Can you clarify what you think the central assumption in the article is? After typing the rest of this I think I may have misunderstood you. I think the central

Can you clarify what you think the central assumption in the article is? After typing the rest of this I think I may have misunderstood you. I think the central assumption is stated in the first sentence: "One oft-underestimated threat to epistemic rationality is getting offended." I don't think it's much of a leap to generalize that to "noticing others getting offended is an indicator that their epistemic rationality isn't up to snuff." But it is inconsistent with a common theme around LW that learning about biases etc. should only be about improving yourself, not about noticing errors in other people. (I say why not both, while just being careful and not over-confident with the latter?) I should clarify that I don't think someone's "I'm offended" level completely measures their rationality, it's just a useful indicator. (They may have a Blue v. Green issue that just hasn't come up yet, or they may fail hard in the many other ways humans can be irrational, or my "this person is offended" detectors might be screwy.) I agree with the assumption that many, and probably most, seemingly rational people can become quite seemingly irrational when prompted with the right subjects. Or as Tesla put it, "The scientists of today think deeply instead of clearly. One must be sane to think clearly, but one can think deeply and be quite insane." Blue v. Green dynamics, and generally group identification, are pretty embedded in our neurological makeup, and to me only seem connected to "being offended" in the way that such a mind state makes it easier to fall into those (and other usually undesirable) modes of thought. I think all bridges can be burnt given enough time, but not all of them through trying to offend alone. With close friends who share your mental structure of finding it hard to be offended, let alone difficult to express it without a trace of irony, it takes other methods to destroy that friendship. The act of incessant trolling by itself could work depending on the person, regardless of if that person finds anything said offensive or not, simply because if I spend all my time trolling them, then I'm not worth their time. I read Gulliver's Travels around the same time as "A Modest Proposal", I agree it's still relevant and insightful. I also think you're right about esr's post having a problem by not defining "racist", and leaving the reader to do it themselves. (There are many definitions out there.) But I do think people who define "racist" as including something similar to "citing statistics that don't paint a pretty picture of all the various races, where such a pretty picture would show that such a simple observable never correlates with anything negative, and giving reasonable advice assuming those statistics are accurate" are in error, regardless of the accuracy of the statistics.