Wikipedia, the internet’s encyclopedia, is run entirely by volunteers—people who spend large swaths of their personal time making sure the information that hundreds of millions of people access every day stays accurate and up-to-date. Of those volunteers, 77 percent of Wikipedia articles are written by just one percent of Wikipedia editors. As such, tensions tend to get a little high, because these editors are often highly invested. They’ve been arguing about corn for nearly a decade, for example, and there’s a long-running edit war about the meaning of neuroticism.
When editors disagree about an edit to be made on a Wikipedia article, they start by discussing it on the article’s Talk page. When that doesn’t result in a decision, they can open a Request for Comment (RfC). From there, any editor can choose a side or discuss the merits of whatever edit is up for discussion, and—in theory—come to an agreement. Or at least, some kind of decision about how to make the edit.
But a new study by MIT researchers found that as many as one-third of RfC disputes go unresolved, often abandoned out of frustration or exhaustion. The most common sticking points were chalked up to inexperience, inattention from experience editors, and just plain petty bickering.
The researchers, led by Jane Im, are presenting the study this week at the ACM Conference on Computer-Supported Cooperative Work and Social Computing).
From 2011 to 2017, the team monitored 7,316 RfC discussions and interviewed editors who frequently closed discussions to find out why these conversations went unresolved. They chose to examine RfCs over other discussion methods, like majority votes or moderator panels, because they operate in a much more nuanced way: Reaching a consensus is more important than voting or convincing a majority to side with you.
Nearly 58 percent of the RfCs they monitored ended up being formally closed and resolved. Of the 42 percent that weren’t, almost 80 percent were just left out to go stale—no one was participating or closing the conversation. Waiting around for a resolution is a waste of time and effort, the researchers wrote.
The biggest reasons these Wiki-beefs weren’t resolved included “poorly articulated initial statements by inexperienced discussion initiators, lack of interest from third-party experienced Wikipedia editors, and excessive bickering or contentiousness during the discussion,” the researchers wrote in the study. Often, the fights started because participants had a history of beefing in past edits, and were intentionally trying to sabotage the process.
Other arguments were complicated by a practice called “sock-puppeting,” where users made fake accounts to boost their side. Sometimes, editors will recruit other editors to jump into the RfC on their behalf, which, the study explains, is called “meatpuppeting” if they’re recruited off-Wikipedia and “canvassing” if it happens within the Wikipedia community.
But they didn’t just critique how Wikipedians argue: The researchers developed a tool called Wikum that they say will help resolve more discussions, and make it easier for editors to stay involved when arguments get gnarly. The tool uses the data they found and analyzed in this research, to summarize threads and predict when they’re at risk of going stale.
“The work of closer is pretty tough,” Amy Zhang, one of the authors on the study, said in an MIT press release. “So there’s a shortage of people looking to close these discussions, especially difficult, longer, and more consequential ones. This could help reduce the barrier to entry [for editors to become closers] and help them collaborate to close RfCs.”