Masters of the Wikiverse

Wikipedia’s new editing policy isn’t the end of the encyclopedia’s democratic age. It’s business as usual.

The council of elders that runs Wikipedia confirmed last week that, sometime soon, the unwashed masses will no longer be able to directly edit the profiles of famous living people. The proposed policy, first reported by the New York Times and later clarified by the Wikimedia Foundation, would require an ordained Wikipedia editor to approve changes before they become visible. This was widely reported as a stress fracture in Wikipedia’s sacred anyone-can-edit architecture. Fast Company painted it as Wikipedia’s coming of age. A commenter on Slashdot compared the move to “Lenin abolishing free elections.”

Cooler heads were quick to cast the policy as a fairly minor revision to Wikipedia’s existing defense mechanisms. Currently, anyone can edit most pages without even signing up for a username, in which case the IP address is logged with the changes. Pages that are prone to vandalism, like those for Jesus, George W. Bush, and episodes of iCarly, are protected so that greener users can’t edit them. Still, the bar for access is not terribly high. To edit a protected page, you need to register a username, wait four days, and make at least 10 edits to unprotected pages. Wikipedia’s new proposed policy, called “flagged protection,” gets rid of this detox period and requires that unregistered users have their edits approved by a more seasoned Wikipedian. (The Wikimedia Foundation hasn’t clearly defined how you get to be a reviewer who can approve edits.)

Wikipedia brass says that the site’s new rules will apply only to pages for high-profile living people who are most likely to be harmed by misinformation. Unapproved edits will still be visible for users who sign in, though most Wikipedia browsers probably won’t take that step.

No matter how you spin this new policy, there’s no getting around that it gives more power and control to a small group of people. But if this were a big problem, Wikipedia would have flopped a long time ago. As I’ve argued before, the encyclopedia’s success is largely due to the devoted efforts of a small number of obsessive editors, many of whom are quick to undo the work of trespassing newcomers. Rather than a signal Wikipedia’s coming of age or a shift away from democracy, these new rules merely formalize, for certain pages, what’s already happening on the site.

Take, for example, the extensive analysis of the Wikipedia ecosystem by Ed Chi and his colleagues at the Palo Alto Research Center. About half of all edits, they found, come from users who have made at least 100 changes to the site, and 20 percent of edits come from those who have made 1,000 or more changes. (See the graph here.) On the other end, Chi and company found that by the end of 2008, first-time users had a 25 percent chance of having their change to Wikipedia undone by someone else. (That doesn’t include changes that are obvious vandalism.) That figure dropped to around 15 percent for the user’s next eight edits. Users who’ve made more than 100 edits, meanwhile, have their fixes undone 1 percent of the time.

This wiki-oligarchy presents its own sort of dangers. The movers and shakers of Wikipedia are largely hidden from public view and unaccountable for editorial decisions. It is fairly easy for one person to establish sovereignty over a less-trafficked page through sheer persistence and a solid command of the site’s Byzantine rules for resolving disputes. For example, consider the curious case of sports writer Rick Reilly’s affinity for tooth jokes, which my colleague Josh Levin chronicled last December. Levin’s story made it into  Reilly’s Wikipedia page, where it was removed by a user who turned out to be Reilly’s PR representative. The page’s edit history tells the rest of the story: Someone reinserted a reference to the tooth article, someone else removed it again, someone else added it back, and finally a user named Dayewalker, who has made well more than 1,000 edits on Wikipedia, weighed in: “[H]umorous personal analysis of his style from a dental perspective is undue, and unsuitable for an encyclopedia. Please discuss on talk page and try and gain consensus.” No one took up the fight, and Reilly’s page remains free of Levin’s critique.

Wikipedia is surely full of these little omissions, some of which have a lot more teeth than Rick Reilly’s taste in figurative language. The hegemony of the devoted guardians of Wikipedia, the vast majority of whom have the best of intentions, is the price we pay for a site with such breadth and startling accuracy. Wikipedia isn’t the Web’s largest, most popular reference work because it adheres to the ideals of Web 2.0. It’s become so pervasive because it’s free, turns up high in Google searches, is easy to parse, has entries on just about everything, and, most importantly, is almost always right. (Where else can you turn for a convenient list of which characters get killed in every episode of The Sopranos?) As long as this remains true, people will continue reading it.

While it’s not quite right to say that the move to flagged revisions is a sign that Wikipedia is coming of age, it is true that the encyclopedia is maturing. At this point, the site is no longer growing at exponential rates. The PARC research team found that edits to the site have plateaued in the last 18 months, having peaked around January 2007. While Wikipedia is certainly still growing, it appears that a larger percentage of the work is devoted to maintaining its current quality—something the flagged revision policy aims to make easier.

My guess is that you wouldn’t notice the difference in Wikipedia even if the site restricts editing on a huge number of pages. There would be fewer embarrassing or harmful wiki-scandals, like when the site briefly reported on the day of President Obama’s inauguration that Sens. Ted Kennedy and Robert Byrd had both died. (Both had health problems that day but survived.) Sure, the site would lose a little responsiveness, though probably not much. Stats from the German Wikipedia, where this policy is already widely in place, suggest that most edits are approved or denied quickly, with a median lag of about six-and-a-half hours. Covering breaking news has never been Wikipedia’s strong suit, anyway. That’s why we have newspapers—for now.