May 31, 2013

Wikipedia: We Should Have Known

Wikipedia can’t get out of its own way. Crowds may be wise, but they are also clumsy.

Recently, Wikipedia’s User:Qworty was identified as the novelist Robert Clark Young. Young had spent at least eight years surreptitiously enhancing his reputation and exacting Wikipedia revenge on his rivals. A group at the Web site “Wikipediocracy” unmasked the deception.

Wikipedia’s response: the Wikipedia page for Wikipediocracy has been nominated for deletion because it’s insufficiently notable. The debate has already passed the 11,000 word barrier. And over on the Wikipediocracy page, people are seriously planning to plant stories in the press that are critical of Wikipediocracy, in order to justify a Criticism section on the page should it survive the deletion debate.

Of course, the optics of this are horrible: Wikipedians, embarrassed by a revelation, seem eager to punish their critics by deleting their page. The impulse is natural enough, I suppose. Other large organizations might flirt with a response like that for a minute or two until cooler heads (or professional advisors) pointed out how petty this would appear. In Wikipedia, everyone is free to contribute to making the project look mean and foolish.

Wikipedia was intended to be written collaboratively by the people who use it. Its underlying assumption is that, on average, people are fairly honest and fairly smart.

In some corners, that’s worked well. But these are precisely the corners where USENET worked: specialized and esoteric topics pursued by serious communities that already possess effective disciplinary mechanisms.

Wikipedia’s coverage of abstract algebra is reportedly fairly good. The article on Nero is extensive and detailed, though it relies far too heavily on Suetonius and makes little effort to weigh recent scholarship. But if you ask a random graduate student for a simple but technical opinion in their field, you’ll likely receive a sane and sensible answer. Who else is going to answer questions about Hermite Polynomials or Julio-Claudian policy?

When we move to more accessible areas, things become more difficult. We now know, for example, that Robert Clark Young systematically ensconced himself in Wikipedia to enhance his reputation and to diminish his rivals. He seized on a minor flap about his one novel in order to insert himself into national controversies. Writing on the Wikipedia page of the National Endowment for the Arts on 22 January 2005, he mentioned recent controversy:

However, in 1996, Congress slashed NEA funding to $99.5 million (see Chronology of Federal Support to the NEA) as a result of increasing pressure from right-wing groups such as the American Family Association, who have criticized the agency for funding artists as diverse as Robert Clark Young, Andres Serrano, and Robert Mapplethorpe.

That’s a diverse list indeed: it contains two famous photographers and a very obscure novelist. It was almost certainly inserted by that novelist, and it stayed in in the lede for the NEA’s article for more than eight years. During that time, the author became User:Qworty, a dedicated and prolific wiki-lawyer, editing thousands of pages and consigning many more to the dustheap of deletion.

Qworty was effective because he was dedicated. He had lots of time on his hands and lots of scores to settle. Eventually, almost by accident, he was caught and banned.

Wikipedia ties itself into hoops in order to keep axe-grinders like Qworty at bay. Everything has to be attached to a Reliable Source, and that source must almost always be tied to print. That something is true is no defense, and if something is false but supported by a “reliable source” it can be incredibly hard to correct. A byzantine array of rules and review boards and disciplinarians tries to manage all this, but it’s never been very effective and is perpetually exposed to the risk that the cranks and Pajamas will capture the apparatus.

A community that tries to govern itself through consensus must have a mechanism for addressing those who, through error or stubbornness, cannot or will not accept any consensus. The Wikipedia of myth has Jimmy Wales, but Jimmy hasn’t scaled and Wikipedia has nothing beyond the vague hope that people of good will can outsmart the fools. Unfortunately, in the Qworty affair and many parallel situations, one side of the argument employs zealous ideologues who have years of experience manipulating wiki-law. The other side may be right, but in the new Wiki way, that’s beside the point. The institutional structures favor zealotry over good will and advantage unemployed cranks (who have all the time in the world) over sensible people (who have other things to do).

We can wreck the long tail in precisely this way; this is precisely the scenario for destroying the Web that I identified in my 2011 Web Science paper. Drive lots of traffic to wikipedia, and let a million weblogs wither. Then, let a scandal (or chance) dent Wikipedia and all that’s left are the old broadcast networks and the cable companies and the spammers and the Government of Syria.

Indeed, this is one reason Facebook has been out-competing the open Web: for all its flaws, Facebook does have a good mechanism for letting you check up on your nieces and nephews without having the trolls and spammers and Pajamas in your face. Until Google and Bing turn down the traffic to Wikipedia, Wikipedia is unlikely to change. If we wait for a major scandal to occasion this, the damage to Wikipedia may be irrecoverable.While we wait, the damage to the Real Web continues daily.