On a related topic, see mycomments on whether status differences serve useful community functions. My current guess is that status differences are counterproductive on net for achieving community goals, but I’d be interested to read counterpoint if anyone’s got any (especially you, Mr. High Status Person).
Ideally status could be replaced by domain specific estimates of competence, reliability, trustworthiness etc. But in practice nobody has the time. We have to summarize.
For humans, social status is much more than just an aggregate estimate of competence/reliability/trustworthiness. It motivates us, distorts our thinking, plays a key role in our politics, etc. To take just one example, I suspect that the main reason it’s so hard for most people to change their mind is because they don’t know how to do it in a way that preserves their status. For many people and social groups, admitting you’re wrong means losing face, and most people don’t like to lose face, so they resist publicly changing their mind.
(This is another reason why status differences may be counterproductive for rational communities… they could create an incentive for high-status people to not change their mind about things, since they have something to lose. The evidence may very well justify thinking one thing one week, then something else the next week, then something else the third week. But if you’re changing your mind about critical issues every week, it won’t be long before typical humans take you less seriously. Which is unfortunate.)
Also, this doesn’t sound like your true objection to me. It doesn’t take very many more bits of information to transfer 3 estimates on each of competence, reliability, and trustworthiness than a single aggregate number. And people communicate specific info all the time (“how good is X at Y? do you trust Z?”). It’s not obvious to me that a single aggregate quantity is frequently useful. Let’s say I introduce a friend to you and say his status is 67⁄100; was that useful information? (And in practice, peoples’ status is often determined by relatively silly things like how many friends they have, what status they’re perceived to have, how confident they act, and how confidently they talk. Another reason status sucks: it gives people an incentive to make confident predictions; see Philip Tetlock’s work on how confident experts are more likely to be wrong and more likely to be quoted in the media.)
(I don’t think I’ve got a clear idea of how best to make use of humans’ status wiring; I’m just kind of exploring different ideas at this point. But it seems like an important and neglected topic.)
Eliminating status differences has been tried and failed. If a hiring manager ever tells you “There are no office politics here”, then don’t take the job. There WILL be politics, except that it will be taboo to publicly admit it—and nobody will help you if you have a problem.
“X has been tried and faied” remains true until someone succeeds. If a thing with so many advantages has been tried and failed, then the solution is not to give up and make an equivalent utterance to “man was not meant to fly”; it is to examine why it failed, explore what the underlying rules and mechanics might be, construct a strategy based on those underlying rules and mechanics, and then try again.
Let me rephrase, then: declaring that you’ve eliminated status differences, when, in reality, you haven’t, is a relatively common mistake that tends to cause problems.
declaring that you’ve eliminated status differences, when, in reality, you haven’t, is a relatively common mistake that tends to cause problems.
Aha, much more understandable. Thank you.
In that case: what would you surmise from a hiring manager that said “there’s office politics everywhere, of course, but we try to take an active role in minimizing their impact, and part of you being a good fit here will depend on your ability to help us with that goal.”?
(I regretfully confess that my own reaction to that statement would depend on that hiring manager’s gender, and (if male) how tall he was and how deep his voice was).
Perhaps a good way to deal with the situation in that XKCD comic would be to try to pick a culture that seemed particularly effective and then copy all of its norms, attitudes, etc.? So you’d have something that was battle-tested, if you will.
Well, Valve’s profitability per employee is supposedly higher than Google or Apple’s, and their employee handbook detailing their unconventional corporate culture is available for viewing online. shrug
(For what it’s worth from what I can tell Mormons don’t even formally make the sort of ontological commitments that are typical of (at-least-somewhat-reflective) mainstream Christianity (like, ‘Jesus is my savior and I should have expected Him to show up in all logically possible worlds and all possible minds should be rounded-up-to-infinitely compelled by His story and the seemingly contingent features of Jesus [Jesus’s teachings] are actually universal features of Logos and so it would be an obvious epistemic sin to disregard Him [them]’) and so it’s more plausible that it would be possible to go along with Mormonism in something like good faith, even if only jokingly or subtly-ironically or something.)
Will, out of curiosity; do you enclose your comments in parentheses to give them the quality of a “whispered aside”, as if the camera had cut to a couple of conversants sitting in the back stalls? Because that’s what it does in my brain.
I give a 70% chance that Mormon doctrine holds that Jesus is accidental (in the sense of not existing in all possible worlds). He has a physical body, after all. For that matter, so does God.
Mormon theology is too weird for me to fully grok, though.
“so many advantages” is optimistic in my opinion; I actually think it’s an at least somewhat close call. There are also upsides to status differences, like better group coordination (as I mentioned earlier). If people know there are methods for them to attain high status, and pursuing high status using these methods can have positive side effects (e.g. starting companies that make products people want and generate consumer surplus, or writing blog posts that lots of people benefit from reading), that can be a good thing. Another thing: when you’re having a conversation, you’d probably prefer for the most knowledgeable/intelligent/rational people to talk more than those who are less knowledgeable/intelligent/rational, and status differences often seem to have the side effect of accomplishing this. (But you can also get a suits/geeks type thing where some people are smooth talkers and some people know lots of math.)
(These are just my thoughts, I’m sure there’s more stuff that hasn’t occurred to me.)
If people know there are methods for them to attain high status, and pursuing high status using these methods can have positive side effects (e.g. starting companies that make products people want and generate consumer surplus, or writing blog posts that lots of people benefit from reading), that can be a good thing.
Only in situations where the cost of failure is low. One of the larger failure modes I’ve experienced in status games is that the difference between success and failure is a narrow and often random margin, and yet the status payoffs are insanely amplified and tend towards a positive feedback loop (the Matthew Effect again). So often times, you don’t actually get a proper selection pressure that leads to the more intelligent/knowledgable/rational people acquiring more status; what you get is the people who know how to leverage their current status get more status. And once you have that, you’re “locked in” to an oligarchy for good or ill.
I have an idea for eliminating status on LW, if that’s what people want. My own status is ‘glad I’m allowed in here at all’, so it wouldn’t make a difference for me personally. ;-) What if your posts didn’t show your username, but just a post ID, and you yourself could see your karma, but no-one else could? There might be problems with PMs, but I’m sure there are programmers here who could find a solution to that.
What if your posts didn’t show your username, but just a post ID, and you yourself could see your karma, but no-one else could? There might be problems with PMs, but I’m sure there are programmers here who could find a solution to that.
Your suggestion would indeed eliminate most status and reputation influences from the site. And this would be a bad thing.
I prefer to know who I am reading, even if, as in the case of many usernames here, the knowledge is no more than “this is the same person who wrote these other things”. It gives context to the words: what they mean can depend very much on who is saying them. And one can hardly have a coherent conversation if there is no way to join up separate comments into a single identity.
It’s not a bad idea if that werewhatttt people wanted, but there are people I definitely want to ignore on here, and people who I think worth spending more time on than others.
Geh, got to update in favour of some behaviors being more common than I thought now.
I’m not sure that removing usernames is necessarily a good idea; they have a valid and important benefit.
Let us assume that a person says X. I suspect that X is most likely incorrect. I then look at that person’s username. If:
a) The username is one that I recognise, and belongs to a person who I have found is right far more often than wrong; then I take a closer look at X, and ask the person to explain, and generally put some effort into investigating X. It is likely that X is not as wrong as I thought, and I would learn something.
b) The username is one that I recognise, and belongs to a person who often posts things that are incorrect. I don’t bother to waste time trying to research X, since I am now even more confident that X is wrong.
c) The username is not one that I recognise, or it is one that I recognise but have not formed an opinion on yet. I may spend a small amount of effort thinking about X; but I am likely to nudge the username a little closer to category b.
In this way, I can optimise the amount of effort I put into trying to see which statements are correct, by putting the most effort into statements from which I am most likely to learn something new.
On the contrary, high-status people can countersignal by publicly changing their mind on things in light of new evidence. You just have to show the evidence as well as the changing of your mind. I mean, if someone’s right, that’s one thing—but publicly changing your mind distinguishes you from people who are merely right by demonstrating the process behind getting things correct.
On the contrary, high-status people can countersignal by publicly changing their mind on things in light of new evidence.
Sometimes. In particular circumstances. With difficulty. Even in circumstances that are abnormally in favour of sanity the status signal is still arguable. But note that effectively gaining social power isn’t about just signalling high status a lot. It’s about navigating social interactions with whichever signals are most effective. Someone who only signals high status comes across as ‘rigid’ or ‘brittle’. I suggest that much of the signalling benefit for mind changing is actually signalling competence and increasing likeability rather than by directly signalling high status in the moment.
On the contrary, high-status people can countersignal by publicly changing their mind on things in light of new evidence.
I agree. And there’s a trick to it, which you described pretty well. I’m just giving that as an example of how big a deal status is: if you don’t know the trick for changing your mind and staying high status, then it can be hard to change your mind, and difficulty changing one’s mind may be the #1 rationality failure mode in the general population.
On a related topic, see my comments on whether status differences serve useful community functions. My current guess is that status differences are counterproductive on net for achieving community goals, but I’d be interested to read counterpoint if anyone’s got any (especially you, Mr. High Status Person).
Ideally status could be replaced by domain specific estimates of competence, reliability, trustworthiness etc. But in practice nobody has the time. We have to summarize.
For humans, social status is much more than just an aggregate estimate of competence/reliability/trustworthiness. It motivates us, distorts our thinking, plays a key role in our politics, etc. To take just one example, I suspect that the main reason it’s so hard for most people to change their mind is because they don’t know how to do it in a way that preserves their status. For many people and social groups, admitting you’re wrong means losing face, and most people don’t like to lose face, so they resist publicly changing their mind.
(This is another reason why status differences may be counterproductive for rational communities… they could create an incentive for high-status people to not change their mind about things, since they have something to lose. The evidence may very well justify thinking one thing one week, then something else the next week, then something else the third week. But if you’re changing your mind about critical issues every week, it won’t be long before typical humans take you less seriously. Which is unfortunate.)
Also, this doesn’t sound like your true objection to me. It doesn’t take very many more bits of information to transfer 3 estimates on each of competence, reliability, and trustworthiness than a single aggregate number. And people communicate specific info all the time (“how good is X at Y? do you trust Z?”). It’s not obvious to me that a single aggregate quantity is frequently useful. Let’s say I introduce a friend to you and say his status is 67⁄100; was that useful information? (And in practice, peoples’ status is often determined by relatively silly things like how many friends they have, what status they’re perceived to have, how confident they act, and how confidently they talk. Another reason status sucks: it gives people an incentive to make confident predictions; see Philip Tetlock’s work on how confident experts are more likely to be wrong and more likely to be quoted in the media.)
(I don’t think I’ve got a clear idea of how best to make use of humans’ status wiring; I’m just kind of exploring different ideas at this point. But it seems like an important and neglected topic.)
Eliminating status differences has been tried and failed. If a hiring manager ever tells you “There are no office politics here”, then don’t take the job. There WILL be politics, except that it will be taboo to publicly admit it—and nobody will help you if you have a problem.
“X has been tried and faied” remains true until someone succeeds. If a thing with so many advantages has been tried and failed, then the solution is not to give up and make an equivalent utterance to “man was not meant to fly”; it is to examine why it failed, explore what the underlying rules and mechanics might be, construct a strategy based on those underlying rules and mechanics, and then try again.
Let me rephrase, then: declaring that you’ve eliminated status differences, when, in reality, you haven’t, is a relatively common mistake that tends to cause problems.
See also.
Aha, much more understandable. Thank you.
In that case: what would you surmise from a hiring manager that said “there’s office politics everywhere, of course, but we try to take an active role in minimizing their impact, and part of you being a good fit here will depend on your ability to help us with that goal.”?
(I regretfully confess that my own reaction to that statement would depend on that hiring manager’s gender, and (if male) how tall he was and how deep his voice was).
Perhaps a good way to deal with the situation in that XKCD comic would be to try to pick a culture that seemed particularly effective and then copy all of its norms, attitudes, etc.? So you’d have something that was battle-tested, if you will.
...Mormons? I don’t wanna. Even though it would probably work.
Well, Valve’s profitability per employee is supposedly higher than Google or Apple’s, and their employee handbook detailing their unconventional corporate culture is available for viewing online. shrug
Eh, it seems worth investigating to me.
(For what it’s worth from what I can tell Mormons don’t even formally make the sort of ontological commitments that are typical of (at-least-somewhat-reflective) mainstream Christianity (like, ‘Jesus is my savior and I should have expected Him to show up in all logically possible worlds and all possible minds should be rounded-up-to-infinitely compelled by His story and the seemingly contingent features of Jesus [Jesus’s teachings] are actually universal features of Logos and so it would be an obvious epistemic sin to disregard Him [them]’) and so it’s more plausible that it would be possible to go along with Mormonism in something like good faith, even if only jokingly or subtly-ironically or something.)
Will, out of curiosity; do you enclose your comments in parentheses to give them the quality of a “whispered aside”, as if the camera had cut to a couple of conversants sitting in the back stalls? Because that’s what it does in my brain.
More or less, yeah. Vladimir Nesov has a similar but distinct habit.
Hmmm, I hadn’t thought of that before.
I give a 70% chance that Mormon doctrine holds that Jesus is accidental (in the sense of not existing in all possible worlds). He has a physical body, after all. For that matter, so does God.
Mormon theology is too weird for me to fully grok, though.
(Eat some sauce.)
“so many advantages” is optimistic in my opinion; I actually think it’s an at least somewhat close call. There are also upsides to status differences, like better group coordination (as I mentioned earlier). If people know there are methods for them to attain high status, and pursuing high status using these methods can have positive side effects (e.g. starting companies that make products people want and generate consumer surplus, or writing blog posts that lots of people benefit from reading), that can be a good thing. Another thing: when you’re having a conversation, you’d probably prefer for the most knowledgeable/intelligent/rational people to talk more than those who are less knowledgeable/intelligent/rational, and status differences often seem to have the side effect of accomplishing this. (But you can also get a suits/geeks type thing where some people are smooth talkers and some people know lots of math.)
(These are just my thoughts, I’m sure there’s more stuff that hasn’t occurred to me.)
Only in situations where the cost of failure is low. One of the larger failure modes I’ve experienced in status games is that the difference between success and failure is a narrow and often random margin, and yet the status payoffs are insanely amplified and tend towards a positive feedback loop (the Matthew Effect again). So often times, you don’t actually get a proper selection pressure that leads to the more intelligent/knowledgable/rational people acquiring more status; what you get is the people who know how to leverage their current status get more status. And once you have that, you’re “locked in” to an oligarchy for good or ill.
I have an idea for eliminating status on LW, if that’s what people want. My own status is ‘glad I’m allowed in here at all’, so it wouldn’t make a difference for me personally. ;-)
What if your posts didn’t show your username, but just a post ID, and you yourself could see your karma, but no-one else could? There might be problems with PMs, but I’m sure there are programmers here who could find a solution to that.
Well, there is the LW anti-kibitzer, which can be enabled via the Preferences page.
Your suggestion would indeed eliminate most status and reputation influences from the site. And this would be a bad thing.
I prefer to know who I am reading, even if, as in the case of many usernames here, the knowledge is no more than “this is the same person who wrote these other things”. It gives context to the words: what they mean can depend very much on who is saying them. And one can hardly have a coherent conversation if there is no way to join up separate comments into a single identity.
It’s not a bad idea if that werewhatttt people wanted, but there are people I definitely want to ignore on here, and people who I think worth spending more time on than others.
Geh, got to update in favour of some behaviors being more common than I thought now.
I’m not sure that removing usernames is necessarily a good idea; they have a valid and important benefit.
Let us assume that a person says X. I suspect that X is most likely incorrect. I then look at that person’s username. If:
a) The username is one that I recognise, and belongs to a person who I have found is right far more often than wrong; then I take a closer look at X, and ask the person to explain, and generally put some effort into investigating X. It is likely that X is not as wrong as I thought, and I would learn something. b) The username is one that I recognise, and belongs to a person who often posts things that are incorrect. I don’t bother to waste time trying to research X, since I am now even more confident that X is wrong. c) The username is not one that I recognise, or it is one that I recognise but have not formed an opinion on yet. I may spend a small amount of effort thinking about X; but I am likely to nudge the username a little closer to category b.
In this way, I can optimise the amount of effort I put into trying to see which statements are correct, by putting the most effort into statements from which I am most likely to learn something new.
Sometimes you need to do things like ban a troll...
On the contrary, high-status people can countersignal by publicly changing their mind on things in light of new evidence. You just have to show the evidence as well as the changing of your mind. I mean, if someone’s right, that’s one thing—but publicly changing your mind distinguishes you from people who are merely right by demonstrating the process behind getting things correct.
Sometimes. In particular circumstances. With difficulty. Even in circumstances that are abnormally in favour of sanity the status signal is still arguable. But note that effectively gaining social power isn’t about just signalling high status a lot. It’s about navigating social interactions with whichever signals are most effective. Someone who only signals high status comes across as ‘rigid’ or ‘brittle’. I suggest that much of the signalling benefit for mind changing is actually signalling competence and increasing likeability rather than by directly signalling high status in the moment.
I agree. And there’s a trick to it, which you described pretty well. I’m just giving that as an example of how big a deal status is: if you don’t know the trick for changing your mind and staying high status, then it can be hard to change your mind, and difficulty changing one’s mind may be the #1 rationality failure mode in the general population.
Another risk of status differences is that good ideas from low status people may get ignored.
My impression is that LW is fairly good about taking people’s behavior one item at a time.