Wikipedia has two kinds of problems. The first category includes problems it recognizes and realizes how to fix, sometimes through a policy change but more often, in recent years especially, by administrative actions or PR activities led by the Wikimedia Foundation. For example, educators once warned students away from Wikipedia, but now editing Wikipedia is an increasingly common pedagogical tool, for which a great deal of credit is owed to the Wiki Education Foundation.

The second type of problem comprises those issues it cannot or will not fix, for reasons as diverse as the problems themselves. This past week brings us another example, highlighted by a September 29 column in the Washington Post Magazine by Gene Weingarten, titled “Dear Wikipedia: Please change my photo!” This comes more than four years after Philip Roth published “An Open Letter to Wikipedia” online at The New Yorker. In each case, both men found fault with their biographical entries on Wikipedia, and used their access to the mainstream media to call attention to the changes.

I am not a vain guy. I am ugly and know it. But this was still a hard column to write. https://t.co/9NyIlfy7oj

— Gene Weingarten (@geneweingarten) September 29, 2016

The problem we are highlighting is that anyone who is written about in a Wikipedia entry typically has no idea what they can or cannot do if they have a problem with said entry. There is some awareness that editing one’s own biography is fraught with peril—“(One is evidently not allowed to alter one’s own entry.)” Weingarten explains in an aside that is effectively true, technically false, and debatable as a matter of Wikipedia guidelines, so who can blame him—but there is little understanding of what one is supposed to do instead:

I tried asking Wikipedia to change or delete this picture. No answer. So I did what any user can do, and deleted it myself, on seven occasions — which, yes, was in blatant and shameful contravention of all Wikimedia Commons policies blah, blah, blah.

Absent a clear path to offering feedback, Weingarten and Roth did they only thing they could imagine: they tried editing the “encyclopedia anyone can edit”. Oddly enough, this didn’t work. Looking at Weingarten’s edits, it’s not hard to see why his attempts to remove the photo were overturned: more than once he simply deleted the entire infobox. He might have been successful if he’d just removed the actual image link (but then again maybe not) however it stands to reason a middle-aged newspaper humor columnist might not be the most adept with markup languages. In Roth’s case, he asked his biographer to make the changes for him, which were overturned because available news sources contravened Roth’s preferred version.

When editing Wikipedia didn’t work, each finally turned their media access to their benefit, and this time they got results. Within hours of Weingarten’s article becoming available, Wikipedia editors gathered on the discussion page of his biography to determine what could or should be done about his plight. Meanwhile on Twitter, longtime Wikipedia contributor (and DC-based journalism professor) Andrew Lih engaged Weingarten in a conversation, trying to get a better photo for him, and explaining why his Washington Post headshot could not be used. Soon, another photo satisfying Wikipedia’s arcane image use policies was identified and added to the article, although it doesn’t seem Weingarten isn’t especially happy with it, either. Lih had previously invited Weingarten out to lunch and a quick photo shoot, and it sounds like this may still happen.

In Roth’s case, it was a more complicated matter: several book reviews had identified a character in Roth’s The Human Stain as “allegedly inspired by” a writer whom Roth denies was the character’s inspiration. In the short term, Roth’s objection was noted, but sometime after the entire matter was relocated to a subsection of the novel’s Wikipedia entry as “Anatole Broyard controversy”, explaining the matter more fully. This seems like the right outcome.

So, everything worked itself out, right? That’s just how Wikipedia works? Mostly, and yes, and this is nevertheless somewhat regrettable. The fact is Weingarten and Roth are both able to command a major media audience via a “reliable source” platform that the vast majority of people (and bands, brands, teams, companies, nonprofits, &c.) do not. The method they used to get action not only doesn’t scale, it rarely happens at all due to most article subjects’ fear of a “Streisand effect” bringing undue attention to their article. As Weingarten writes in his piece:

[I]it is also possible that this column will serve as a clarion call to every smart aleck and wisenheimer and cyber-vandal out there. Anyone can make ephemeral changes to my Wikipedia page, any time.

Fortunately, that hasn’t happened, but it isn’t an unreasonable worry. Fortunately for Weingarten, as a white male whose writing doesn’t really take sides on controversial issues, he’s not much of a target for the Internet’s troll armies and political agitators.

The causes of this failure are many. We can assign some blame to Wikipedia’s strict policies regarding copyrights and reliance on crowdsourced images which has made its often-poor celebrity headshots both a source of angst and amusement. We can assign some to Wikipedia’s confusing discussion pages, which are forbidding; a project was once in development to overhaul them, only to be mothballed after facing community critcism. We can assign some as well to the contradictory message of Wikipedia as the encyclopedia anyone can edit—just not when the subject is the one you know about best, yourself. And we cannot let Wikipedia’s editing community escape blameless; even as they are not an organized (or organizable) thing, the culture is generally hostile to outsiders, unless of course said outsiders can get their criticism of Wikipedia into a periodical they’ve heard of before.

In the four years since the Roth episode, Wikipedia has had time to come up with a process for accepting, reviewing, and responding to feedback. I’ve argued previously for placing a button on each entry to solicit feedback, feeding into a public queue for editorial review. The reasons not to do this are obvious: most of it would be noise, and there wouldn’t be enough editor time to respond even to those requests which might be actionable.

I still think the feedback button is a good idea, but I recognize it is not sufficient: it would also needs an ombuds committee set up to triage this feedback. Perhaps this could be community-run, but this seems too important to be left up to volunteers. This work could be performed by WMF staff even if, for complicated reasons every Wikipedia editor understands but would need a lengthy paragraph to explain, they could not implement them outright. And it’s not just a matter of making sure Wikipedia is accurate—though you’d think that would be enough!—it’s also a matter of making sure Wikipedia is responsible and responsive to legitimate criticism.

Of course, Wikipedia already operates on this very model, in a way: it solicits edits from its readership, and then also spends a lot of time reverting unhelpful edits, and the difference between bad edits with good intentions and bad edits with bad intentions is often impossible to tell. Providing a clear option for expressing a specific concern rather than forcing the expression of that problem to be an edit rather than a request is something Wikipedians should think about again. When someone is unhappy with their Wikipedia entry, that they have no idea what can be done about it isn’t really their fault. Ultimately, it’s Wikipedia’s. And it’s not just an abstract information asymmetry problem—it’s a PR problem, too.