Wikipedia Ain't Broke, but Needs Fixing
By Chris Maxcer
03/09/07 7:48 AM PT
With Wikipedia under fire yet again -- this time for allowing edits from a user who lied about his academic credentials -- can the online encyclopedia remain true to its commitment of user-generated content? Wikimedia's advisory board members say the wiki model isn't broken. However, the incident shows the perils of allowing anonymity and accepting one's credentials without question.
What's the big deal with having a false persona? We live in a world of fake names, and "World of Warcraft" and Second Life are only part of it. Want to participate in an Internet discussion forum? You'll have to create a user ID, and most likely it won't have your first and last name in it.
What about MySpace ? Does anyone believe all those MySpace pages represent real personalities? Or do they represent the personality that the owners simply want for themselves? You know, the one they adopt and project to the world and hopefully get their "friends" to reinforce?
How different are these false identities from the anonymous identities assumed by thousands of Wikipedia contributors around the world? How different are they from the "EssJay" persona adopted by Ryan Jordan, the community college dropout who claimed to be a tenured professor at a private university?
Wikipedia's Trust Reliance
It turns out there is a world of difference, and it's based on the simple social concept of trust. Users of Wikipedia trust that anonymous contributors will only add true and accurate knowledge to the free online encyclopedia.
However, not everyone is trustworthy. Contributors have made mistakes, of course, and then there's the widely-reported inconsistencies and outright vandalism of entries.
How do you do trust someone you don't know?
Credentials go a long way to build trust. The first time you visit a new doctor's office, you don't know the quality of the staff, but you trust that if they are indeed doctors then they've been officially trained. To back up their authority, doctors sometimes even post framed certificates on their office walls.
All of this leads to the reason EssJay most likely claimed to be a professor. His credentials gave his Wikipedia entries weight, and this appears to be the key that has caused all the uproar.
Wikipedia co-founder Jimmy Wales' initial response to the news of Jordan's invented persona was that he regarded it as a pseudonym and didn't really have a problem with it, according to Wikipedia (note the irony of citing Wikipedia here).
"As a result of the controversy," the key "EssJay controversy" Wikipedia entry states, "Wikipedia users began a review of EssJay's previous edits and discovered evidence he had relied upon his fictional professorship to influence editorial consideration of edits he made."
Want Wiki? Disallow Anonymity
The solution of disallowing anonymity makes it difficult for most people to outright lie or conduct revision wars on wiki content. In fact, Gartner (NYSE: IT) , in its report, "What We Learned From the Wikipedia Experience," recommends that IT professionals looking at implementing wiki models in their organizations start by disallowing anonymity.
Gartner also noted that, despite Wikipedia's current success, "Anonymity is contrary to success -- for example, the Los Angeles Times' wikitorial system failed in 2005 partly because it allowed anonymous posts. It's difficult to exert social control when participants aren't identified, and it's difficult to assume that contributors will act within the boundaries of social constraints if they don't have to identify themselves."
To help direct the beleaguered Wikipedia, the Wikimedia Foundation, which essentially runs Wikipedia, started up an advisory board drawn from experts among its board of trustees. This announcement came in January, however, before the EssJay debacle hit.
Having It Both Ways
If Wales wants contributors to remain anonymous unless they cite professional credentials, it begs the question, how will this fundamental change affect Wikipedia's commitment to being a user-generated and user-monitored site?
Moreover, how will a loose federation of volunteers successfully manage a global project?
"The core problem is accountability and quality control. The fact they didn't even do a background check when they hired [Jordan] suggests severe quality problems," Rob Enderle, president and principal analyst at the Enderle Group, told LinuxInsider.
"I think this points to an overall flaw in the thinking -- while people are certainly willing to contribute, too many are either motivated to falsify the information positively or negatively, or they simply don't do good work. That means you need a very strong quality control process, and Wikipedia simply doesn't have that," Enderle explained. "Without it you'll eventually lose the trust of your audience and become kind of a joke."
No Laughing Matter
There is a breach of trust when someone misrepresents himself as having credentials and then having those credentials count in arguments with others, which is why EssJay was asked to resign, said Jay Rosen, associate professor of New York University's Department of Journalism, founder of NewAssignment.Net and author of PressThink and member of the Wikimedia Advisory Board.
That misuse of credentials only tangentially attacks the idea of user-driven content. "If I saw any movement at Wikipedia that threatened its basic commitment to open participation and 'edit this page,' then I would speak up about it," Rosen told LinuxInsider. "Others on the advisory board, which is new, would speak, too. It's not a perfect system but I think it can work."
In the long run, Mark Pellegrini, a Wikipedia administrator, bureaucrat, arbitrator, and PhD candidate in electrical and computer engineering at the University of Delaware -- according to his Wikipedia page -- doesn't believe that the EssJay debacle will have an affect on Wikipedia's commitment to remain user generated.
"Simply put, we think the system works pretty well," Pellegrini told LinuxInsider. "The basic premise upon which Wikipedia is built -- that many people participating in the process tends to make our articles more comprehensive and accurate -- is now born out by multiple studies."
The most recent study, conducted by Dennis M. Wilkinson and Bernardo A. Huberman at HP Labs, found that overall quality, particularly on highly visible topics, was quite good.