This is the last of a three-part post on the Wikitrust project and what it implies about the way information will be socially organised in the future. See the first post, “Taking Your Word” for Wikitrust’s aims and the second, “Distributed Expertise”, for the methods behind WikiTrust’s attempt to crowdsource trustworthiness. Today I’m going to try to analyse what Wikitrust implies for the future of the encyclopedia and for information-sharing as a whole.
Wikipedia’s editing culture has come into a lot of criticism for its Darwinian self-organisation. The status of “super-editors” has prompted cries about the stagnation and death of Wikipedia. But “super-editor” status is popular status: they are tribunes, not Caesars. Authority on Wikipedia is based on the intangibles of social currency: a network of respect and other reputation relationships.
On familiar vanilla Wikipedia an editor’s reputation had to be earned, at least in theory, through regular, reasoned and courteous engagement with other editors on the discussion page. Of course this system had its tinpot notability czars, but the constant debate over the best truth was what made Wikipedia so enormously successful.
So is Wikitrust going to take the humanity out of the system?
The very helpful and all-round classy guy Bo Adler (@thumper17) who works on the WikiTrust Project took my cynical suggestions of how their system might be “gamed” with good grace, and pointed out where such attacks have been anticipated. For instance, the algorithm is designed to resist “sock puppet” multi-user exploits; and because each word is tracked individually a contribution will have to be rewritten, not just re-ordered, for another user to “steal” authorship in order to profit from someone’s work.
But the crucial anti-gamer idea WikiTrust have is that the will not display outcome of an editor’s reputation calculation, except through colour-coding. So said Bo Adler:
Thumper17: @AboutThisLater I think our Wikimania group settled on not needing to present the actual rep num, which hope will discourage competitiveness
Now, it will of course be very easy for editors to gauge their status and to identify those who have affected it. But the vital difference is that editors will not have a WikiTag or a “score”. They will not have a title, nor a special hat. Their numbers will be crunched and their pungent utterances rated on a scale from comforting, vanilla pod transparency to a sort of surly, bloodshot puce. But they will not be on a leaderboard. It will not be possible to win the wikis. Reputation will still mainly be mediated through human relationships.
This is a mere psychological difference, but then trust is a psychological business. Wikipedia is the single most powerful poster child for the notion that truth is experimental, that truth is built through common consent, not derived from ideals above.
It may be that WikiTrust will make Wikipedia a more social space, better able to tap and encourage individuals and to generate truth more organically. It may be that trying to quantify something as intangible as reputation will make Wikipedia into a more chilly, points-obsessed place. Who can say? WikiTrust are approaching the process as a fascinating experiment, and after a while you have to stop worrying about what’s at stake. If you believe in crowdsourcing, you have to believe that the best solution will out.
But whatever you think, it seems likely that in the future viewing the internet will not just be a question of where you go, but of whose views you to use, and whose consensus you choose to trust.
