Three years ago this month, historian Richard Carrier explained at some length why it is okay for us to conclude someone is guilty of sexual misconduct based on online claims republished from an (at the time) unnamed source:
I also believe the preponderance of evidence is sufficient to conclude [named person] probably has crossed moral lines. I have seen enough evidence to establish, in my own mind, at least a 50.1% chance that [named person] has not just cheated or fooled around, but has left a wake of victimized women in his path, that he has not conducted himself morally, and that he is probably not good or safe company (especially for women). Again, I am not witness to this. I am only inferring it from what has been said online by those who do claim to have witnessed evidence of it.
These days, Dr. Carrier has become much less sanguine about those who take such a reckless approach:
It is very disheartening to see feminists in our movement act exactly as MRA’s and other anti-feminists claim they would: by believing any claim told them without investigating it, and exaggerating ordinary and relatively harmless behavior into dangerous sexual harassment. This is a fundamentally broken epistemology. I have always said competent feminists don’t just “believe the accuser.” They take accusations seriously, by conducting or calling for a formal or at least competent investigation. That’s sound epistemology. Whatever epistemology The Orbit and Freethoughtblogs is deploying, it’s not sound. It’s dangerous. And I think this actually puts them now in a position where no one can trust them to accurately describe anything. And that severely harms the cause of feminism in the atheism movement.
When someone else’s reputation was on the line, Carrier did not call for a formal or competent investigation, but rather took up his pen to “viscerally describe rape and shame scenarios,” concluding that the accused in question “took advantage of a woman and destroyed a human being’s happiness” at the very least. Now that it is his own reputation on the line, Carrier has at long last discovered the virtues of epistemic circumspection. Better late than never, naturally, but it is rather difficult to take lessons in “sound epistemology” from someone who was more than happy to just “believe the accuser” despite not knowing her name or any of the relevant background details at the time of his original posting.
One final point. While Carrier was happily tearing down the reputation of another prominent skeptic, he roundly condemned the “use of the law to silence people from saying what happened to them” as comparable to “attempts to silence abuse victims with frivolous defamation suits.” These days, he sees no problem with using threats of lawsuits to silence his own accuser, as you can see from the cease and desist letter posted by Amy Frank. I would say Carrier was right about this the first time around, back when he condemned the use of libel law in a situation where much graver accusations were being put forward.
If you are free to change your mind about both the propriety of public accusations and the use of libel law against such accusations in ways that serve your own interests—as soon as they are threatened—there is a good chance that you are working from a fundamentally broken epistemology. An epistemology where you work backward from a foregone desirable conclusion. This is, of course, not how a Bayesian rationalist is supposed to proceed, nor is it a good example to set as a leading thinker in the freethought community.