I had a pretty negative initial reaction to this comment, and I am currently trying to decipher why. Here are my thoughts:
I felt some sense that you were personally threatening Connor, or trying to get him to admit a mistake. The first sentence was:
This is why you should produce offline, then upload it online.
To me this was spoken in a condescending tone, with somewhat of an “I told you so” intonation. The beginning of the next sentence read similarly to me:
I don’t know why you haven’t been doing it
Which to me was spoken in an ironic voice, meaning more something like “I don’t know how you missed such an obvious thing”.
And I think after those two sentences I had already settled on that being the tone of the comment, and that tone just felt really aggressive to me, so I downvoted it.
If that would have been the tone in which you intended the comment to be read, then I think the comment should have been justifiably downvoted. Reading your response I have a sense that you did not intend the comment to be read in the tone that I ended up reading it.
I am not really sure what the best way to deal with this is. Subtext and tone is much harder to communicate in text-only communication, and so I think authors will have to inevitably put in a bunch of effort to ensure that text does not get read in the wrong tone. The reaction of me reading the comment in this more aggressive tone happened very quickly at a highly subconscious level, and while I will try to calibrate myself here better, I am not sure how tractable an intervention on the level of “let’s just have everyone try to be more charitable when reading comments” actually is, or whether that makes sense.
For now, I think it is reasonable that this comment got downvoted, because I expect the majority of readers to read it in an aggressive tone, but also think that there is a very closeby comment that does not require significant rewriting that would get read in the correct tone.
I agree with both Elizabeth and Said (just saw all of this thread in the last ten minutes).
The advice was something like 85% obvious and stuff I had already belatedly realized, and so getting it in long form was a little bit like being lectured or being kicked while I was down. There are two status signals in a comment like that (the first being “I, DragonGod, know better” and the second being “you, Conor, are not clever enough to have already figured this out, nor will you in the future unless I tell you”) and for that reason my monkey brain wanted to downvote reflexively (I didn’t, in the end).
But a) it’s obvious from you asking this question that those status moves were not at all on your radar/were not intended, and b) I completely agree that it’s good to have those thoughts and heuristics up where others can see them and benefit from them. Your comment included models and details that make it concretely useful, and it did, in fact, include at least one point that I hadn’t considered.
Interestingly, as I tried to answer your question, I found myself skirting the edge of the exact same surface-level problem that I think brought on the downvotes. So as I went to try to tell you what I think was good and bad about your comment, I found myself wanting to couch with phrases like “you probably already thought of this, but for the sake of anyone who doesn’t” and “I could be wrong here, let me know if your models differ” and “I hope this doesn’t come across as lecturing/condescending; I just figured it was better to err on the side of caution and completeness.”
If I were in your position, that’s probably the update I’d take away from this (that the skin of the comment matters a lot, and there’s a large swath of people who don’t want to be treated like white belts). I’d probably still have the same “speak up and share advice” module, but I’d add a function to the front that injects some gentleness and some status-dynamic-defusing words and phrases.
> I’d probably still have the same “speak up and share advice” module, but I’d add a function to the front that injects some gentleness and some status-dynamic-defusing words and phrases.
In this case, I have to object to this advice. You can tie yourself in knots trying to figure out what the most gentle way to say something is, and end up being perceived as condescending etc. anyway for believing that X is “obvious” or that someone else “should have already thought of it” (as again, what is obvious to one person may not be obvious or salient to another). Better to just state the obvious.
First, I definitely agree that getting tied in knots is both a) bad and b) something at least some people are vulnerable to.
But 80⁄20 seems like a valuable principle to adhere to, especially if you’ve already been downvoted/punished for your normal style and then, on your own initiative, you asked for clarification/updated heuristics. The claim I made about what I would do in DragonGod’s shoes wasn’t an imperative for literally everybody, it was simply … well, what I would do in their shoes.
I think I would be with you re: objecting if the line you quoted were broadcast as general advice that everyone needs more of. But in context, my brain is trying to round off your objection to “you can’t please everyone, so don’t even try.” I grant that I may have misunderstood you, but in fact signaling and status moves are real, and in fact (as evidenced by group opinion) DragonGod’s first comment was readily parseable as containing information that DragonGod didn’t intend to put there, and in fact a good patch for preventing that from happening in the future is adding something like 5-25 words of gentling and caveatting. That wasn’t a call to tie oneself in knots, nor a claim that this would please everybody.
“Add 5-25 extra words,” for someone who’s already gotten data that they’re not at the ideal point and was intrinsically motivated to investigate further, does not seem to me like a dangerous heuristic that’s likely to suck up lots of attention or effort.
And as for “stating the obvious,” well—I’d wager $50 at 5-1 odds that if DragonGod had prefaced the comment with “I may be stating the obvious here, but,” then the downvotes would not have happened (because that does away with the implicit claim that I or other readers don’t know that these things are low-hanging fruit and easily thinkable).
For what it’s worth, I’m strongly of the same opinion as Conor on this one. It’s just too easy to get annoyed by criticisms, and I think increasing the risk of tying oneself in knots is worth it for the sake of decreasing the risk that someone gets offended.
I thought it was an excellent comment, and upvoted it!
Edit: In particular (though not only)—I found the analogy to software development insightful, and new to me (has this comparison been made elsewhere? if so, I’ve not seen it); and I also didn’t know about the snowflake method, so that’s a new thing I’m reading about now, thanks!
And I definitely agree with the core point about writing offline, and all the reasons you list for why it’s a good idea.
Not everything is obvious to everyone; and what is obvious, may grow less obvious with time, as new generations, and new waves of people who come into a community, come of age without ever hearing it said aloud (since it’s too obvious to say!). This is why even obvious things need to be said, periodically, lest another piece of collective knowledge be lost.
Then, too, things that are obvious when you think about them might not be salient, in the moment of doing; saying the obvious thing increases its salience, and thus benefits everyone.
Finally, a comment on a forum/blog post is not a one-to-one communication between OP and commenter; it’s a public speech act. The advice is condescending, useless to the OP? Perhaps, perhaps not. But that has little bearing on whether it is useful to others who read it!
It will, I think, lead to far more lastingly useful content on this site, if we treat comments not just, and not even primarily, as utterances in a conversation, but also and importantly, as public utterances, spoken before the whole forum, and addressed to the collective. For who might care about the details of some conversation that took place long ago? But what was said to a public audience—that will keep.
Why did this comment get downvoted? I’m willing to update f I did something wrong, but I think I was being genuinely helpful.
I had a pretty negative initial reaction to this comment, and I am currently trying to decipher why. Here are my thoughts:
I felt some sense that you were personally threatening Connor, or trying to get him to admit a mistake. The first sentence was:
To me this was spoken in a condescending tone, with somewhat of an “I told you so” intonation. The beginning of the next sentence read similarly to me:
Which to me was spoken in an ironic voice, meaning more something like “I don’t know how you missed such an obvious thing”.
And I think after those two sentences I had already settled on that being the tone of the comment, and that tone just felt really aggressive to me, so I downvoted it.
If that would have been the tone in which you intended the comment to be read, then I think the comment should have been justifiably downvoted. Reading your response I have a sense that you did not intend the comment to be read in the tone that I ended up reading it.
I am not really sure what the best way to deal with this is. Subtext and tone is much harder to communicate in text-only communication, and so I think authors will have to inevitably put in a bunch of effort to ensure that text does not get read in the wrong tone. The reaction of me reading the comment in this more aggressive tone happened very quickly at a highly subconscious level, and while I will try to calibrate myself here better, I am not sure how tractable an intervention on the level of “let’s just have everyone try to be more charitable when reading comments” actually is, or whether that makes sense.
For now, I think it is reasonable that this comment got downvoted, because I expect the majority of readers to read it in an aggressive tone, but also think that there is a very closeby comment that does not require significant rewriting that would get read in the correct tone.
I agree with both Elizabeth and Said (just saw all of this thread in the last ten minutes).
The advice was something like 85% obvious and stuff I had already belatedly realized, and so getting it in long form was a little bit like being lectured or being kicked while I was down. There are two status signals in a comment like that (the first being “I, DragonGod, know better” and the second being “you, Conor, are not clever enough to have already figured this out, nor will you in the future unless I tell you”) and for that reason my monkey brain wanted to downvote reflexively (I didn’t, in the end).
But a) it’s obvious from you asking this question that those status moves were not at all on your radar/were not intended, and b) I completely agree that it’s good to have those thoughts and heuristics up where others can see them and benefit from them. Your comment included models and details that make it concretely useful, and it did, in fact, include at least one point that I hadn’t considered.
Interestingly, as I tried to answer your question, I found myself skirting the edge of the exact same surface-level problem that I think brought on the downvotes. So as I went to try to tell you what I think was good and bad about your comment, I found myself wanting to couch with phrases like “you probably already thought of this, but for the sake of anyone who doesn’t” and “I could be wrong here, let me know if your models differ” and “I hope this doesn’t come across as lecturing/condescending; I just figured it was better to err on the side of caution and completeness.”
If I were in your position, that’s probably the update I’d take away from this (that the skin of the comment matters a lot, and there’s a large swath of people who don’t want to be treated like white belts). I’d probably still have the same “speak up and share advice” module, but I’d add a function to the front that injects some gentleness and some status-dynamic-defusing words and phrases.
> I’d probably still have the same “speak up and share advice” module, but I’d add a function to the front that injects some gentleness and some status-dynamic-defusing words and phrases.
In this case, I have to object to this advice. You can tie yourself in knots trying to figure out what the most gentle way to say something is, and end up being perceived as condescending etc. anyway for believing that X is “obvious” or that someone else “should have already thought of it” (as again, what is obvious to one person may not be obvious or salient to another). Better to just state the obvious.
First, I definitely agree that getting tied in knots is both a) bad and b) something at least some people are vulnerable to.
But 80⁄20 seems like a valuable principle to adhere to, especially if you’ve already been downvoted/punished for your normal style and then, on your own initiative, you asked for clarification/updated heuristics. The claim I made about what I would do in DragonGod’s shoes wasn’t an imperative for literally everybody, it was simply … well, what I would do in their shoes.
I think I would be with you re: objecting if the line you quoted were broadcast as general advice that everyone needs more of. But in context, my brain is trying to round off your objection to “you can’t please everyone, so don’t even try.” I grant that I may have misunderstood you, but in fact signaling and status moves are real, and in fact (as evidenced by group opinion) DragonGod’s first comment was readily parseable as containing information that DragonGod didn’t intend to put there, and in fact a good patch for preventing that from happening in the future is adding something like 5-25 words of gentling and caveatting. That wasn’t a call to tie oneself in knots, nor a claim that this would please everybody.
“Add 5-25 extra words,” for someone who’s already gotten data that they’re not at the ideal point and was intrinsically motivated to investigate further, does not seem to me like a dangerous heuristic that’s likely to suck up lots of attention or effort.
And as for “stating the obvious,” well—I’d wager $50 at 5-1 odds that if DragonGod had prefaced the comment with “I may be stating the obvious here, but,” then the downvotes would not have happened (because that does away with the implicit claim that I or other readers don’t know that these things are low-hanging fruit and easily thinkable).
For what it’s worth, I’m strongly of the same opinion as Conor on this one. It’s just too easy to get annoyed by criticisms, and I think increasing the risk of tying oneself in knots is worth it for the sake of decreasing the risk that someone gets offended.
I thought it was an excellent comment, and upvoted it!
Edit: In particular (though not only)—I found the analogy to software development insightful, and new to me (has this comparison been made elsewhere? if so, I’ve not seen it); and I also didn’t know about the snowflake method, so that’s a new thing I’m reading about now, thanks!
And I definitely agree with the core point about writing offline, and all the reasons you list for why it’s a good idea.
I haven’t voted on this comment, but if I had to guess:
The author did not ask for advice. The loss was mentioned as part of context for a different decision.
It is too late to do anything about losing this particular post.
“Write offline” is an obvious solution to the problem, suggesting it comes across as condescending and victim-blaming.
Not everything is obvious to everyone; and what is obvious, may grow less obvious with time, as new generations, and new waves of people who come into a community, come of age without ever hearing it said aloud (since it’s too obvious to say!). This is why even obvious things need to be said, periodically, lest another piece of collective knowledge be lost.
Then, too, things that are obvious when you think about them might not be salient, in the moment of doing; saying the obvious thing increases its salience, and thus benefits everyone.
Finally, a comment on a forum/blog post is not a one-to-one communication between OP and commenter; it’s a public speech act. The advice is condescending, useless to the OP? Perhaps, perhaps not. But that has little bearing on whether it is useful to others who read it!
It will, I think, lead to far more lastingly useful content on this site, if we treat comments not just, and not even primarily, as utterances in a conversation, but also and importantly, as public utterances, spoken before the whole forum, and addressed to the collective. For who might care about the details of some conversation that took place long ago? But what was said to a public audience—that will keep.