Wikipedia talk:Wikipedia Signpost/Drafts/Disinformation report

Mistake or systematic failure?

Thanks for this very interesting and important article, Your Friendly Neighborhood Sociologist. Cantor went out of his way to encourage me as a newbie editor, so it's saddening to learn that he was and is using Wikipedia to further bigotry.

Anyway, I happened to come across this article before it was published, so I hope you don't mind if I leave some thoughts here for you to take or leave.

I sadly don't think we can chalk up ArbCom's failure in the Sexology case to a mistake. That is, I don't think it's as simple as the 2013 ArbCom just fucking up, and that if circumstances were different or if it were a different set of people they'd have got it right. Crucially, I also highly doubt that, faced with a similar case, we'd do any better now.

I'm not too familiar with the Sexology case but it has strong parallels to the 2019 antisemitism in Poland case, where the committee of the time (which included me) also fucked it up and missed an opportunity to put a stop to damaging disinformation targeted against a minority. I deeply regret that, and I don't mean to excuse myself or by implication the 2013 committee on Sexology, but I do think that the main factors in both cases were structural. More specifically, the bad decision grew out of ArbCom's framing as a body that can only pay attention to "conduct" and not content. So not only are arbs not selected to have the skills and knowledge necessary to deeply engage with content disputes, they are actively discouraged from doing so. This happens both through the structure of the arbitration process and what the community expects of arbs. In antisemitism of Poland, the real substance of the case was a dispute over reliability and NPOV in historiographical sources. But even if I had the time or expertise to dive into that, I had a strong feeling that doing so wasn't my job and that if I tried I would be criticised. So instead we looked at conduct—who said what, who reverted who, and so on—and as a result missed the fact that the guy who was behaving the worst was actually the one who was right on the issues.

I see something similar in Sexology. To reach the right conclusion, the arbs should have dived into the sources and the content arguments that were presented to them and seen that Cantor's views were fringe. Instead, they took the easy route and looked at which side was the worst-behaved. And I wouldn't discount the influence of Cantor's credentials here: he was one of very, very few active editors who could also claim to be a tenured and published expert (with a Wikipedia page!), and that produced an aura. For a group of pseudonymous Wikipedians to challenge him and say he's wrong in his area of purported expertise would have been a bold move. You're right that the COI aspect should have countered this, but we were also much less serious about COI back in 2013. Again, it's not to say they couldn't do these things or didn't have a choice, but the deck was stacked against the proper outcome. Maybe had the case happened in 2024, now that trans issues are better understood by the general public, it would have be easier to spot and name Cantor as a fringe figure, but I wouldn't put money on it.

So I don't, unfortunately, believe that we'll better with issues like this in the future. Not until the community accepts the fact that, just like their a intractable conduct disputes that can't be satisfactorily solved through consensus, there are intractable content disputes that can't be satisfactorily solved by amateurs discussing what reliable sources say. These should be sent to a body that is equipped to deal with them, not shunted to an ArbCom that deliberately has its hands tied behind its back when it comes to content. – Joe (talk) 10:34, 6 July 2024 (UTC)

P.S. I've shamelessly stolen your analogy in the last section for Wikipedia:Al Capone effect, I hope you don't mind. – Joe (talk) 11:58, 6 July 2024 (UTC)

@Joe Roe No problem, thank you so much for the feedback! You put it very well, particularly the structural issues with Arbcom focusing on conduct over content. Additionally, the notes on the COI issues being treated differently in WP's early years and the weight of credentials - others noted those too but it's been hard to find a RS or policy or etc to cite to bring it up. Would you mind if I refer to your note and explanation in the piece? Also, if you have RS touching on any of those concerns (conduct v content, early COI editing, credentialing) I'd appreciate those too!

Finally, good artists copy, great authors steal :p - don't mind at all and very much appreciated the essay! Best, Your Friendly Neighborhood Sociologist ⚧ Ⓐ (talk) 23:13, 8 July 2024 (UTC)