Reading about danah boyd's recent O'Reilly Gov 2.0 Expo talk on "Information literacy" has finally inspired me to kick off a project I've been talking about for a while now.
I'm an ex-wikifiddler because the politics drove me away, but I still believe in the ideal of a crowd-sourced Big Book of All World Knowledge. The real problem with Wikipedia (politics aside) is, of course, bias. It's inevitable, inescapable. But then, isn't all meaning subject to bias? And of course, as any good postmodernist knows, listener bias is as critical as speaker bias. To confront our own biases and fully understand information, we need to understand the provenance of that information. In Wikipedia's case, all the content comes from its editors. If we could know more about the biases of those individuals, we'd be in a better position to understand and evaluate what they've written, and Wikipedia's knowledge would be significantly more valuable to us.
One of the great strengths of a wiki is that it does identify the source of all its content, and stores details of all changes to that content in a perpetual history. At least in the case of Mediawiki—the open source software which runs Wikipedia—the traditional mechanism for accessing that detail is not conducive to casual investigation. Nevertheless the history of each claim, phrase or word can be made apparent to the reader, although the data manipulation required is non-trivial.
There's obviously a lot (an awful lot) of design and development work between inspiration and completion on this thing, and it's a project I'm fairly certain I won't ever complete on my own. Still, here's a first step.