Long-term Wikipedia self-promotion exposed

Update: we will of course be covering this story in the next episode of Wikipedia Weekly.
Another update: I’ve removed the claim that the second user was a sockpuppet, based on Psychonaut‘s opinion on the matter.
Update 3: Welcome, slashdot! Please read the opposing view!

Wikipedia, as everyone knows, can be vulnerable to certain types of activity that compromise its integrity. As a collaboration between tens of thousands of ordinary people, without a rigorous credential policy like the current pilot of Citizendium. While on the one hand that has enabled Wikipedia to grow to its amazing size today, and the combined forces of AntiVandalBot and the Recent Change Patrollers generally keep vandalism to a minimum (although not entirely without issue), there are some kinds of vandalism that people simply aren’t prepared for.

The article on “NPA personality theory”, a theory developed by retired physician Anthony M. Benis, was recently proposed for deletion. The move came as something as a surprise, since the article had previous been listed as a Good Article, one of the best 0.2% of all articles, and one of just six in the field of psychology. Now, it was accused of being an example of long-standing and accepted self-promotion at the website, having been largely written by a user called “ABenis” — Anthony M. Benis himself.

Self-promotion
The article was originally listed as a Good Article some months ago, thanks to an extremely effective campaign of astroturfing and deception. Not only was the article written almost entirely by Benis himself (and an associated user, named “D-katana”), but it took advantage of the fact that Good Article and peer reviewers aren’t really supposed to ensure that the articles they are asked to look at aren’t vandalism. The assumption is that, if they make it as far as GA review, they’re probably quite good.

Unfortunately, the problems run deeper than simply taking advantage of the system. At several points during the course of the article’s creation at Wikipedia, users expressed worrying attitudes towards NPA personality theory and its article. To quote from the AfD discussion:

One editor in particular expressed rather directly that his or her intention was to use Wikipedia to promote the theory, as evidenced by this statement: “And, in turn, Wikipedia has the honor of the recognition for championing the theory before any other group of scholars took it on for further development and propagation.” —Cswrye

The article, and the one on its author (both still kept in Google’s cache, if you’re interested) remained in Wikipedia and even rose to great heights entirely on the work of a single editor — who happened to be creator of the theory.

And they remained there for months, and months, and months:

This is precisely one of the things that troubles me about Wikipedia even more than reg’lar spam — bogus knowledge slipped in between the cracks and woven into the article matrix. Yikes. —Dhartung

I’m not sure what there can be done with this. The fault is not with any of the users who were asked to review the article, since their work does not involve checking for notability, verifiability or hoaxes; they make good-faith efforts to improve the article as requested. The fault doesn’t lie with any of the newpage patrollers, who had just one chance to catch the article upon creation and were not in a position to make a snap decision on its accuracy.

That said, the blame doesn’t even entirely fall with Benis: while he certainly violated Wikipedia’s vanity guideline (and common sense) and the GNU FDL under which Wikipedia content is licensed (reproducing the article without reference or the terms of the license at npatheory.com), he wasn’t even the only editor involved, nor was he — apparently — the one that instigated the article’s “improvement.” These things happen, and it’s certainly not possible to find a single point of failure.

The fault here lies partially with the structure of Wikipedia, too. The website’s openness both allows an enormous amount of vandalism to filter through every day, not all of which can be removed within minutes, and relies on ordinary people to catch hoaxes like this and remove them.

In this case, the articles were caught by a professional scientist after one of them had been a Good Article since 29 May 2006 — and while it’s comforting to think that this is now gone, it’s a little unsettling to know that we might have missed more.