The Epistemology of Mass Collaboration: Wikipedia as a source of knowledge
Back in 2009 the philosophy journal, Episteme, published an issue focussing on Wikipedia as a form of mass collaboration. The volume is no longer available online. However, Wikipedia published an extensive review of the four articles that dealt with Wikipedia, including one by Larry Sanger, founder of Wikipedia, entitled “The Fate of Expertise after WIKIPEDIA.”
The other three articles are “Wikipedia and the Epistemology of Testimony” by University of Memphis philosopher Deborah Perron Tollefsen, “The Epistemic Cultures of Science and Wikipedia: A Comparison” by State University of New York philosopher, K. Brad Wray, and “On Trusting Wikipedia” by another State University of New York philosopher, P.D. Magnus (who has the greatest personal website I’ve seen).
Tollefsen argues that mature Wikipedia articles, that is, those that have had a number of contributors and have reached a stable state of agreement, can be considered as representing a form of group testimony. Group testimony can be fundamentally different from the testimony of any one of the individuals who contribute, which makes it a fundamentally different form of knowledge. Nothing startling there for anyone who has facilitated any kind of group learning exercise. The end is always very different from the beginning that any single person in the room could have imagined. However, and this is the bit I don’t agree with, she argues that Wikipedia is an “immature epistemic agent”, the claims of which should be interrogated like that of a child. While I agree that everything we read on Wikipedia needs to be interrogated, the same can be said of any source. The beauty of mature group testimony written by many rather one is that no-one owns history. That seems to me to be a particularly powerful and wonderful thing.
Wray adopts the old economics argument posited by Smith several hundred years back, that Wikipedia may be trustworthy because of an “invisible hand“. In the same way that Darwinian market forces kills of the weak, poorly executed articles on Wikipedia as also quickly edited to such an extent that the final product is unrecognisable from the starting point. There are two arguments against this position that immediately spring to mind: First, it assumes that any else out there in the world actually cares enough about a subject to make editorial changes; Secondly, it assumes that people are not malicious or self-interested in their editing. Neither is a safe assumption in my estimation.
Strangely, Sanger goes straight into bat for another of projects, Citizendium (which I must admit, I had never heard of, there’s a good description on, you guessed it, Wikipedia). Suggesting that it might offer a happy balance between more traditional privileged knowledge sources (like the journal in which the article appeared) and contemporary online wikis. Citizendium has different rules regarding editing rights and responsibilities that privileges academics. It would seem to suggest, that Sanger himself, has concerns about the veracity of the Wikipedia model as a global knowledge base.
Finally, Magnus, explores strategies for determining article quality on Wikipedia; which, given the know variability seems like a pretty good idea. His five strategies are to look at (1) the authority of the author, (2) the plausibility of the writing style, (3) the plausibility of the content, (4) calibrating a subset of claims against a secondary source, and (5) check one claim against multiple other sources. Seems a lot of work to me!