This is a great point, but maybe I’m just saying that because it’s the exception that proves the rule. Just by arguing with someone with a fringe viewpoint, you’ve granted them roughly equal status, so they will be highly hostile as a status grab. However, in a group, these fringe viewpoints have a history of rewarding their advocate with exile—so the advocates will make a show of giving away status in that circumstance.
Compare this vs. mainstream, acceptable views—say, conservative vs. liberal in private vs. public. It’s much easier to have a productive conversation in private about these views than in public.
“the exception that proves the rule” seems like a very un-Bayesian thing to say. The implication is that both X and ~X provide evidence for the hypothesis. (Not that I always communicate my actual and complete hypothesis—sometimes that is a distraction from my main point.)
I think the implication is not that both X and ~X provide evidence for the hypothesis, but rather something like, “yes, there are a few exceptions to the rule, but if you look at what the exceptions are they’re so unusual that they just underline the fact that the rule is generally (though not universally) applicable.”
This is a great point, but maybe I’m just saying that because it’s the exception that proves the rule. Just by arguing with someone with a fringe viewpoint, you’ve granted them roughly equal status, so they will be highly hostile as a status grab. However, in a group, these fringe viewpoints have a history of rewarding their advocate with exile—so the advocates will make a show of giving away status in that circumstance.
Compare this vs. mainstream, acceptable views—say, conservative vs. liberal in private vs. public. It’s much easier to have a productive conversation in private about these views than in public.
“the exception that proves the rule” seems like a very un-Bayesian thing to say. The implication is that both X and ~X provide evidence for the hypothesis. (Not that I always communicate my actual and complete hypothesis—sometimes that is a distraction from my main point.)
I think the implication is not that both X and ~X provide evidence for the hypothesis, but rather something like, “yes, there are a few exceptions to the rule, but if you look at what the exceptions are they’re so unusual that they just underline the fact that the rule is generally (though not universally) applicable.”