while doing so. Having thoughts like “the only reason my interlocutor could think X is because they are obtuse due to their biases” rather than “maybe they have I point I don’t fully realize” (*raises hand*).
I want to avoid people being harmed and also feeling like they won’t be harmed (but in a truth-tracking way: if you’re likely to be attacked, you should believe it). I also think that protective measures are extremely risky themselves for truth-seeking. There is a legitimate fear here a) people can use the protections to silence things they don’t like hearing, b) it may be onerous and stifle honest expression to have to constrain one’s speech, c) fear of being accused of harming others stifles expression of true ideas, d) these protections will get invoked in all kind of political games.
I think the above a real dangers. I also think it’s dangerous to have no protections against people being harmed, especially if they’re not even allowed to object to be harmed. In such an arrangement, it becomes too easy to abuse the “truth-seeking free speech” protections to socially attack and harm people while claiming impunity. Some of it’s truth-seeking ability lost to becoming partly a vicious social arena.
I present the Monkey-Shield Allegory (from an unpublished post of mine):
I find it really non-obvious what the established norms and enforced policies should be. I have guesses, including a proposed set of norms which are being debated in semi-private and should be shared more broadly soon.Separate from the question I have somewhat more confidence in the following points and what they imply for individual.
1. You should care about other people and their interests. Their feelings are 1) real and valuable, and 2) often real information about important states of the world for their wellbeing. Compassion is a virtue.
Even if you are entirely selfish, understanding and caring about other people is instrumentally advantageous for your own interests and for the pursuit of truth.
2. Even failing 1, you should try hard to avoid harming people (i.e. attacking them) and only do so when you really mean to. It’s not worth it to accidentally do it if you don’t mean to.
3. I suspect many people of possessing deep drives to always be playing monkey-political games, and these cause them to want to win points against each other however they can. Ways to do that include being aggressive, insulting people, etc, baiting them, and all the standard behaviors people engage in online forums.
These drives are anti-cooperative, anti-truth, and zero-sum. I basically think they should be inhibited and instead people should cultivate compassion and ability to connect.
I think people acting in these harmful ways often claim their behaviors are fine by attributing to some more defensible cause. I think there are defensible reasons for some behaviors, but I get really suspicious when someone consistently behaves in a way that doesn’t further their stated aims.
People getting defensive are often correctly perceiving that they are being attacked by others. This makes me sympathetic to many cases of people being triggered.
4. Beyond giving up on the monkey-games, I think that being considerate and collaborative (including the meta-collaborative within a Combat culture) costs relatively little most of the time. There might be some upfront costs to change one’s habits and learn to be sensitive, but long run the value of learning them pays off many times over in terms of being able to have productive discussions where no one is getting defensive + plus that seems intrinsically better for people to be having a good time. Pleasant discussions provoke more pleasant discussions, etc.
* I am not utterly confident in the correctness of 4. Perhaps my brain devotes more cycles to being considerate and collaborative than I realize (as this slowly ramped up over the years) and it costs me real attention that could go directly to object-level thoughts. Despite the heavy costs, maybe it is just better to not worry about what’s going on in other people’s minds and not expend effort optimizing for it. I should spend more time trying to judge this.
5. It is good to not harm people, but it also good to build one’s resilience and “learn to handle one’s feelings.” That is just plainly an epistemically virtuous thing to do. One ought to learn how to become less often and also how to operate sanely and productively while defensive. Putting all responsibility onto others for your psychological state is damn risky. Also 1) people who are legitimately nasty sometimes still have stuff worth listening to, you don’t want to give up on that; 2) sometimes it won’t be the extraneous monkey-attack stuff that is upsetting, and instead the core topic—you want to be able to talk about that, 3) misunderstandings arise easily and it’s easy to feel attacked when you aren’t being, some hardiness to protection again misunderstandings rapidly spiralling into defensiveness and demonthreads.
6. When discussing topics online, in-text, and with people you don’t know, it’s very easy to miscalibrated on intentions and the meaning behind words (*raises hand*). It’s easy for their to be perceived attacks even when no attacks are intended (this is likely the result of a calibrated prior on the prevalence of social attacks).
a. For this reason, it’s worth being a little patient and forgiving. Some people talk a bit sarcastically to everyone (which is maybe bad), but it’s not really intended as an attack on you. Or perhaps they were plainly critical, but they were just trying to help.
b. When you are speaking, it’s worth a little extra effort to signal that you’re friendly and don’t mean to attack. Maybe you already know that and couldn’t imagine otherwise, but a stranger doesn’t. What counts as an honest signal of friendly intent is anti-inductive, if we declare to be something simple, the ill-intentioned by imitate it by rote, go about their business, and the signal will lose all power to indicate the friendliness. But there are lots of cheap ways to indicate you’re not attacking, that you have “good will”. I think they’re worth it.
In established relationships where the prior has become high that you are not attacking, less and less effort needs to be expended on signalling your friendly intent and you can get talk plainly, directly, and even a bit hostilly (in a countersignalling way). This is what my ideal Combat culture looks like, but it relies of having a prior and common knowledge established of friendliness. I don’t think it works to just “declare it by fiat.”
I’ve encountered push back when attempting to 6b. I’ll derive two potential objections (which may not be completely faithful to those originally raised):
Objection 1: No one should be coerced into having to signal friendliness/maintain someone else’s status/generally worry about what impact their saying true things will have. Making them worry about it impedes the ability to say true things which is straightforwardly good.
Response: I’m trying to coerce anyone into doing this. I’m trying to make the case you should want to do this of your own accord. That this is good and worth it and in fact results in more truth generation than otherwise. It’s a good return of investment. There might be an additional fear that if I promote this as virtuous behavior, it might have the same truth-impeding effects as it if was policy. I’m not sure, I have to think about that last point more.
Objection 2: If I have to signal friendly intent when I don’t mean it, I’d be lying.
Response: Then don’t signal friendly intent. I definitely don’t want anyone to pretend or go through the motions. However I do think you should probably be trying to have honestly friendly intent. I expect conversations with friendly intent to be considerable better than those without (this is something of a crux for me here), so if you don’t have it towards someone, that’s real unfortunate, and I am pessimistic about the exchange. Barring exceptional circumstances, I generally don’t want to talk to people who do not have friendly intent/desire to collaborate (even just at the meta-level) towards me.
What do I mean by friendly intent? I mean that you don’t have goals to attack, win, or coerce. It’s an exchange intended for the benefit of both parties where you’re not the side acting in a hostile way. I’m not pretending to discuss a topic with you when actually I think you’re an idiot and want to demonstrate it to everyone, etc., I’m not trying to get an emotional reaction for my own entertainment, I’m not just trying to win with rhetoric rather than actually expose my beliefs and cruxes, if I’m criticizing, I’m not just trying to destroy you, etc. As above, many times this is missing and it’s worth trying to signal its presence.
If it’s absent, i.e. you actually want to remove someone from the community or think everyone should disassociate from them, that’s sometimes very necessary. In that case, you don’t have friendly intent and that’s good and proper. Most of the time though (as I will argue), you should have friendly intent and should be able to honestly signal it. Probably I should elaborate and clarify further on my notion of friendly intent.
There are related notions to friendly intent like good faith, questions like “respect your conversation partner”, think you might update based on what they say, etc. I haven’t discussed them, but should.