Wiki-Limits

"Einstein could not have posted his theory of relativity on Wikipedia. Wikipedia is not the place for novel theories and ongoing work. Wikipedia is here to consume new material only after it has found its way into the 'mainstream' via multiple reliable secondary sources."

This unremarkable statement comes from one of the self-appointed volunteers who help police Wikipedia to ensure accuracy and objectivity. It's what we all expect from something purporting to be a reliable repository of knowledge, and anyone who has ever contributed to a Wikipedia entry knows how strictly they are monitored. On one level, this is very reassuring and makes me wonder why "professional" publications and associations refuse to accept Wikipedia as a citation source. 

Yet on another level, the above statement belies, in part, the lofty claims attending Wikipedia and its cousins, "crowdsourcing" and "folksonomy." (And yes, I have no compunction about linking to Wikipedia; I use it quite often, as I imagine most readers do as well.) 

In Wikinomics, Don Tapscott and Anthony Williams review the explosion of mass collaboration and peer production, from open source software to Wikipedia. These are supposed to lead to more participation and, ultimately, greater levels of innovation. Wikipedia founder Jimmy Wales calls Wikipedia a "Darwinian evolutionary process, where content improves as it goes through iterations of changes and edits."

As the Einstein statement above makes clear (and this particular volunteer is quite prolific in his or her monitoring--hundreds of revisions a day), places like Wikipedia are not always synonymous with innovation and we shouldn't make the mistake of assuming that mass collaboration always generates more innovation. Places like InnoCentive have a different structure, but when many organizations talk about collaboration and peer production, they often mean a model along Wikipedia's lines, with multiple people contributing to a single project.

For one thing, as we've noted elsewhere, innovation frequently, if not always, means an idea that is initially quite marginal. It is absolutely true that Einstein's theory of relativity would not pass mass collaboration muster. The structure of Wikipedia, as ingenious as it is at soliciting self-organization and self-regulation, is somewhat antithetical to innovation. And that's OK.

But it means mass collaboration as a phenomenon to be applied everywhere to stimulate innovation is a bit misguided. We're all familiar with human dynamics: any group of people, no matter how diverse, often gravitates toward a consensus view of things. Few people enjoy being the odd one out, or want to risk fraying a relationship over opposing viewpoints. And consensus, as valuable as it may be in some arenas like politics, is an antonym for innovation. Wikipedia and other peer production ventures may be seen as evolutionary, but variation and adaptation (non-teleological, by the way) do not have to meet a consensus standard to survive. And "spandrels," the nooks and crannies of natural selection, are often fertile sources of variations. 

I don't doubt the potential for innovation out of some models of peer production--creative thought, after all, springs out of the recombinations of existing ideas. The coming together of different people can be an efficient way to do that. If we want an economy based on innovation, we need as much variation from as many people as possible. Mass collaboration, depending on how it's done, can as easily stimulate that as it can stultify it.

Comments

[Michael F. Martin]

The older I get, the more it has dawned on me that creative workers can be profoundly antisocial. But it takes a community to appreciate the new way of looking things. The tension between these two makes innovation look random.

[Gregory Kohs]

Jimmy Wales opines that Wikipedia is:
"...where content improves as it goes through iterations of changes and edits".

Is there any scientific evidence that Wikipedia content is improving over time? There was a University of Minnesota study that found that Wikipedia, over time, generates a higher and higher probability of presenting a page to a reader in a "damaged" state.

Link:
http://chance.dartmouth.edu/chancewiki/index.php/Chance_News_31#The_Unbreakable_Wikipedia.3F

Also, a study of the 100 articles about the hundred U.S. senators found that (in the 4th quarter of 2007) these articles were wrong -- usually maliciously -- about 6.8% of the time.

Link:
http://www.mywikibiz.com/Wikipedia_Vandalism_Study

Are there other bodies of systematic evidence that "prove" Wikipedia is on an improvement trajectory?

[Dane Stangler]

Gregory: Thanks for pointing these out; I wasn't aware that people had systematically studied Wikipedia's error rate, though it doesn't surprise me. Many pages are flagged as incomplete. I failed to mention the potential for errors and inaccuracies--they are important, but weren't the focus of my post.
Michael: I hope it didn't sound like I was saying innovation is random. I, and many others at Kauffman, am a big believer in interactive dialogue as a route to innovation. And I always paraphrase Malcolm Gladwell on the importance of "structuring conditions for successful spontaneity."

[Roman]

Wikipedia founder Jimmy Wales calls Wikipedia a "Darwinian evolutionary process, where content improves as it goes through iterations of changes and edits."

I tried to come up with a source for this one.. but he actually says the complete opposite:
http://www.youtube.com/watch?v=9CtaVkzNPiA&feature=PlayList&p=711889FE73B02C1F&index=7

[Dane Stangler]

Roman,
Sorry, I should have been more precise in my citation. That quotation about Wales' view is from Wikinomics, page 73, and it's the authors' description. Thanks for the link,
Dane

[mbt shoes]

I love your blog very much, more more info, I will concern it again!