Reading this again three years later: I'm still a massive arsehole - I think it's an intrinsic personality problem - but I do occasionally succeed in catching myself and not fucking up my relations with others yet again. (This is actually a conscious thing I actually do.)

That's the really good news: This stuff is actually susceptible to thoughtful consideration!

Excellent point. Something I've observed is that one's ability to deal with frank, no bullshit criticism is highly context dependent, and well approximated by the magnitude of the bayesian update it requires to the status of those involved.

I find myself perfectly willing to give and receive sentences like 'here's why that's bullshit', from colleagues of roughly equal social status. I've noticed that the few other people I can do that with share with me a fair amount of confidence in their own intelligence. I don't have to revise my social status relative to theirs if I correct them or they correct me.

If however a superior who's opinion of me I'm not sure of corrects me, my stomach falls out.

It's not just that other's have different values, or are less focused on truth seeking. It's also that they may be occupying a different, more vulnerable position in the game. Imagine a very attractive person tells you that you have the seductive prowess of a pug in heat. You'd take it harder than they would.

I've also noticed since I started trying to modify my interactions with people, that they've begun voicing more of their opnions to me. They used to avoid doing that out of intimidation (or so I'd been told several times).

Trying to prove yourself the smartest kid in the class is a childish goal. It's something we all have to let go of sooner or later if we want to be effective.

Random Tip:

If you intend to criticize an idea, then I agree that it is socially productive to first point out something you liked about that idea, and if you didn't like its contents at all, then go with "I like that you brought up this topic/point, because I too find it important, however / yet I think..."

The magic words in the sentence above are "however" and "yet", the latter being superior. Notice how the same sentence would sound if I replaced "yet" with "but" to link the praise/concession with my criticism. "But" can (and often enough is) perceived in contexts like this as a harsh word, and is parsed as if anything you mentioned in the sentence beforehand is completely negated by this single word. The reason why it can feel this way is because many people actually use it in exactly this way. They say something and offer concessions or even praise, and then use the word "but" to really mean "I didn't actually mean any of what I just said, so here's what I actually think on this matter..."

I believe I picked up this simple trick from "How to Win Friends and Influence People" what must have been close to 10 years ago, and every time I'm about to use a "carrot and stick" sentence I remind myself to use "however" or "yet" instead of "but" to link them. I believe the book even offered "and" (though I was reading the German translation of that book) as a potential linking word, but I can't quite warm up to that and my intuition says it might actually even weird out your conversation partner.

If you use "but" the person you talk to is likely to assume a defensive role, whereas if you use "yet" it can feel more like an invitation for a joint venture to genuinely explore the topic.

Really enjoyed the article, and thanks for the link to the nerds article. I think it is easy to underestimate how big an effect this has. When growing up my mother was always incredibly helpful with schoolwork, but because she focussed on the negative, stating mistakes directly rather than praising the good first and then carefully broaching the subject, she came across as very harsh and demanding. Despite the best possible motives her delivery made me less happy and made me more resistent to suggestions and mutinous.

Unfortunately I think I've fallen into the same trap (I wrote a comment on my sister's blog which rather upset her because I didn't follow the advice here) so thanks for flagging this and pointing me to some reading material.

It's funny, I don't remember seeing this post initially. I just followed a link from a more recent discussion post. Just yesterday I had the experience of reading a comment I posted on a popular blog and realizing that I was being a jerk in precisely this way. I only wish I could have edited it after I caught myself, but posting an apologetic followup was helpful anyway.

I learned this general principle a long, long time ago, and it has made a huge difference in the way people respond to me.

That said, to this day, I haven't been able to fully ingrain the habit. When I don't think about my presentation, it's very easy to fall into the habit of being brusque with corrections and arguments where there is no need to be combative.

The problem is that your examples already go overboard. You describe a good upper bound of how nice to be, but you can usually get away with less, and not have to constantly be constructing bullshit in your brain. This is what I think causes people to object.

For example, is this comment really going to upset you? I seriously doubt it. There would be no reason for me to write, say, "You have made a really good point, but I wonder if you perhaps went a tiny bit overboard in your examples, and thus this decreased your effectiveness."

The basic message, and the one I agree with, is to not be quite so combative when you criticize, and to make sure your criticism would cause more good than harm.

Also, you didn't need to spend so much time explaining, either. That probably also led to the idea that your methodology has a low signal-to-noise ratio. I mean, I personally skipped most of the middle, as you were just repeating yourself. I think you'd be more convincing if you learned about brevity.

Very, very good post.

I realized I had this problem, and how much it could cost me, a few months ago.

I was going to write a small column for a newspaper made by students from my faculty. The subject was a delicate one, and the positon I would argue against was strongly popular among one of the colors in the faculty. I already saw the irrationality in the old false dillemas and the groupthink that guided most color politics, and I thought I could use a chance to criticize the "movement" as well as the specific cause.

I wrote a text with a strongly ironical introduction. I showed it to my father, and he said "This is going to get you in trouble with these people in the future. You probably don't want that." My first reaction was something like this. But as I thought about it, I realized there was no need for provoking the "greens". It would accomplish nothing and would possibly gain enemies. I rewrote that first paragraph, and the tone of the text changed completely, without removing or changing a single argument. When it was published, I even received compliments from one of the "greens".

Since then, I'm careful about what I say, and more so about what I write. I think one thing aggravates the problem you described: when we write something, we can "hear" the specific tone we would use to say that, but the reader can't. Normal statements may seem ironic, and ironic statements may seem to carry much more criticism than intended.

-- Edited for markdown syntax correction.

Haha! At first glance, I thought the title read, "defecating by accident."

This post could have been custom-written for me. I have the problems you listed; I was definitely "that kid" in school,... and just five minutes ago I wrote a comment pointing out someone's unclear phrasing. I just went and edited it.

Current comment count is past 300, and now the most common criticism has become "You're condescending to us, how dare you act like we never try to be polite in our everyday lives." What I would actually say is the biggest problem is that in this form of communication, what you're communicating to the other person is completely distinct of and perhaps even the exact opposite of the exact meaning your sentence has. It becomes a "the wine before me" situation over and over. Examples that happen to me all the time:

English: "Oh, this story of yours isn't that bad. I liked the part where they were talking in the meadow while it was raining. You've been making a lot of progress."

Transmission: "I really don't like your story that much. It had one good part in it, but I want to see you get better."

English: "Uh, the twenty-first? That's a Saturday, right? I think I'm going to have a lot of chemistry homework then... probably some other stuff too. Maybe another time?"

Transmission: "Did I hear you correctly? No, I don't want to go at all, but I don't want to socially reject you completely."

The transmission many of us are receiving from you is "you're being stupid jerks right now, you have to actually try the superior method of actual politeness and you'll see immediately that it'll be way better for you". It's one I'd rather not hear, and I'm a little ashamed of that, on multiple levels.

I like this post a lot, it speaks to some of my concerns about this community and about the sorts of people I'd like to surround myself with. As an analytical/systematizing/whatever (I got a 35 on the test Roko posted a while back, interpret from that what you will), I felt very strange about all of these rhetorical games for most of my life. It was only when I discovered signalling theory in my study of economics that it started to make sense. If I frame my social interactions as signalling problems, the goals and the ways I should achieve them seem to become a lot clearer. It's also a useful outsider perspective; I find that I can (occasionally) recognize what people are really trying to do and how better than them simply because they have a richer and more complicated view of human socializing.

While I agree with most everything you've posted in the abstract, assuming your whole goal really is to have a reasoned discussion to provide evidence on which to update your beliefs, I think we often misunderstand our own intentions. Although the most common counterargument to "you should be nicer in how you frame your responses" is "I don't have time for all that fluff", there is a whole lot more going on there. First there are the personal signaling goals; not just to demonstrate that you are clever, but also that you are the sort of person who is upfront and honest (something people value) and not overly concerned with status in the eyes of your peers. This second one is amusing, since the purpose of the signal is to raise your status with the people who value not valuing status too much. I think children acting "too cool" to try hard in school is a good example of this, among other things. Additionally, there is the enforcement of group norms, not all of which have to be in perfect harmony. For example, LW has a group norm of open discussion with an eye toward gathering information about and improving human rationality. However, we also have the norm of not suffering fools gladly (what would happen if some nut posted stuff about The Secret as a comment on the quantum physics sequence?). Oftentimes these two sync up, since we don't want the discussion polluted by noisy nonsense, but sometimes they don't; I've seen more than a few people with valuable ideas leave this community because of hostile treatment.

None of this is to say that people shouldn't be nicer if they want to achieve their higher goals. That's a principle I've tried to operate on, and I think it's a good one. My point is that, just like the player who seems to play the ultimatum game irrationally by denying unfair offers, people who respond acrimoniously to their peers are often playing a game with a different goal, whether they know it or not. It may still be irrational, but its irrational in a complex and multifaceted way, and I doubt one post, or even a whole sprawling sequence of posts, would be enough to touch on all the complicated signals involved in this. The fact that the norm enforcement algorithm feels a lot more individualistic and noble from the inside than it looks from the outside makes the whole problem a lot worse.

Something I have trouble remembering:

To someone for whom it is normal to choose words carefully and connote respect, it's obvious that this is the right way to go about things--it gets other people on your side, so you don't have to fight as much to get what you want or convince people of something. It's also more pleasant to be around, and is the way you wish to be treated.

To someone for whom it is normal to be as direct, clear, and efficient in language as possible, it's obvious that this is the right way to go about things--it's much more honest than insincere fluff and doesn't waste everybody's time. It's also more pleasant to be around, and is the way you wish to be treated.

The bit I particularly have trouble remembering is that neither person would believe that if their method didn't work for them. If you practice one of the above policies, I can reasonably assume that you get results from it which are acceptable to you. There appears to be a fairly strong division between people who have that experience with the former strategy, and those who have that experience with the latter strategy. Seeing that, I find myself wondering which of these scenarios is most accurate:

• Does one strategy work better for some people, and the other work better for other people? Each is certainly easier for different people, but I'm asking about results.
• Is the difference instead in the situation--so people who often find themselves in situations where one strategy is better prefer that strategy?
• Are these just two different tools which are both often valuable?
• Or of course, is it really just that one of these is a better strategy, and a lot of people are just so stuck to the other one that they won't accept it?

The reason I care about which of these scenarios is most accurate is that they change what kind of conversation about them is appropriate. If different strategies work for different people, then trying to convince someone else to use yours is unproductive other-optimizing. If they apply to different situations, it might be interesting to examine the differences in experience that lead to each preference. If they're both good tools, that suggests it could be worthwhile for each of us to work on the one we're worse at. And, finally, if one of them is actually better, we can carry on trying to convince each other of which one that is.

My intuition is that defaulting to courtesy, if it is not uniformly the more useful strategy, is at least more useful in the vast majority of cases. However, I do not trust my intuition on this, because that is the strategy which is much easier and more enjoyable for me to use. When I try to think of real reasons to hold that belief, other than "it's worked for me," two come to mind: one is that the courteous strategy seems like it requires more skill/effort (even people who don't prefer it seem to think so), and I don't see why that would have developed if it weren't valuable; the other is that I've read or heard several people (including the OP) say that they used to prefer the blunt strategy, but have learned to use and now see the value in the other one; I have never seen anyone describe the opposite experience.

If I'm wrong, I would like to be convinced that I'm wrong, and I feel strongly enough that I'm right that I don't think I can do that on my own. Here are some ways you could convince me:

• Describe a common situation where there is clearly more utility in stating x bluntly than stating x politely. Note that I don't consider the extra few seconds to say a few more words to be significantly detrimental to utility. If you can come up with one which is common for you but not for me, this will lend weight to the second scenario; if you want to convince me that the blunt strategy is really generally better, I would want an example that's common even when you're not, say, working in a technical field.
• Are you someone who is naturally inclined towards being more consciously respectful/courteous, but has switched to the other strategy because you found it more useful? What experiences led to this switch and what's different now that you've switched?

This is the kind of thing I'd like to see more of on here; explaining how to use e.g. language to get along with people who are socially unlike you is hard and being able to do it is important. I laughed at this:

The most common criticism seems to be that adding fluff is a waste of time, insincere, and reduces signal:noise ratio.

... because as I was reading the post, I was hearing in my head every time I've heard that excuse from a friend. "It's just fluff, it shouldn't matter, it doesn't mean anything if you just say it automatically ..."

It does mean something. It means that you respect the person you're talking to enough to follow these social mores and support them rather than just sticking pins in their ego. Speaking directly and simply isn't neutral--it's actively disrespectful, with the results you describe.

This reminds me of another post Yudkowsky made on a very similar topic.

I would consider most of the people here to be informally operating on at least a milder version of Crocker's rules.

I do understand the idea that criticism is best given in a way that considers social implications rather than just the practical ones, but... what if you have nothing to praise? I do some critique (informal, granted, and of nonpublished writing) and I have noticed that, surprise, hammering people with everything that's wrong with their work is not the best way to get them to fix things. Even when I know I'm right and have more experience than they do. But oftentimes the work is of such a quality that there really isn't anything to compliment. So, are you suggesting that it's best to /make up/ a compliment or strain yourself to find something to praise rather than risk offending the writer? (Or speaker, or whichever.)

On a related note, is it necessary to use such indirect means of communicating a change you'd like to see made? Indirect meaning "/Maybe/ some examples from your experience as an accountant?" (note the question mark). Is that more or less effective than directly stating, "I'd like to see some examples from your experience as an accountant."?

I don't think so. I'd say that sometimes--oftentimes--it's best to be direct. That encourages understanding, and with indirect requests or criticism you run the risk of the recipient having a sense that your criticism is somehow optional, that since you're not stating it forcefully they must have a choice whether or not to make the change. This is circumventive and, in my opinion, gets nothing done.

Interesting article, and solidly written, but I do have to say that I'm not convinced you're completely right.

--Rachel Q.

The golden rule prevents me from padding my criticism with a layer of poetic fluff. I (and I assume others) readily see through obligatory rhetorical nonsense. I find it irritating and insulting, because it implies that you think you're tricking me into thinking that I'm not receiving criticism, and that I would have irrationally overreacted to your criticism if you hadn't cunningly veiled it.

If we are working together, I want your honesty, and I do not want us to fail together because I've done something wrong and you were too "polite" to express your genuine disapproval. If I'm dealing with a child, or someone who is very emotional and not very smart, perhaps, I should adopt your patronizing scheme. But I will not treat my technical coworkers like children, and I expect the same of them.

"Hey Sebastian, I wanted to give you a heads up. I saw your recent post, but you spelled "wisen" as "wizen" - easy spelling error to make, since they're uncommonly used words, but I thought you should know. "Wizen" means for things to dry up and lose water. Cheers and best wishes."

"Hey, thanks... I don't worry about spelling too much, but yeah that one's embarrassing, I'll fix it. Much appreciated. Anyways, what are you working on? How can I help?"

I find this insanely inefficient. The proper protocol is as follows (communicated privately, so as not to take other people's time; the emotional reception on both sides is neutral):

"Typo: "wizen""

"Thanks"

My speculation: people in "our" personspace cluster tend to be pattern mismatchers/polarity responders (NLP lingo, there are probably some googleable descriptions). Whereas "normals" get good emotions from rapport, "we" are the opposite. A lot of nerd awkwardness probably comes from the failure to understand and utilize rapport.

About your edit:

Don't just guess here. Try it out for a month. I think you'll be amazed at how differently people react to you

This is not a valid argument, and using it puts you in the same reference class with many nasty quack cures.

Great post. I think this form of self-sabotage is one that many analytical people don't realize they are engaging in. As a computer programmer and a mathematician, I definitely fall into the category of analytical people.

One way I've managed to reduce this problem in my life comes from public speaking. When we give other speakers feedback, we always commend something they've done, recommend how they can improve, and finish with another commendation.

This kind of feedback is much more effective than just praise alone, which can be rejected especially in certain cultures as being unrealistic or insincere. The feedback is also more effective than pure criticism, which, as you pointed out in your post, can make the speaker hostile and therefore less likely to listen to you.

This post helped coalesce a number of observations I had made in the past, so I would like to leave aside the debate over whether the examples of politeness are optimal and look at a couple of other points.

One point which I haven't seen much of in comments is the relationship between how well people know each other and how polite they need to be. If people know you well, then they know enough to give you the benefit of the doubt if a comment can be taken multiple ways. If, however, they have only just met you or interact with you mostly in formal settings, that extra bit of politeness can go a very long way.

• A little politeness is particularly effective when dealing with people who are being paid to do something for you, such as waiters or salespeople, or with people in bureaucracies from whom you need something. While it is not strictly necessary to be polite in these cases, it will often get you better service.

• Unless you have a particularly close-knit workplace, that is also an arena where a little extra politeness is a good idea. You certainly don't want to offend your boss, and your coworkers will likely react better to constructive criticism than straight criticism.

• The last, and most overlooked, of the areas where social niceties are important is the internet. Here, not only are you often talking to people who don't know you well, but they are also deprived of tone of voice and facial expressions, which provide important clues about intent. A comment which would be perfectly polite with a pleasant smile can seem downright rude in plain text.

Another point is that Less Wrong is not a place without social codes, it simply has it's own. Whether these codes are more rational than standard social mores is irrelevant when it comes to dealing with people who don't know them. Social codes are another type of language, and the most important factor in communication with language is speaking a common tongue. If the lojbanists succeeded in creating a perfect, logical, utilitarian language, the fact remains that it would not be particularly useful in ordering a milkshake at McDonald's. When someone butchers the English language, even if I did understand what they meant to say, I am bothered, and it affects my impression of their statement. This holds true for politeness as well.

This is really good advice for the workplace and how I would write criticisms for people who I didn't know but wanted the help of. But it is a really terrible suggestion as a norm for Less Wrong.

Here, we've all more or less agreed to not take arguments personally and reward people for admitting they are wrong. Part of what is special about this place is that while it is good to be nice I can focus on whether comments are right instead of whether or not I am threatening a poster's status. As much as possible we try to avoid status maneuvers here - so following your suggestion that we undermine the community's signal to noise ratio in order to make allies (in a way other than being right) is a rather straightforward defection.

This doesn't make being mean is acceptable. I agree with Alicorn's classic post. But you seem to be advocating not just taking steps to avoid being mean but to expend extra efforts and page space on meaningless niceties instead of making forthright and respectful comments. While perhaps too confrontational in most workplaces nearly all of the examples are just fine here. If you're going to bother correcting spelling at all (and only in certain cases is it worth it) be brief for goodness sake!

I hereby pre-commit to downvoting anyone who corrects one of my many spelling errors by writing a paragraph like

Hey Jack, I wanted to give you a heads up. I saw your recent post, but you spelled "wisen" as "wizen" - easy spelling error to make, since they're uncommonly used words, but I thought you should know. "Wizen" means for things to dry up and lose water. Cheers and best wishes."

...

Wisen, not wizen.

will suffice. If you can't leave it like that "Wisen, not wizen, you idiot." does less psychic damage than making me read that paragraph. The funny joke that was made in the original example would be the best, though.

I've struggled with this for years and have done a lot of the things you've described, both the bad behaviors and the good suggestions. I still have trouble with it, especially over the Internet.

I'd like to point out that, at least in my experience, it takes a lot of practice to avoid offending others without offending your sense of self. I've also found it much easier in person, where I can use tone and a natural lack of seriousness to paper over any rough edges.

I generally like this essay on this topic: http://pdf23ds.net/implications-and-debate/

If you read the comments, however, please note that the original essay contained a lot of language that was pretty aggressive and insulting to feminists and sex/gender writers. Some writers (including myself) called out that language in the comments. The essay was then edited multiple times, but no notes were left that it had been edited. This was a great way to make commenters who had complained about the original essay (such as myself) look like crazy bitches, which doesn't seem like a very charitable debating tactic to me. ;)

Otherwise, though, yeah, it's a good essay.

I am neurotypical in the sense that I'm not on the Asperger's/autism spectrum at all. And I think, because I'm a woman, I've internalized a fair bit of social finesse/politeness/ways to signal deference and avoid ego challenges. Given that overestimating our own competence is a common bias, I'm hesitant to say "my social skills are good" or anything like that, but I can at least report that I don't generally provoke hostility where I'm not anticipating it.

I agree that niceness is important, for all the reasons Alicorn has laid out. But I also agree with Alicorn and other commentators that the examples you give are off-putting. To me, they do not actually read as nice. They read as smarmy and condescending. At the same time, your advice: "Don't just guess here. Try it out for a month. I think you'll be amazed at how differently people react to you" is off-putting for its condescension. You are dismissing all your critics as not knowing what they're talking about ("just guessing"), and implying that people react poorly to them now--or at least much more poorly than they react to you. In this way you're implicitly claiming superior social status, which is ego-challenging behavior, and likely to provoke a hostile reaction.

So, like Alicorn etc, I agree with your premise but disagree with your suggestions for successful execution of the goal. Also, as others have pointed out, it's important to tailor your style for the audience. The three examples of criticism which you brought up as overly blunt strike me as perfectly tailored to the audience--the first uses friendly humor, which is actually a great method of softening criticism, and the second two are direct and succinct without being hostile.

Consider dropping it altogether if it's not a big deal. This about learning to prioritize - I had someone comment on my site thinking mistakenly that The Richest Man in Babylon and The Greatest Salesman in the World were by the same author. It wasn't, but who cares? It makes no difference. It's not worth pointing it out - almost everyone has an aversion to being corrected, so only do it if there's actually tangible gain. Otherwise, go do something more important and not engender the potential bad will.

I consider this particularly selfish and evil. If you know there's a place where someone's map doesn't correspond to the territory, you should tell them before they inadvertently drive off a cliff. Even if it would be a status hit to yourself. And you can't tell me that's a part of their map they weren't even using, as they had just used it!

Why call this "defection"? I interpret "defection" as meaning not just "a bad thing people do" but as deliberately deviating from a previous agreement. The relationship between the prisoner's dilemma and not being sufficiently polite seems forced, or at least like it could have used more thorough explanation.

I agree with Alicorn and others who find the sort of forced extreme politeness of some of the suggested responses (especially to #1) off-putting. I can't quite explain why, but if I had to guess, it would be for two reasons. First, politeness level indicates status, and when someone uses excessive politeness that ascribes to me extremely high status that I don't feel I've earned, I suspect they're trying to manipulate me. Given that many of the arguments in this post are explicitly about politeness as a tool for manipulating people, this seems to be a valid suspicion.

Second, lack of politeness is a countersignalling method to indicate friendship and community by showing you are close enough to a person that politeness is unnecessary (consider the relatively common story of the friends who greet each other with racial slurs, like "Hey n*gga!", "Hey cracker!" partly as a bonding mechanism to show that they're close enough to allow what would normally be offensive). If everyone is on board with this, someone who goes around being extremely polite and telling others to be extremely polite might seem to be distancing eirself from the community and denying friendship.

I realize you don't intend polite comments that way, but as you say in this post, it doesn't matter what you intend so much as how other people interpret it. I'd agree that politeness is generally a good thing, but carpet-bombing all social interactions with politeness will tend to freak a lot of people out, and Less Wrong is probably more prone to that than other places.

Oddly enough, I actually became very popular for some time in school, largely by accident and without developing my social fluency in the process. It was intensely unpleasant having so much social attention, even though it was mostly positive, without having the faculties to deal with it effectively. I think this may have delayed my process of developing my social skills considerably, because I became convinced that being popular was not a desirable state of affairs, and lost interest in pursuing the approval of others.

Can anyone recommend an online discussion group which displays neurotypical communication at its best? It seems to me that the politeness examples in this discussion are a little hypothetical.

This is a great point you're making. At the risk of being perceived as antagonistic, I've gotta say I don't think the prisoner's dilemma is the best frame for it. It's not that people are trying to cooperate and failing, it's that they just haven't developed the skills they need to accomplish the ends they seek. It's rather like not working on your short game in golf because you're really good at the long game. I think it's a good effort to make the issue topical, but the issue is already so topical that I think it distracts from rather than enhances your point.

I have a theory about "dumb blonde syndrome", the idea that beautiful women are dumb. Folk psychology says that everybody gets the same number of character points to distribute among their attributes, so some people get intelligence while others get beauty. But reality says that beauty is correlated with <edit>health</edit>, which is correlated with intelligence. Beautiful people should tend to be smart. I think there is some positive beauty/intelligence correlation.

But I remember taking a class with a stunningly beautiful woman, who every week would loudly make some inane comment or question, and not realize it was inane because no one would tell her so. And I developed the theory that beautiful people don't learn to self-censor, because they don't need to. Anybody else would get ridiculed when they said something stupid, and learn to be more shy.

Maybe this also applies to smart people. They're more likely to be correct, and so less likely to be made fun of when speaking their mind, and so less needful of learning how to phrase a question in a way that reduces their chances of being made fun of.

I agree with your point where it concerns interactions in most social forums, but strongly disagree with your implied suggestion of a social norm for Less Wrong. I prefer a community that produces the given examples to one that produces your suggested corrections.

The karma system somewhat mitigates the "silent approval, vocal dissent" phenomenon here, though I agree with NihilCredo that there still exist confusing side effects, and that a more nuanced system would be superior. But as it stands, the community norm for an author of a post should be that criticism is meant at face value without further implied disapproval, and that the post karma is the better clue to readers' overall impressions.

Good advice for the real world, maybe. But consider that here on LW, we are among analytical people, and wouldn't have it otherwise.

You know, if you wanted to look at conversation between two people as a prisoner's dilemma, I think it's quite the other way around, isn't it?

Consider the "payoff" of this situation to be the amount which the person you're talking to has considered and valued your concerns, suggestions, ideas, or requests.

C = Say what you mean as clearly and accurately as possible, and interpret the other person's actions as such

D = Choose your words and actions to be polite and non-confrontational

(You C, I C) = We both understand each other and we both wind up communicating to the best of our ability.

(You C, I D) = I am irritated with you for being rude to me and don't care for what you said very much. You probably heard me out, although you think I'm too sensitive and beat around the bush.

(You D, I D) = We both say what we have to say, but it might take longer and there might be minor misunderstandings that neither of us wanted to step on.

We'd all be better off, I think, if we could learn to C, and give and receive criticism in an accurate way, in good faith, without playing status and politeness games about it. But since the majority of the world is D, and probably won't change, we might need to D as well if we ever want anyone to put up with us.

So if you buy this interpretation, we're not defecting by accident -- we're cooperating by default, even when we're playing against Defection Rock and getting poor outcomes as a result.

If I say an idea in public that doesn't work for some reason such as lack of resources, I'd rather someone bluntly tell me. The issue in many of these examples doesn't seem to be so much "defecting" as much as some people preferring their own status over honest discussion. The long-term solution is to change the culture, and in the meantime, just make sure not to other-optimize when trying to talk to people.