This is why it’s taken me so long to write about these things on LW: because I found the prospect of dealing with ever more nitpicking and straw-manning to be exhausting. So often people don’t meet me half way and put in the work of steelmanning my arguments that would be necessary to make it cost-effective for me to engage.
Two points:
First, I think the recurring advice to state your main thesis, and then motivate it, applies. Among other reasons, it makes it easier for people to not make leaps in the wrong direction. If you show me some bizarre theorem, and then explain the pieces that make up that theorem, I can keep returning to the bizarre theorem, adjusting my concept of it with the new explanation until it clicks. If you just show me the pieces that make up the theorem, the part of me that’s trying to model your motives in the conversation has to search many possibilities for why you might be introducing any particular piece. Unless I can independently discover the theorem you want to talk about, I’m probably going to get it wrong even once I have all the pieces! While I have only a subset of the pieces, how do I have any hope?
Remember, a steelman is when one takes an argument that reaches a particular conclusion and says “alright, what is the most forceful version of this argument that I can make?”—which implies optimization holding the conclusion constant. If the conclusion is unknown, it’s not really steelmanning that’s called for, but a sort of suspension of disbelief.
Second, these sorts of writing difficulties are typically addressed by editing. Among other things, the editor can point out what readers are likely to misunderstand (possibly because they misunderstood it). For example, this post looks like you intended it to just claim “Scott Alexander, because he has great aesthetic sense, could become a good research mathematician through taking a particular approach to learning” but some people have read it as “there is not such a thing as mathematical ability that differs among people” or “if Alexander had worked harder, he would have gotten further,” neither of which it looks like you intended to me. I think those misunderstandings are a predictable outcome of the way you communicated, though, and reorganization or rewriting could prevent those misunderstandings and lead to a better reception (here and elsewhere).
This is helpful feedback. I do recognize that I have a lot of room for improvement in these regards. But making comments like
This lowers my expectation of you getting around to a sensible recommendation.
should be against community norms, not for my sake, but for the sake of the commenters – this is not a good mode of operation for overcoming bias and becoming less wrong (!!). Commenters should be inquisitive and open-minded rather than combative and dismissive.
I dislike the trend to cuddlify everything, to make approving noises no matter what, then framing criticisms as merely some avenue for potential further advances, or somesuch.
On the one hand, I do recognize that works better for the social animals that we are. On the other hand, aren’t we (mostly) adults here, do we really need our hand held constantly? It’s similar to the constant stream of “I LOVE YOU SO MUCH” in everday interactions, it’s a race to the bottom in terms of deteriorating signal/noise ratios. How are we supposed to convey actual approval, shout it from the rooftops? Until that is the new de facto standard of neutral acknowledgment?
A Fisherian runaway, in which a simple truth is disregarded: When “You did a really good job with that, it was very well said, and I thank you for your interest” is a mandatory preamble to most any feedback, it loses all informational content. A neutral element of speech. I do wish for a reset towards more sensible (= information-driven) communication. Less social-affirmation posturing.
But, given the sensitive nature of topics here, this may be the wrong avenue to effect such a reset, invoking Crocker’s Rules or no. Actually skipping the empty phraseology should be one of the later biases to overcome.
This is a thing because we have complex brains, with only a part devoted to processing information of the kind you mean, and others worried about contingent social facts: dominance/submission/status/etc.
I think the broadly right response is to make peace-via-compromise between those parts, and that involves speaking on multiple bandwidths, as it were. This, to me, is a type of instrumental rationality in interpersonal communication.
Citing phatic expressions is not really enough. The issue is what creates the signal: presence or absence of something.
If the default is “Thanks” then saying nothing is the negative signal and saying “Thank you, you did such a great job!” is a positive signal.
But if the default is “Thank you, you did such a great job!” then just “Thanks” becomes a negative signal and for a positive signal you have to escalate to “Oh my God this was the greatest thing ever I thank you so much how could I ever...”
It’s easy to see how this could get to be very inefficient and, frankly, ridiculous.
“Collegiality and tact” is not a function of how extended the social default is—it is a function of knowing that social default (or, more specifically, the expectations of the other party) and not being an asshole.
Whether the runway process gets off the ground depends on the circumstances. Some cultures have (or had) elaborate long and greeting rituals for high-status/caste/position people and skipping some part would have been a serious offense. As a more prosaic example, notice how the film credits take the pain to include the names of all second assistants to the third helper of the aide to the attendant who carried the purse of the co-star...
The default amount of “gratitude” expressed on LW seems to be considerably less than that expressed by even “thanks”. Actually, most of the time it seems that the default response is to find some flaw of wording to nitpick, and usually such a flaw is only tangentially related to the thrust of the argument. That’s not what we should be encouraging.
Congratulatory comments, even of the empty sort like “Great job!”, serve as positive Pavlovian reinforcement, which helps to motivate/encourage people to post. In addition, they signal appreciation and gratefulness at the fact that someone was willing to make a top-level post in the first place. The fact that the people on LessWrong are at times so damn unfriendly is in my opinion a non-trivial part of the cause of LW’s too often insular atmosphere.
Furthermore, studies consistently show that humans respond better to positive reinforcement than to negative reinforcement, regardless of age. This isn’t about whether we’re “adults who don’t need our hands held”. It’s about how to motivate people to post more. If Jonah gets a torrent of criticisms every time he posts something, that’s going to create an ugh field around the idea of posting. If he then points this out in a comment, and people respond by saying what effectively amounts to “Well, it’s your own fault for not being clear enough,” well, you can imagine how it might feel. This is an issue entirely separate from that of whether the criticisms are right.
The bottom line is that transmission of useful information isn’t the only kind of transmission that occurs in human communication. “This post is so messy and obfuscated as to be nearly unreadable” and “I think your point may benefit from some clarification” are denotationally similar, but connotationally they are very different. If you insist on ignoring this distinction or dismissing it as unimportant (as it seems so many LWers are wont to do), you run the risk of generating an unpleasant social atmosphere.
Seriously. This isn’t rocket science. (See what I did there?)
I don’t want approval, I want to help people. If people think that they can offer helpful feedback (as Vaniver did), they should do so. Empty praise is just as useless as empty criticism. Vaniver’s feedback had substantive information value – that’s why I’m glad that he made his comment. If I fail to help people because I’m not receptive enough to critical feedback, it’s my own fault. I accept responsibility for the consequences of my actions.
Two points:
First, I think the recurring advice to state your main thesis, and then motivate it, applies. Among other reasons, it makes it easier for people to not make leaps in the wrong direction. If you show me some bizarre theorem, and then explain the pieces that make up that theorem, I can keep returning to the bizarre theorem, adjusting my concept of it with the new explanation until it clicks. If you just show me the pieces that make up the theorem, the part of me that’s trying to model your motives in the conversation has to search many possibilities for why you might be introducing any particular piece. Unless I can independently discover the theorem you want to talk about, I’m probably going to get it wrong even once I have all the pieces! While I have only a subset of the pieces, how do I have any hope?
Remember, a steelman is when one takes an argument that reaches a particular conclusion and says “alright, what is the most forceful version of this argument that I can make?”—which implies optimization holding the conclusion constant. If the conclusion is unknown, it’s not really steelmanning that’s called for, but a sort of suspension of disbelief.
Second, these sorts of writing difficulties are typically addressed by editing. Among other things, the editor can point out what readers are likely to misunderstand (possibly because they misunderstood it). For example, this post looks like you intended it to just claim “Scott Alexander, because he has great aesthetic sense, could become a good research mathematician through taking a particular approach to learning” but some people have read it as “there is not such a thing as mathematical ability that differs among people” or “if Alexander had worked harder, he would have gotten further,” neither of which it looks like you intended to me. I think those misunderstandings are a predictable outcome of the way you communicated, though, and reorganization or rewriting could prevent those misunderstandings and lead to a better reception (here and elsewhere).
This is helpful feedback. I do recognize that I have a lot of room for improvement in these regards. But making comments like
should be against community norms, not for my sake, but for the sake of the commenters – this is not a good mode of operation for overcoming bias and becoming less wrong (!!). Commenters should be inquisitive and open-minded rather than combative and dismissive.
I dislike the trend to cuddlify everything, to make approving noises no matter what, then framing criticisms as merely some avenue for potential further advances, or somesuch.
On the one hand, I do recognize that works better for the social animals that we are. On the other hand, aren’t we (mostly) adults here, do we really need our hand held constantly? It’s similar to the constant stream of “I LOVE YOU SO MUCH” in everday interactions, it’s a race to the bottom in terms of deteriorating signal/noise ratios. How are we supposed to convey actual approval, shout it from the rooftops? Until that is the new de facto standard of neutral acknowledgment?
A Fisherian runaway, in which a simple truth is disregarded: When “You did a really good job with that, it was very well said, and I thank you for your interest” is a mandatory preamble to most any feedback, it loses all informational content. A neutral element of speech. I do wish for a reset towards more sensible (= information-driven) communication. Less social-affirmation posturing.
But, given the sensitive nature of topics here, this may be the wrong avenue to effect such a reset, invoking Crocker’s Rules or no. Actually skipping the empty phraseology should be one of the later biases to overcome.
http://en.wikipedia.org/wiki/Phatic_expression
This is a thing because we have complex brains, with only a part devoted to processing information of the kind you mean, and others worried about contingent social facts: dominance/submission/status/etc.
I think the broadly right response is to make peace-via-compromise between those parts, and that involves speaking on multiple bandwidths, as it were. This, to me, is a type of instrumental rationality in interpersonal communication.
Citing phatic expressions is not really enough. The issue is what creates the signal: presence or absence of something.
If the default is “Thanks” then saying nothing is the negative signal and saying “Thank you, you did such a great job!” is a positive signal.
But if the default is “Thank you, you did such a great job!” then just “Thanks” becomes a negative signal and for a positive signal you have to escalate to “Oh my God this was the greatest thing ever I thank you so much how could I ever...”
It’s easy to see how this could get to be very inefficient and, frankly, ridiculous.
But in reality this runaway process doesn’t get off the ground, and peters out at something called “collegiality and tact.”
“Collegiality and tact” is not a function of how extended the social default is—it is a function of knowing that social default (or, more specifically, the expectations of the other party) and not being an asshole.
Whether the runway process gets off the ground depends on the circumstances. Some cultures have (or had) elaborate long and greeting rituals for high-status/caste/position people and skipping some part would have been a serious offense. As a more prosaic example, notice how the film credits take the pain to include the names of all second assistants to the third helper of the aide to the attendant who carried the purse of the co-star...
The default amount of “gratitude” expressed on LW seems to be considerably less than that expressed by even “thanks”. Actually, most of the time it seems that the default response is to find some flaw of wording to nitpick, and usually such a flaw is only tangentially related to the thrust of the argument. That’s not what we should be encouraging.
Congratulatory comments, even of the empty sort like “Great job!”, serve as positive Pavlovian reinforcement, which helps to motivate/encourage people to post. In addition, they signal appreciation and gratefulness at the fact that someone was willing to make a top-level post in the first place. The fact that the people on LessWrong are at times so damn unfriendly is in my opinion a non-trivial part of the cause of LW’s too often insular atmosphere.
Furthermore, studies consistently show that humans respond better to positive reinforcement than to negative reinforcement, regardless of age. This isn’t about whether we’re “adults who don’t need our hands held”. It’s about how to motivate people to post more. If Jonah gets a torrent of criticisms every time he posts something, that’s going to create an ugh field around the idea of posting. If he then points this out in a comment, and people respond by saying what effectively amounts to “Well, it’s your own fault for not being clear enough,” well, you can imagine how it might feel. This is an issue entirely separate from that of whether the criticisms are right.
The bottom line is that transmission of useful information isn’t the only kind of transmission that occurs in human communication. “This post is so messy and obfuscated as to be nearly unreadable” and “I think your point may benefit from some clarification” are denotationally similar, but connotationally they are very different. If you insist on ignoring this distinction or dismissing it as unimportant (as it seems so many LWers are wont to do), you run the risk of generating an unpleasant social atmosphere.
Seriously. This isn’t rocket science. (See what I did there?)
I don’t want approval, I want to help people. If people think that they can offer helpful feedback (as Vaniver did), they should do so. Empty praise is just as useless as empty criticism. Vaniver’s feedback had substantive information value – that’s why I’m glad that he made his comment. If I fail to help people because I’m not receptive enough to critical feedback, it’s my own fault. I accept responsibility for the consequences of my actions.