One-Click Content, No Guarantees

Wikipedia users who logged on to the free online encyclopedia this January to do research on current members of the U.S. Congress may have been surprised to encounter the online equivalent of a playground shouting match. The official entry for Rep. Eric I. Cantor (R-Va.) noted that he smelled of "cow dung," and the blurb for Sen. Tom Coburn (R-Okla.) mentioned a dubious-sounding biographical detail: "Coburn was voted the most annoying senator by his peers in Congress. This was due to Senator Coburn being a huge douche bag."

Within hours, Wikipedia administrators had intercepted the renegade edits--but not before the incident provoked a nationwide media furor, spurring questions about the encyclopedia's credibility. As the first-ever major reference work with a democratic premise--that anyone can contribute an article or edit an entry--Wikipedia has generated shared scholarly efforts to rival those of any literary or philosophical movement in history. Its signature strength, however, is also its greatest vulnerability. User-generated articles are often inaccurate or irrelevant, and vandals like the political jokesters are a constant threat. As a result, the role of the encyclopedia's gatekeepers assumes added importance. Who are they, and how do they go about the business of deciding which new content will pass through their crucible?

Founded in 2001 by Jimmy ("Jimbo") Donal Wales, a former Chicago options trader, Wikipedia has morphed into a cultural phenomenon on a par with Google or Friendster. Internet users have contributed more than 3 million articles in 200 languages to the site, and every few seconds, a new article or edit is added to Wikipedia's 180-gigabyte database. Overseeing the entire gargantuan knowledge machine are the Wikipedia elite: about 800 longtime contributors who have volunteered to maintain the site and help ensure its accuracy.

The influx of information is so great that it's easy to characterize content-control efforts as potshots into a crowd, but Wikipedians--as regular contributors like to call themselves--claim the review process is actually carefully executed and multilayered. The first line of defense is the so-called recent changes patrol, an online SWAT team made up of hundreds of volunteers who comb new or recently modified content for errors. "If there's outright vandalism, the recent changes patrol will avert the situation fairly quickly," Wales says. "An easy deletion would be an article with the title 'asdfasdf' and content reading 'Hi Mom.'"

In many cases, however, the decision to keep or cut is not as straightforward. "A lot of stuff is borderline," Wales says. "Is it verifiable? Is it important enough to go into the encyclopedia?" Disputes among administrators--senior Wikipedians who have the power to block or roll back edits on an entry, or even to delete an entry outright--about the validity or relevance of a fact or article can lead to pages-long online debates. When Florida author and programmer Rogers Cadenhead wrote an entry about himself, for instance, the question at issue was not whether Cadenhead was guilty of self-promotion (that was a given), but whether he was notable enough to warrant his own entry. "Keep--author of popular books," one Wikipedian weighed in. "Writing a book itself does not mean the person should be included," another administrator fired back. "I looked up the books on Amazon, and [Cadenhead's] sales rankings are 30 000 and 80 000." In the end, Cadenhead's entry was kept--along with a note about the controversy.

Wales describes the give-and-take review process as similar to a collegiate debate round. After each Wikipedian speaks his or her piece, all administrators familiar with the issue are polled for a consensus, and changes are made accordingly. "If there's a dispute, it's all about being dutiful, citing sources, and doing the research," Wales says.

Unlike advisors at elder-statesman publications like the World Book Encyclopedia and the Encyclopaedia Britannica, both based in Chicago, Wikipedia administrators need not have scholarly credentials--the only requirements for the positions are keen research skills, a critical eye, and lots of spare time. This unconventional hierarchy is reflective of Wikipedia's core ethos: rather than coming directly from an established authority, content should emerge collaboratively and organically in what one administrator has dubbed a "social Darwinian evolutionary process." The more users and gatekeepers who weigh in on an entry, the thinking goes, the more detailed and accurate it becomes, ideally producing a whole greater than the sum of its parts.

Terrence Seamon, a human resources executive from New Brunswick, N.J., who has contributed dozens of entries, testifies to the unique alchemy of Wikipedia. "Sometimes I'll start an article," he says, "and when I revisit it days or weeks later, someone else has added interesting stuff to it that I couldn't have come up with on my own."

Many publishers and academics, however, have criticized the Wikipedia model on the grounds that it generates the informational equivalent of sludge. The lack of formal gatekeeping procedures, they say, ensures that the lowest common denominator will prevail--and since no experts or editors are hired to vet articles, no clear standards exist for accuracy or writing quality. "If you take any group of people that claim to know something, many of them will be wrong, but all of them will be equally confident," says Robert McHenry, former editor in chief at Britannica. "Leaving the [Wikipedia] encyclopedia open for anyone to contribute guarantees that its content and accuracy will tend toward the mediocre." McHenry also points out that the Darwinian strategy falls through when contributors submit articles on obscure topics, as few people have the knowledge necessary to criticize or edit the articles.

Wikipedia's power structure has spawned other problems as well. Like most other encyclopedias, it derives its reputation from readers' faith in its objectivity. However, according to Jeremy Hunsinger, a professor at Virginia Polytechnic Institute and State University, in Blacksburg, who has contributed articles on political science topics, this faith is sometimes misguided. Hunsinger has witnessed members of the Wikipedia establishment overlook the contributions of established authorities--the kinds of experts World Book and Britannica hire to evaluate their content. "You can give your academic credentials, but that doesn't necessarily travel very far in the Wikipedia culture," Hunsinger says. "The administration is a tight community that's had a lot of the same members for years, and it can be hard for new people to break in."

Still, many users and contributors agree that the system works well, if not perfectly, in practice. Seamon was pleasantly surprised at the administrators' ability to show restraint and vigilance by turns. "I was fearful that anything I'd write would be tampered with and ruined, but that's never happened," he says. "Sometimes people will add stupid stuff, almost the equivalent of graffiti, but before I know it, it's been removed." And for those who assume that Wikipedia's policies translate into general inaccuracy, in a head-to-head comparison of Wikipedia and Britannica in the journal Nature last year, Britannica had an average of three errors per published science article, while Wikipedia had four--a difference so slight it left the primacy of Britannica's venerated review process in question. (Independent reviews of Wikipedia's nonscientific articles have not yet been conducted. At press time, Britannica is demanding that Nature retract the article on the grounds of "sloppiness [and] indifference to basic scholarly standards." Nature has retorted that it is standing by the article and that the comparisons drawn are "fair.")

That's not to say Wikipedia users should ever feel so confident as to take the encyclopedia's content on faith. Wales advises readers to check their online finds against other sources and to be aware of Wikipedia's unique strengths and weaknesses, especially when gathering information for research projects. "No encyclopedia article is intended to be a primary source--it's just an introductory summary, and people should approach it that way," he says. "Wikipedia's timeliness is really impressive, and so is the sheer amount of brainpower we bring to bear on complicated questions. But because everything is so open and fluid, you have to be aware that anything on the site could be broken at any given moment. It's a live work in progress."