Friday, December 16, 2005

The Wikipedia War

A recent high-profile dispute over the user-written encyclopedia's veracity has the site rethinking some of its rules.

By Eric Hellweg

In 1728, Ephraim Chambers, a mapmaker in London, published the Cyclopaedia -- the first encyclopedia to include cross references. Almost 275 years later, Wikipedia, the online, user-created encyclopedia, is championed for its volunteer model and its elaborate system of cross-referencing and verifications. But the Wikipedia model has had a rough couple of weeks -- ever since the mainstream media jumped on the story of Robert F. Kennedy confidante John Seigenthaler’s outrage at false information posted in his Wikipedia entry.

The concept of Wikipedia -- and the general idea of "wikis," or user-edited Web pages -- is notable because it embodies one of the key cultural movements of the decade. Called Web 2.0, it's the notion that the Internet should be fueled by user-generated content -- blogs, wikis, podcasts, and user-created add-ons to programs such as Google Maps and Firefox. The desire of people to contribute online content is as old as the medium itself; now software improvements and broadband adoption make it easier than ever. With the tools for content creation in the hands of single individuals, not just companies and institutions, Web 2.0 means a more democratic Internet, and a truer voice, or at least a broader one.

“When people think of Web 2.0, they’re thinking of Wikipedia,” says Larry Sanger, a co-founder of Wikipedia, who's no longer with the enterprise and currently involved in a new online encyclopedia project.

But the Web 2.0 concept, and Wikipedia itself, still have some growing up to do. In part because of the Seigenthaler flap, Wikipedia is going through a very public coming-of-age. “The last two weeks, it’s been hard to get any work done,” says Jimmy Wales, founder of Wikipedia and president of the non-profit Wikimedia Foundation.

At the heart of the issue is a problem that many startups would kill to have: overwhelming interest in their product. In 2001, Sanger predicted that, with a little luck, Wikipedia would have 80,000 articles by 2008. The site currently has more than 850,000 in English alone. Throw in 200 other world languages, and the total rises to 3.7 million.

Wikipedia was founded with a radical premise: thousands of enthusiastic volunteers would scan the entries regularly, ensuring that they remained truthful -- or at least not erroneous for very long. In addition to being reviewed by the general public, Wikipedia articles would also be vetted by administrators and “page watchers.”

That's exactly what has come about. But with the huge proliferation of Wikipedia entries, as well as the built-in ability for anyone to contribute to the site, the occasional vandal has appeared -- and gone undetected for too long.

Evidently, the motive for encyclo-vandalism is sometimes just humor. According to a report in the New York Times, the contributor who entered false information about Seigenthaler (claiming he'd played a role in Robert Kennedy's assassination) was playing a joke on some friends. Since then, the perpetrator, Brian Chase, who was ferreted out, has apologized to Seigenthaler, who accepted his apology.

I tested the Wikipedia correction process while reporting for this article. After logging on, without giving an e-mail address, I edited the entry dedicated to musician Tom Waits. In a section on the artist in the 1990s, I wrote that Waits had played a concert with Elvis Costello, Elvis Presley, and Mr. Ed (the talking horse). Within 24 hours, the Presley and Mr. Ed references were removed, but the Elvis Costello citation -- also false, but not as glaringly so -- remained.

There are real, perhaps inevitable weaknesses in Wikipedia's system. But the Siegenthal ordeal seems to have unleashed a disproportionate number of Wikipedia critics. News site Official Wire, run by Baou.com, is now posting stories that allege “Nazi-style behavior” among Wikipedia contributors and editors.

And this week, a site called WikipediaClassAction.org went live, soliciting feedback and, more significantly, instances of monetary damages caused by Wikipedia. Their goal is to launch a class-action lawsuit against the site. When I called the phone number listed on that site, the person who answered refused to give his name, then rattled off a long series of allegations against Wikipedia. The charges felt specious to me, and were quite vitriolic.

A quick piece of sleuthing turned up a likely explanation: Baou.com also runs an organization called QuakeAid, the Wikipedia entry for which cites some questionable circumstances surrounding the organization and its founders. Furthermore, some of the anti-Wikipedia articles found on Official Wire are written by “Jennifer Monroe,” the same name listed as having registered the domain WikipediaClassAction.org.

Although Baou's actions imply a multi-pronged revenge campaign, some anti-Wikipedians appear to have more reasonable complaints. Daniel Brandt, the man behind wikipedia-watch.org (and also Google-watch.org), says that until Wikipedia drops its policy of allowing anonymous posts and edits, the quality of the site will suffer. “For research purposes you ought to be able to find [authors],” he says.

But Brandt, too, has a personal reason to be upset with Wikipedia. He admits his opposition to the site came after he learned that it included a page about him with links he considered unflattering. Brandt was a prominent draft resister in the 1960s.

Now Wikipedia president Wales has enacted a policy that requires users to log in before creating new articles. However, since there’s no e-mail address required to do so, anyone can make up a name and create an article without a way to verify their identity.

In any case, Wales insists that the vast majority of the articles on the site are correct, and that anonymity isn’t the issue. Still, he's working on other measures aimed at eliminating the possibility of false information being added.

One is a “holding zone,” where contentious information or topics -- those prone to vandalization -- can be queued for review before going live. Another is a community-based rating system, scheduled to go live on January 1, 2006. Wales says he’s also considering soliciting experts to submit ratings on articles, to see how those ratings jibe with the community’s.

A couple of years ago, Wales suggested to the Wikipedia board that they adopt a “real name” policy, similar to Amazon.com’s review system. At Amazon, anyone can review a book; but after the site was hit with allegations that authors were writing glowing reviews of their own books and slamming competitors’ works, Amazon decided that giving people the option to use their real names -- and having Amazon certify it with a “Real Name” logo beneath it -- would lend credibility. The Wikipedia board rejected the idea a few years ago; but today Wales thinks “anything’s possible.”

Others with experience allowing masses of people to create content say some policing is necessary to prevent chaos and to ensure the validity of information. Today, Rich Skrenta is CEO of Topix.net, a news-gathering site that just launched a citizen’s journalism effort, where visitors can write news stories that pertain to their interests. Prior to founding Topix, though, Skrenta was a founder of DMOZ.com, a community-created site directory, where 60,000 "editors" index sites and write site descriptions.

“We had issues where editors were being bribed to write good reviews and spammers tried to get in, so we had to develop a fairly sophisticated set of policies to keep this running,” Skrenta says. “With our new project we have both social and technical designs to facilitate good information. They’re patrolled, monitored, and managed.”

Would a stricter log-in process have stopped me from falsifying the Tom Waits page? Probably not; I was under deadline and knew I'd change it back. But for the few visitors who come to the site with a casual ill intent, it might.

Skrenta puts the Wikipedia situation into some perspective: “Our citizen journalism effort doesn’t have any problems yet because it doesn’t have any traffic. When people are corrupting entries, that’s a problem you only have when you’ve succeeded.”

If Wikipedia wants to become the trusted, open-source repository of all the world’s knowledge -- and it probably does -- it may have to adopt some of the successful tactics of Amazon and others who have gone before it.