How Wikipedia Got It Wrong On Chelsea Manning, And Why


On August 22, the soldier formerly known as Bradley Manning announced she wanted to be called Chelsea, and recognized as female. Within an hour, the English-language Wikipedia had renamed the article and changed its pronouns. That kicked off a week-long often-acrimonious debate among Wikipedia editors, which resulted, over the weekend, in the article title being reverted back to Bradley Manning. (For further context, here’s a news story and some detailed analysis.)

This disappointed me. I think Wikipedia did the right thing in promptly responding to Manning’s announcement, and I feel like reverting the change was, for many reasons, a bad call. And, I’m dismayed by subsequent on-wiki discussions that look like they may result in revisions to editorial policy that would support, in the future, similar bad decisions on trans issues.

The purpose of this blog post though, isn’t to argue why Wikipedia should respect that Chelsea Manning is a woman. (I’ve already done that extensively this past weekend on-wiki, as have others.) The purpose is to talk about the underlying factors that I think led Wikipedia to make a bad decision, and point to where I think solutions might lie.

First, I want to question some of the frames people are currently using to understand what happened here. Some editors have been characterizing the debate as political in a narrow sense — with left-leaning Wikipedians favouring use of Chelsea and right-wing Wikipedians favouring Bradley. Others are framing it as accuracy versus kindness — saying that editors who support using Chelsea are motivated by a desire to not hurt her feelings, and are prioritizing that goal over truth. I think both those frames are wrong. I think what the debate is actually revealing is a blind spot that Wikipedia has about gender.

Last summer I drafted a blog post about Wikipedia’s coverage of rape. I haven’t published it yet, and at this point I think I probably never will: its moment may have passed. But the similarities between what I observed in our discussions about rape, and what I’ve observed over the past few days as we talked about trans issues, are I think instructive. In both cases a non-negligible chunk of our editing community, all of whom ordinarily value accuracy and knowledge and expertise, seem instead to be going on gut feeling and knee-jerk assumptions. We aren’t applying our normal judgement and standards.

On rape and trans issues, the typical editor probably doesn’t have much expertise. That’s not at all unusual — people edit on topics they don’t know much about, all the time. We fix typos, we smooth out writing styles, we dig up and add citations. What’s unusual here is that rather than deferring to people who had read and thought a lot about the article topic, as we normally do, instead a substantial chunk of the community seemed to let itself be swayed by prejudice and unexamined assumptions. (To be clear: I do not mean that a large chunk of Wikipedians are themselves prejudiced. I mean that they let prejudiced ill-informed people establish a tone, an overton window if you will, of what was acceptable to think and say.)

Here’s what’s normal: When Pluto was declared to not be a planet, Wikipedia deferred to the experts, and reflected what they said in the article.

Here’s what’s not normal: When Dominique Strauss Kahn was accused of having raped a hotel cleaner, and when Todd Akin made pseudo-scientific claims about rape and pregnancy, many Wikipedians’ discussions were (I thought) remarkably ill-informed. Some editors seemed to believe that false accusations of rape were common. Some didn’t seem to realize that rape is seriously underreported. They didn’t recognize that there’s a body of knowledge on rape that’s well-sourced and reliable.

It took me a while to connect this to systemic bias — to realize that rather than Wikipedians being unusually lacking in knowledge about what rape is and how it works, I might better understand it as me being more-knowledgeable-than-the-average-Wikipedian on the topic. Because I’m a woman, and also a journalist, I’ve followed rape issues pretty closely in the media, I’ve talked about it a fair bit with my female friends, and I’ve read a couple of dozen books and studies on it and related topics. It took me a while to realize that that level of interest, and therefore expertise, is unusual on Wikipedia, presumably at least partly because our editor community skews so heavily male.

The same is true for transgender issues. A number of editors have made truly ignorant comments over the past week or so, comparing Chelsea Manning to someone who woke up one morning believing herself to be a dog, a cat, a Vulcan, Jesus Christ, a golden retriever, a genius, a black person, a Martian, a dolphin, Minnie Mouse, a broomstick or a banana. In saying those things, they revealed themselves to be people who’ve never thought seriously about trans issues — who have never read a single first-person account of growing up transgendered, or a scholarly study or medical text, or maybe even the Wikipedia article itself. That in itself is perfectly okay: different things are interesting to different people, and I for one know nothing about trigonometry or antisemitism in the 19th century or how a planet is determined to actually be a planet. But I don’t deny that there is stuff on those topics worth knowing, nor do I mock the knowledge of others, nor accuse them of bias and POV-pushing.

Wikipedians normally don’t either. Wikipedians won’t ordinarily defer to someone just because of their credentials, but we do normally attach extra credibility to people who’ve demonstrated they know more about a topic than we do. In this instance though, Wikipedians are considering sanctioning the two thoughtful and well-informed editors who originally made the change from Bradley to Chelsea. Which to me suggests systemic bias fuelled by groupthink.

So what needs to happen now?

The entire controversy has been referred to Wikipedia’s Arbitration Committee, the body of experienced editors who’ve been elected by the community to adjudicate difficult issues.

My hope is that ArbCom will clarify Wikipedia policy, and affirm that we have a responsibility to respect the basic human dignity of article subjects, to not mock or disparage them, and to attempt to avoid doing them harm. That we must not participate in or prolong their victimization.

I also hope ArbCom will weigh in on how Wikipedia handles trans issues in general. I’d be particularly interested in an examination of the role that subject-matter expertise is playing in our current discussions, and an exploration of how editors might choose to conduct themselves in disputes in which they have little expertise, and in which systemic bias risks skewing outcomes. In the Manning situation, for a variety of reasons that almost certainly include systemic bias, discussion didn’t achieve a result consistent with our desire to protect the dignity of an article subject.

So. Here’s the question. Given that Wikipedia makes decisions by consensus, how can majority-culture (male, young, Western, heterosexual, cisgendered) editors best participate in discussions in ways that work towards good decision-making, rather than groupthink?

For anybody reading this who doesn’t know: I’m the ED of the Wikimedia Foundation, and I’m also a Wikipedia editor. It’s in my latter, volunteer capacity that I wrote this blog post. What I say here is mostly informed by my experiences editing, but of course my experiences as ED have also shaped my opinions. Also: everything I say here, I say with lots of respect for the Wikipedia community. This is a rare misstep: an unusual and unfortunate blind spot.

Thank you to the people who vetted and gave feedback me on the draft version of this post — I appreciate it.

 

Leave a comment