Monday, December 12, 2005

Wikipedia needs safeguards that work

The free, online encyclopedia Wikipedia is widely used, but also increasingly criticized. I personally use it and will think it ok to put up a link to a post there about, say, Julius Caesar. However, it gets a lot weaker when you move closer to contemporary and politically sensitive issues. That's why Faith Freedom International has decided to make their own, online encyclopedia about Islam, for instance:

Wikipedia needs safeguards that work

Jimmy Wales has often described himself as the constitutional monarch of Wikipedia, the free online encyclopedia written and edited by volunteers. Queen Elizabeth II of England and other constitutional monarchs can only intervene in affairs of state on rare occasions of great crisis. Wales, who co-founded the non-profit Wikipedia in 2001, exercised his royal power last week. But he only took a tiny step in the midst of a crisis where bolder leadership is required. They're not enough to resolve Wikipedia's fundamental dilemma: It can't meet what Wales calls the project's primary goal -- producing ``a free, high-quality encyclopedia'' -- while also clinging to the utopian concept that anyone can contribute without restrictions. How this crisis plays out will reverberate across the emerging landscape of ``social media,'' where loose groups -- such as Wikipedia's volunteer contributors -- come together through the Web to create news, community forums and information.

The result of Wikipedia's open editing system is predictable: Most contributors provide useful material, while a small number of ``trolls'' repeatedly deface the encyclopedia. Wikipedia is also plagued by endless ``revert wars,'' where dueling groups keep reversing each other's changes to controversial articles. This undermines the credibility of Wikipedia, which now offers an unprecedented 857,000 articles in English, along with versions in more than 100 other languages. Wikipedia is becoming a first reference stop for millions of people, from schoolchildren to journalists, including me. But many of these users don't realize a small percentage of articles are flawed. Even more troubling, there's no way to know when you've hit one of those defective entries. That's why I never put a fact from Wikipedia into one of my columns without first double-checking it elsewhere.

I have a suggestion: Wales should issue a royal decree moving Wikipedia to a ``gatekeeper'' model, borrowed from successful open-source software projects such as the Linux operating system and the Firefox browser. These projects are administered by networks of trusted volunteers who carefully review additions and changes before they are made, and there's a hierarchy to resolve disputes. Wikipedia is now big enough, with a core group of 13,000 active volunteers, to pre-screen all of its contents. New entries and edits could still be submitted -- even anonymously -- by any visitor to the Wikipedia site but would be placed in a kind of holding pen until one of the trusted volunteers took a look and said OK.

Bjørn Stærk: Why we need professional encyclopedias

Quality is holistic, it requires understanding. Getting individual facts right isn't good enough. There is no gradual slope of quality as there is with factual accuracy, only a long plain of mediocricity followed by a sharp rise at the end. Britannica is at the top of that rise because it is written by individuals who know how to write, who know their subject, and know how to present it clearly in few words. Wikipedia is at the bottom of the same rise because it is written by committee. Britannica is consistently good. Wikipedia is good only by accident. Britannica gives experts the freedom to do their job, Wikipedia drowns their experts in tedious committee work. Wikipedia should perhaps emulate the free software movement. A free software system like Linux is not written by any Joe R. Newbie with a text editor. Linux is controlled by an elite of brilliant hackers. They decide which contributions to accept, and which to reject. They ensure that the whole isn't sacrificed in favor of the parts, and they decide when to rewrite code that has deteriorated in a larger sense than just containing individual bugs. This way of working actually works, and should be emulated by Wikipedia. Let anyone make a contribution, but use experts to guard the gates.


At December 12, 2005 3:23 PM, Blogger talnik said...

Back in the days when I was a hippie punk snotty jerk, my favorite band was The Doors.I tried to learn everything I could about the band in general and Jim Morrison in particular. He, of course, croaked in Gay Paree in 1971, so it was with great anticipation I awaited the delivery of the yearly update of our Compton's encyclopedia.

Compton's obit on Morrison was so inaccurate (stating, among other things, that he wrote "Light My Fire" and "Touch Me") that I thereafter took everything I read with a grain of salt. It made me aware of the fact that reporters (and encyclopedia editors) aren't always knowledgeable about the subjects they cover.

In the intervening years, this fact has been reinforced for me daily.

I guess my point is that "real" encyclopedias might not be all that much more accurate than Wikipedia.

At December 13, 2005 8:02 AM, Blogger Mike H. said...

Wikipedia⇒Peer review.

BTW, for Bjorn, Linux is written by anyone that wants to write a diff or a driver. It has to undergo a review by the individual and others in charge of the module that it fits into.

At December 13, 2005 7:52 PM, Blogger Bjoern said...

"BTW, for Bjorn, Linux is written by anyone that wants to write a diff or a driver."

Depends on what you mean by Linux. What I was referring to here was the kernel, but even the free software world as a whole is full of gatekeepers. Projects have official or unofficial owners, a small number of people often do most of the hard work and make the important decisions, the distros have high standards for what they'll include, etc. Free software works because there are a lot of good programmers who make decisions about what they'll accept in projects they're responsible for .. and that's what Wikipedia should emulate. It's not enough to let people contribute, you need a way to extract quality from those contributions.

- Bjørn Stærk

At December 16, 2005 7:35 AM, Blogger ik said...

Wikipedia vs Encyclopedia Britannica

Jimmy Wales' Wikipedia comes close to Britannica in terms of the accuracy of its science entries, a Nature investigation finds.

Wikipedia's Accuracy Compared to Britannica

Nature magazine recently conducted a head-to-head competition between Wikipedia and Britannica, having experts compare 42 science-related articles. The result was that Wikipedia had about 4 errors per article, while Britannica had about 3. However, a pair of endevouring Wikipedians dug a little deeper and discovered that the Wikipedia articles in the sample were, on average, 2.6 times longer than Britannica's - meaning Wikipedia has an error rate far less than Britannica's."


Post a Comment

<< Home