Sometimes when you work at a large tech-focused company, you'll be pulled into a required-but-boring all-day HR meeting to discuss some asinine topic like "communication styles".

If you've had the misfortune fun of attending one of those meetings, you might remember that the topic wasn't about teaching a hypothetically "best" or "optimal" communication style. The goal was to teach employees how to recognize when you're speaking to someone with a different communication style, and then how to tailor your understanding of what they're saying with respect to them. For example, some people are more straightforward than others, so a piece of seemingly harsh criticism like "This won't work for XYZ reason." doesn't mean that they disrespect you -- they're just not the type of person who would phrase that feedback as "I think that maybe we've neglected to consider the impact of XYZ on the design."

I have read the many pages of debate on this current disagreement over the past few days. I have followed the many examples of linked posts that were intended to show bad behavior by one side or the other.

I think Zack and gjm have a good job at communicating with each other despite differences in their preferred communication styles, and in particular, I agree strongly with gjm's analysis:

I think this is the purpose of Duncan's proposed guideline 5. Don't engage in that sort of adversarial behaviour where you want to win while the other party loses; aim at truth in a way that, if you are both aiming at truth, will get you both there. And don't assume that the other party is being adversarial, unless you have to, because if you assume that then you'll almost certainly start doing the same yourself; starting out with a presumption of good faith will make actual good faith more likely.

And then with Zack's opinion:

That said, I don't think there's a unique solution for what the "right" norms are. Different rules might work better for different personality types, and run different risks of different failure modes (like nonsense aggressive status-fighting vs. nonsense passive-aggressive rules-lawyering). Compared to some people, I suppose I tend to be relatively comfortable with spaces where the rules err more on the side of "Punch, but be prepared to take a punch" rather than "Don't punch anyone"—but I realize that that's a fact about me, not a fact about the hidden Bayesian structure of reality. That's why, in "'Rationalist Discourse' Is Like 'Physicist Motors'", I made an analogy between discourse norms and motors or martial arts—there are principles governing what can work, but there's not going to be a unique motor, a one "correct" martial art. 

I also agree with Zack when they said:

I'm unhappy with the absence of an audience-focused analogue of TEACH. In the following, I'll use TEACH to refer to making someone believe X if X is right; whether the learner is the audience or the interlocutor B isn't relevant to what I'm saying.

I seldom write comments with the intent of teaching a single person. My target audience is whoever is reading the posts, which is overwhelmingly going to be more than one person.

From Duncan, I agree with the following:

It is in fact usually the case that, when two people disagree, each one possesses some scrap of map that the other lacks; it's relatively rare that one person is just right about everything and thoroughly understands and can conclusively dismiss all of the other person's confusions or hesitations. If you are trying to see and understand what's actually true, you should generally be hungry for those scraps of map that other people possess, and interested in seeing, understanding, and copying over those bits which you were missing.

Almost all of my comments tend to focus on a specific disagreement that I have with the broader community. That disagreement is due to some prior that I hold, that is not commonly held here.

And from Said, I agree with this:

Examples?

This community is especially prone to large, overly-wordy armchair philosophy about this-or-that with almost no substantial evidence that can tie the philosophy back down to Earth. Sometimes that philosophy gets camouflaged in a layer of pseudo-math; equations, lemmas, writing as if the post is demonstrating a concrete mathematical proof. To that end, focusing the community on providing examples is a valuable, useful piece of constructive feedback. I strongly disagree that this is an unfair burden on authors.

EDIT: I forgot to write an actual conclusion. Maybe "don't expect everyone to communicate in the same way, even if we assume that all interested parties care about the truth"?

Thanks for writing this!! There's a number of places where I don't think you've correctly understood my position, but I really appreciate the engagement with the text I published: if you didn't get what I "really meant", I'm happy to do more work to try to clarify.

TEACH, so that B ends up believing X if X is right and Y if Y is right.
CONVINCE, so that B ends up believing X.
EXPOUND, so that the audience ends up believing X.

I'm unhappy with the absence of an audience-focused analogue of TEACH. In the following, I'll use TEACH to refer to making someone believe X if X is right; whether the learner is the audience or the interlocutor B isn't relevant to what I'm saying.

this amounts to supposing that everyone is trying to WIN, hopefully with constraints of honesty, which they do partly by trying to CONVINCE and EXPOUND. We may hope that everyone will LEARN but in Zack's presentation that doesn't seem to be present as a motive at all. [...] you're only going to be able to WIN if you start out disagreeing and argue vigorously for your dissenting opinion

That's not what I meant. In terms of your taxonomy of motives, I would say that people who don't think they have something to TEACH, mostly don't end up writing comments: when I'm only trying to LEARN and don't have anything to TEACH, I usually end up silently reading (and upvoting contributions I LEARNed from) without commenting. LEARNing can sometimes be a motive for commenting: when I think I think an author might have something to TEACH me, but I can't manage to infer it from the text they've already published, I might ask a question. But I do think that's a minority of comments.

The relevance of WINning is as a reward for TEACHing. I think we should be trying to engineer a culture where the gradients of status, esteem, and WINning are aligned with good epistemology—where the way to WIN is by means of TEACHing truths, rather than CONVINCEing people of falsehoods. I am fundamentally pessimistic about efforts to get people to care less about WINning (as contrasted to my approach of trying to align what WINning means in the local culture). If I claimed that my motive for commenting on this website was simply that I altruistically wanted other users of this website to have more accurate beliefs, I would be lying; I just don't think that's how human psychology works.

but this isn't about diverging-the-opposite-of-converging at all, it's using "diverge" to mean "differ at the outset" [...] Converging means getting closer, not starting out already in agreement.

You know, that's a good point! Now that you point it out, I see that the passage you quote is bad writing and sloppy thinking on my part. As a result of you pointing this out, I think it's healthy for people to think (slightly) more of you for pointing out a flaw in the text I published, and (slightly) less of me for publishing flawed text. (And very slightly more of me for conceding the point as I am doing now, rather than refusing to acknowledge it.) You WIN. I think it's okay for you to WIN, and to enjoy WINning. You've earned it!

in the case where I'm right the others' problems needn't be "something wrong with their epistemic process" in any sense that requires disrespect

Isn't it, though?—at least for some relevant sense of the word "disrespect." Previously, you thought Tao was so competent that you never expected to find yourself in the position of thinking he was wrong. Now, you think he was wrong. If you're updating yourself incrementally [LW · GW] about Tao's competence, it seems like your estimate of his competence should be going down. Not by very much! But a little. That's the sense in which disagreement is disrespect. (The phrase comes from the title of the linked Robin Hanson post, which is also linked when I used the phrase in my post; Hanson explicitly acknowledges that the degree of disrespect might be small.)

he's conflating "not aiming to converge on truth" with "being wrong"

So, I tried to clarify what I meant there in the parenthesized paragraph starting with the words "This is with respect to the sense". Did that suffice, or is this more bad writing on my part? I'm not sure; I'm happy to leave it to the reader to decide.

disagree on some difficult question about, say, theoretical physics. The cooperate move [...]

I like the physics debate analogy, but the moral depends on how you map real-world situations to a payoff matrix. When expressing disapproval of the Prisoner's-Dilemma-like frame, it's because I was worried about things analogous to finding errors in the other person's calculation being construed as "defection" (because finding fault in other people's work feels "adversarial" rather than "collaborative").

Zack is more willing than Duncan thinks he should be to decide that other people are bozos

I would particularly emphasize that people can be bozos "locally" but not "globally". If I'm confident that someone is wrong, I don't want to pretend to be more uncertain than I actually am in order to make them feel respected. But I'm also not casting judgement on the totality of their worth as a person; I'm just saying they're wrong on this topic.

I don't think it usually looks like Zack's optimistic picture

I'm glad you noticed the optimism! Yes, I would say that I'm relatively optimistic about the possibility of keeping discussions on track despite status-fighting instincts—and also relatively pessimistic about the prospects of collaborative norms to actually fix the usual problems with status-seeking rather than merely disguising them and creating new problems [LW(p) · GW(p)].

When having a discussion, I definitely try to keep in mind the possibility that the other person is right and I'm wrong. But if the person I'm currently arguing with were to tell me, "I don't feel like you're here to collaborate with me; I think you should be putting in more effort to think of reasons I might be right," that actually makes me think it's less likely that they might be right (even though the generic advice is good). Why? Because giving the advice in this context makes me think they're bluffing: I think if they had an argument, they would stick to the argument (telling me what I'm getting wrong, and possibly condemning my poor reading comprehension if they think they were already adequately clear), rather than trying to rules-lawyer me for being insufficiently collaborative.

That said, I don't think there's a unique solution for what the "right" norms are. Different rules might work better for different personality types, and run different risks of different failure modes (like nonsense aggressive status-fighting vs. nonsense passive-aggressive rules-lawyering). Compared to some people, I suppose I tend to be relatively comfortable with spaces where the rules err more on the side of "Punch, but be prepared to take a punch" rather than "Don't punch anyone"—but I realize that that's a fact about me, not a fact about the hidden Bayesian structure of reality. That's why, in "'Rationalist Discourse' Is Like 'Physicist Motors'" [LW · GW], I made an analogy between discourse norms and motors or martial arts—there are principles governing what can work, but there's not going to be a unique motor, a one "correct" martial art.

I have only read through this once, somewhat quickly, and have not had time to deeply grok and savor it.

But from that one quick read-through, I basically endorse gjm's presentation of my position; it passes my ITT with at least a B+ and plausibly if I read through carefully it would prove to have full marks.

I'm not quite sure how I would rephrase 5, given pushback and time to mull (it is more the case that the given phrasing doesn't do a good job expressing the thing than that I think the thing isn't part of the basics of good rationalist discourse).

There are a handful of pieces to the puzzle, including but not limited to:

1. Good rationalist discourse, as opposed to good unilateralist truth-seeking inquiry, values and prioritizes some kind of collaboration. There are many different ways that a good collaborative truth-seeking conversation can look, ranging from sparring/debate to long trying-to-pass-one-another's-ITT's. But there has to be some kind of being-on-the-same-page about how the conversation itself works, some sort of mutual consent regarding the local rules of engagement, or you end up failing to get anything out of the discussion at best, or end up in conflict or wasting time/energy at worst.
2. Setting "convergence with your partner" as the target/goal/thing to optimize for is wrong, as clearly stated in the original post. Most ways of converging are orthogonal to truth; if your intention is to smooth or strengthen a social connection you're not doing truth-seeking.
3. It is in fact usually the case that, when two people disagree, each one possesses some scrap of map that the other lacks; it's relatively rare that one person is just right about everything and thoroughly understands and can conclusively dismiss all of the other person's confusions or hesitations. If you are trying to see and understand what's actually true, you should generally be hungry for those scraps of map that other people possess, and interested in seeing, understanding, and copying over those bits which you were missing.

So from the perspective of being on a forum about clear thinking, clear communication, and collaborative truth-seeking, i.e. if you came here to play the game that is being played here, there is some combination of prescriptions that goes something like:

1. Don't be an antagonistic, suspicious, hostile dick; if you don't have some amount of baseline trust and charity for the median LWer then you're in the wrong place. Good faith should not be extended infinitely or blindly; people can in fact prove that they don't deserve it, but you should start from a place of "we're probably both here to try to get closer to the truth" because if you don't, then you're going to ruin the potential for collaboration before it has a chance to be realized. Analogy: don't join a soccer team full of people you do not trust to play soccer well and then immediately refuse to coordinate with them because you think they're untrustworthy teammates.
2. That being said, don't prioritize "being a good teammate." Prioritize playing a good game of soccer/doing correct things with the ball. Trust others to receive the ball from you and pass the ball to you until they show you that they don't deserve that trust; don't make "evaluating whether they deserve that trust" or "bending over backwards to prove that you deserve that trust" high priorities. Part of the point of assembling in the garden of LessWrong rather than just doing everything out on the naked internet is that there's a higher baseline of good players around and you can just get down to playing soccer quickly, without a lot of preamble. Figuring out what's true is the actual game, getting closer to truth is the actual goal, and if you're focused on that, you will tend to move closer to others who are also doing that as a side effect.
3. Share the labor of making progress with your conversational partners. This loops back to the Zeroth guideline, of expecting it to take some work, but there's a thing where, like, one person sits back and keeps expecting their partner to (basically) do their bidding, and do all of the heavy lifting, and this isn't playing soccer, this is plopping yourself down in the middle of the field and getting in the way of the other people who now have to play soccer around you and tune out your attempts to call plays from a seated position. 
   
   (Another way to say this is "it isn't anybody's job to convince or educate you—it's both of your jobs to figure out what's true, and the best way to do that is to hand information back and forth to each other, not to demand that other people hand their information to you.")

"Aim for convergence on truth" was my rough attempt to try to say these all in a single mouthful, because I do think there's One Thing here, even though it has some component parts.  It's the thing that people are failing to do if they're prioritizing social smoothness above all else, but it's also the thing that people are failing to do if they act like a hostile jerk; it's something like "effectively leveraging the presence of two humans each in possession of different scraps of the underlying truth."

Some people fail on this one by not even trying to leverage the presence of two humans, and others fail on this by not doing so effectively. But like. There's some potential that is present in [the interaction of two people] that is bigger and better than what a unilateralist can expect to achieve by diving into enemy territory.  That's what the Fifth guideline was trying to gesture at/codify.

This debate feels to me like husband and wife yelling at each other, starting with some object-level thing like "you didn't do the dishes again", but quickly escalating to meta and more meta ("it's not just the stupid dishes, but your general attitude towards...", "no, the real problem is how you communicate...", "no, it's you who cannot communicate, because..."), where almost everything that could be said was already said in the past.

Prediction: the only way this will be helpful is people choosing sides, based on whom they like more. This may solve the situation in the sense that the losing side may admit their loss and leave, or the winning side may feel socially empowered to ban the losing side... but if this is what actually happens, I would hope we are adult enough to realize it (even if commenting on it explicitly may not be the smartest idea, because it adds more potential fuel to the drama, yet another thing to go meta about, etc.).

What should be done instead... no idea, just a general suspicion that going meta/abstract is actually avoiding the painful part.

I like this post! Positive reinforcement. <3

(Why do I think Duncan means that rather than, e.g., “explicitly aim to converge”? Because he wrote this: “If two people disagree, it’s tempting for them to attempt to converge with each other, but in fact the right move is for both of them to try to see more of what’s true.” And, next paragraph: “If you are moving closer to truth [...] then you will inevitably eventually move closer and closer to agreement with all the other agents who are also seeking truth.” I think that’s a little too optimistic, incidentally.)

I’m afraid that it remains unclear to me how this interpretation may be simultaneously described as “aiming for convergence” (in any sense) and “aiming for truth”.

Yes, it seems more or less^[1] correct that if two agents are moving closer to the truth, then they will also move closer to each other. But this does not seem to require any effort whatsoever from these two agents to converge on each other. Indeed it seems to require precisely the opposite: if either agent ever stops having convergence on truth as his only goal, he may well (indeed, is likely to) veer from the path—which, by hypothesis, will actually impede his progress toward convergence with his counterpart!

You say:

It is possible to aim for truth without aiming for convergence even in this sense.

And my question remains: what does it mean to “aim for convergence” “in this sense”, other than aiming for truth? It seems to me that not only is “aim[ing] for truth without aiming for convergence ” possible, it is in fact the only thing that’s possible. In other words, it does not seem possible to aim for truth with aiming for convergence!

For instance, if A is very confident that X is right but has given up hope of convincing B, he may aim to EXPOUND and secondarily to WIN, promoting truth by convincing the audience that X (which A believes) is true and also making them less likely to be (as A sees it) taken in by B. Or A may not care what anyone else ends up thinking and aim only to LEARN, which will produce convergence if it happens that B is right and A is wrong but will likely do little to help B if A was right from the start.

… or, having given up hope (or having never had any hope or interest) of convincing B, A may pursue some other goals [LW(p) · GW(p)] which, however, are not “dishonorable” (as “EXPOUND” and/or “WIN” may be described as being). (Which includes “LEARN”, certainly. But not only that!)

The problem with this decomposition of motives and behaviors is that "truth" is actually elusive, and often poorly-defined in most real debates.  In the case of propositional, testable predictions, prediction markets are awesome because they ALIGN "win" motive with "truth-publication" motive.  But just having multiple participants perform the calculations or share their evidence probably works too.  It's pretty easy to recognize if someone is handwaving or not actually trying to improve their prediction, so walking away is feasible as well.

For discussions of generality or framing or modeling of a situation, that's a whole lot harder.  Even policy or action debates often turn on which generalized principle(s) are most salient, for which "truth" doesn't actually apply.  For these, the best debates seek cruxes, not necessarily convergence.  The disagreement CAN be legitimate from different priors, or different weighting of interpretations of observations, not from incorrect epistemology.  

Here are some goals A might have. …

Additional possible goals that A might have:

• Learn and/or refine his thinking about things other than just X/Y
• What you call “TEACH”, not B, but rather the audience—so that the audience ends up believing whichever of X/Y are true
• Give the audience the opportunity to learn and/or refine their thinking about things other than just X/Y
• Signal (to the audience) his own and/or B’s status, membership in some faction, adherence to some philosophy, possession of some personal characteristics, etc.

Silly nitpick: There is also a non-zero probability that your proof of a statement and Terry Tao's proof of its negation are both valid, which can happen if both of you are making a false assumption. (You prove that Conjecture A implies X, he proves that Conjecture A implies not-X, together you've proven Conjecture A false.)

While somebody might not engage in debate with the goal of gaining status, they can do it with the goal of convincing (whether it be through rationally argued explaining or otherwise), i.e. with the goal of spreading their ideas. That seems to be a very, very powerful motivation in the modern psyche. People are ready to pay considerable sums to obtain the rights of leaving comments on such or such media websites etc. I am always struck by the quasi-Darwinian aspect of this. A lot of us, it seems, seek to spread their ideas just like Darwin says we seek to spread our genes. Ideological Darwinism I would call it - if the name was not already taken for something else.