Mm, perhaps rather than saying that most such people are untrustworthy, I just want to instead make an argument about risk and the availability of evidence.
Some people are very manipulative and untrustworthy and covertly break widespread social norms.
Some people covertly break widespread social norms for good reasons.
Even if you find out one time people are covertly breaking a norm, you do not know how much more often they are covertly breaking social norms, and it’s hard to understand the reasoning that went into the one you have learned about.
Suppose the amount of covert social norm breaking is heavy-tailed, where 90% of people break none, 8% of people break 1, 1% of people break 2-3, and 1% of people break 4+ (and are doing it all the time).
If you find out that someone breaks one, then you find out that they’re not in the first bucket, and this is a 10x multiplier toward them being the sort of person who breaks 10+. So this is pretty scary.
And what’s worse is regardless of which bucket they’re in, they’re not going to tell you which bucket they’re in. Because they’re not going to volunteer to you info about other norms they’re breaking.
So (if this model/distribution is accurate) when you find out that someone has covertly broken a widespread social norm, you need to suddenly have your guard up, and to be safe you should probably apply a high standard before feeling confident that the person is not also violating other norms that you care about and keeping that from you.
(I just want to acknowledge in my comments I’m doing a lot of essentialism about people’s long-standing personality traits, I’m not sure I’d endorse that if I reflected longer.)
2 seems both true and obvious to me (and we have a rich historical record of many of those people being vindicated as moral development proceeded apace).
3 seems true and correct to me as well.
Our divergence is after 3, in the rough model. I think that it is waaaaaaay unlikely that a 90% bucket is the right size. I think that 50+% of people covertly break at least 1 widespread norm, and even if someone talks me out of it I do not expect them to talk me even half of the distance down to 8%.
Mm, perhaps rather than saying that most such people are untrustworthy, I just want to instead make an argument about risk and the availability of evidence.
Some people are very manipulative and untrustworthy and covertly break widespread social norms.
Some people covertly break widespread social norms for good reasons.
Even if you find out one time people are covertly breaking a norm, you do not know how much more often they are covertly breaking social norms, and it’s hard to understand the reasoning that went into the one you have learned about.
Suppose the amount of covert social norm breaking is heavy-tailed, where 90% of people break none, 8% of people break 1, 1% of people break 2-3, and 1% of people break 4+ (and are doing it all the time).
If you find out that someone breaks one, then you find out that they’re not in the first bucket, and this is a 10x multiplier toward them being the sort of person who breaks 10+. So this is pretty scary.
And what’s worse is regardless of which bucket they’re in, they’re not going to tell you which bucket they’re in. Because they’re not going to volunteer to you info about other norms they’re breaking.
So (if this model/distribution is accurate) when you find out that someone has covertly broken a widespread social norm, you need to suddenly have your guard up, and to be safe you should probably apply a high standard before feeling confident that the person is not also violating other norms that you care about and keeping that from you.
(I just want to acknowledge in my comments I’m doing a lot of essentialism about people’s long-standing personality traits, I’m not sure I’d endorse that if I reflected longer.)
1 seems both true and obvious to me.
2 seems both true and obvious to me (and we have a rich historical record of many of those people being vindicated as moral development proceeded apace).
3 seems true and correct to me as well.
Our divergence is after 3, in the rough model. I think that it is waaaaaaay unlikely that a 90% bucket is the right size. I think that 50+% of people covertly break at least 1 widespread norm, and even if someone talks me out of it I do not expect them to talk me even half of the distance down to 8%.