Wikipedia's Woes

WSJ.com - Real Time :: Tumultuous Weeks for Internet Encyclopedia Bring Furor Over Anonymity, Accountability - By JASON FRY : "First the online, volunteer-created encyclopedia was blasted by a retired newspaper editor who found a phony bio about himself in its listings. That controversy soon became a furor in which Wikipedia's entire workings were put on trial. While the debate was still going on, Wikipedia then got an unexpected boost from a venerable science publication that said its scientific entries stacked up well against those of Encyclopedia Britannica, the standard Wikipedia's founder has said is his goal.

Not enough? Throw in an Internet detective story and a cast of characters worthy of a pretty fair novel: The wronged editor is a respected free-speech advocate; the Net sleuth turns out to be a fierce Wikipedia critic; and the author of the phony bio is revealed as a prankster with a fair bit of Southern gentleman in him.

The furor began with a Nov. 29 article in USA Today by John Seigenthaler, an administrative assistant to Robert Kennedy in the early 1960s and the founder of the Freedom Forum First Amendment Center at Vanderbilt University. Mr. Seigenthaler had a tale to tell: He'd discovered that a biographical entry about him on Wikipedia said that "[f]or a brief time, he was thought to have been directly involved in the Kennedy assassinations" and added that he moved to the Soviet Union in 1971.

Mr. Seigenthaler related his discussions with Jimmy Wales, Wikipedia's founder, and his anger at finding himself the victim of hit-and-run character assassination that rode high in Google search results for his name for more than four months. He also wrote about his frustrations trying to discover the identity of the entry's author -- a quest that seemed destined to end with an IP number maintained by Bell South.

"Naturally, I want to unmask my 'biographer,' " he wrote. "And, I am interested in letting many people know that Wikipedia is a flawed and irresponsible research tool."

Mr. Seigenthaler seemed genuinely pained by the push and pull between the freedom of speech he's long championed and the lack of accountability on Wikipedia. That helped an already-irresistible story get dissected endlessly in newspapers and magazines and on blogs, message boards and TV. Then the tale got better: Wikipedia critic Daniel Brandt traced the IP address Mr. Seigenthaler had found to a Nashville delivery company. (He tells News.com's Daniel Terdiman how he did it here. For more on Mr. Brandt, check out his Wikipedia Watch site, which is by turns kooky and thought-provoking, and this 2002 Salon article.) Mr. Brandt's efforts led Brian Chase, the delivery firm's operations manager, to resign and hand-deliver a letter of apology to Mr. Seigenthaler. Mr. Chase said he thought Wikipedia was a gag site, and concocted the phony bio as a prank on a co-worker. "I am truly sorry to have offended you, sir," his letter to Mr. Seigenthaler said. "Whatever fame comes to me from this will be ill-gotten indeed." While it in no way excuses what he did, Mr. Chase's gracious apology is worthy of note in our graceless times.

And there was still another act to be played. Along came Nature with a very interesting study: The journal had experts in various fields review 42 relevant encyclopedia entries from Wikipedia and Encyclopedia Britannica, without telling them the source. The experts found about four inaccuracies in each Wikipedia entry -- but Britannica had about three inaccuracies per entry. Mr. Wales has long talked of meeting Britannica standards, but the Nature article suggests those standards aren't quite what we thought they were.

So, whew. Where to begin?

The Seigenthaler affair feels like a miniature version of innumerable debates about the wider Internet. One minute Wikipedia is lauded as revolutionary: a mutating encyclopedia that harnesses the spookily powerful process by which consensus and wisdom emerges from countless individual efforts. The next minute it's a troubling symptom of our digital age: a vehicle for anonymous assaults and a dangerously out-of-control echo chamber for dispute and error.

At the heart of this is the fact that at a very basic level, Wikipedia is out of control. There's next to nothing stopping me, you or anybody else from vandalizing a Wikipedia entry the way Mr. Chase did -- or from editing one in good faith and accidentally introducing mistakes. But too many accounts of the Seigenthaler affair missed that Wikipedia is self-correcting to a remarkable degree. The phony Seigenthaler bio stayed uncorrected (and most likely unread) for so long because he's relatively obscure; a prank entry about Bill Gates -- or Gollum -- would most likely have been fixed within minutes. (This is neither a claim that what happened to Mr. Seigenthaler is of no import nor an endorsement of the nonsensical digerati claim that he was somehow at fault for not simply fixing it himself.) The important thing to note is that this self-correcting mechanism also depends on Wikipedia's being fundamentally out of control. Whether you find that basic truth invigorating or frightening is something of a Rorschach test for how you feel about the Internet as a whole.

This isn't to say that Wikipedia doesn't need better defenses against vandals. Earlier this month, the encyclopedia began requiring users to register before they create articles, and it's mulling new ways of verifying articles' accuracy. But unregistered users can still edit articles, and registering takes less than a minute, no email address required. Mr. Wales has defended anonymity by noting it's extremely easy to conjure up a fake identity and pointing out that anonymity isn't always sinister. The latter is a fair point: Expanding a Wikipedia entry about China and Tibet is a hobby if you're American, but a dangerous business if you're Chinese.

That said, it seems to me that Wikipedia could take steps against vandals and trolls without sacrificing anonymity. If combined with Wikipedia's existing methods for banning troublemakers and policing entries, making all users register and requiring confirmation from a valid email address for that registration seems like a worthy step that would make repeat vandalism not worth the hassle to most troublemakers, while preserving anonymity and keeping registration from being too much of a chore.

But there's one thing that's disturbing about Wikipedia that no amount of tweaking can resolve -- and, not surprisingly, it's something disturbing and impossible to solve about the larger Net as well. And that's the fact that we put far too much faith in technology. Even though we've heard innumerable times that the Web has let anyone be a publisher, too often we treat print on a screen as if it's been through the quality-control processes behind books, magazines or newspapers -- not that such publications are anything close to infallible. To this oft-unconscious bias, add in the magic of search engines that spit out dizzying amounts of information in response to our silliest query. A Web search feels like we're doing something complicated and rigorous, and when what comes back is a well-designed Web page or Wikipedia entry, it feels like we've done something right. Arriving at such a page should just be the start of assessing the information we're given and deciding how much to trust it, but too often we treat it as the finish line.

It'll be interesting to see if another decade or two makes us more skeptical of online information, or less. In the meantime, we should bear in mind Mr. Wales' warning that students and researchers shouldn't cite Wikipedia. But we should also note that he said that as part of a larger warning: "People shouldn't be citing encyclopedias in the first place." Which is a warning I remember from teachers before any of us had heard of the Internet. It was true then and it's true now -- but that doesn't mean encyclopedias aren't valuable tools. Particularly an encyclopedia that's self-correcting, keeps up with current events, provides links to more-basic sources and is already bigger than any encyclopedia has ever been. Check out Wikipedia's entries on deconstructionism, the history of records, or the Replacements, then consider that those entries and some 868,000 others are the work of volunteers. I find that impressive -- revolutionary, even. And if you follow those links and find something you know is wrong, fix it.

As has been done with the entry for John Seigenthaler."