I don’t think there’s an easy fix for that: Owen Cotton-Barratt is not anonymous, and a prominent figure (at least in the UK community). As such, people will want to give their personal testimony of what threat he might pose or not. The women who reported him remain anonymous—no one will come to defend their character or great deeds for the community, since no one knows who they are.
I’d also flag the related risk of cognitive bias here—the women who were affected by Owen’s actions are in a sense abstractions to the vast majority of readers. By that I mean we (understandably) haven’t heard their stories in their own words or in any detail; we largely are reliant on the Boards’ (understandably) concise and not-too-specific factual findings. In contrast, Owen is known and has (mostly) positive history with people in the community.
That’s largely unavoidable, because of course we should honor the women’s desire not to be identified. But I would say that if someone’s reaction to what happened changes too much if they change from evaluating the abstract “a senior EA figure” to “Owen,” they may want to consider that they have been exposed to humanizing/de-abstracting information (I’d like a better word here!) for one participant but not for other participants in the situation.
I don’t know what epistemic strategy is best employed in this kind of circumstance. Having one person as a known entity with the other people feeling like abstractions could lead to a position that is more favorable to the identified person than it would be if humanizing/de-abstracting information existed on both sides. One could argue for falling back to the level of abstractions, deciding how one would react if only aware that “a senior EA figure” committed the actions, and update according to that—but that strategy comes at the cost of disregarding pretty relevant information. Maybe the best that can be done is to evaluate one’s position with and without the Owen-specific contextual information, and update to ~ the median of the two evaluations?
I’d also flag the related risk of cognitive bias here—the women who were affected by Owen’s actions are in a sense abstractions to the vast majority of readers. By that I mean we (understandably) haven’t heard their stories in their own words or in any detail; we largely are reliant on the Boards’ (understandably) concise and not-too-specific factual findings. In contrast, Owen is known and has (mostly) positive history with people in the community.
That’s largely unavoidable, because of course we should honor the women’s desire not to be identified. But I would say that if someone’s reaction to what happened changes too much if they change from evaluating the abstract “a senior EA figure” to “Owen,” they may want to consider that they have been exposed to humanizing/de-abstracting information (I’d like a better word here!) for one participant but not for other participants in the situation.
I don’t know what epistemic strategy is best employed in this kind of circumstance. Having one person as a known entity with the other people feeling like abstractions could lead to a position that is more favorable to the identified person than it would be if humanizing/de-abstracting information existed on both sides. One could argue for falling back to the level of abstractions, deciding how one would react if only aware that “a senior EA figure” committed the actions, and update according to that—but that strategy comes at the cost of disregarding pretty relevant information. Maybe the best that can be done is to evaluate one’s position with and without the Owen-specific contextual information, and update to ~ the median of the two evaluations?