I don't have a specific reference for you, just some comments. :) I think perhaps your original claim is too strong--on both counts. Including the word "can" as you did in your reply is important. Wikis CAN lead to high-quality content. And CAN lead to active communities.
Yes, I came to the same realization after writing my reply. I have started quite a few wiki sites that never achieved either quality nor active communities ;-).
(Clearly there are examples of active communities and examples of wikis that never get used much.) I am not convinced that the technology itself leads to high-quality articles in Wikipedia, and I don't think that the Nature article supports that claim in any way.
When I talk about "wiki", I am not talking about just the technology. I am also talking about the wiki way and philosophy (which, btw, I consider much more significant than the technology that we currently use to support that wiki way and philosophy).
One could hypothesize about characteristics of wiki that help support the goal of producing high-quality articles and try to empirically verify the claim, but to my knowledge no one has done so in any systematic fashion.
Well, there are a number of wiki patterns and anti-patterns that have been written here:
www.wikipatterns.com
But that's not exactly a scientific analysis of what works and doesn't.
Oh welll... Maybe I'll just leave it as a "bold" claim.
Thx Andrea.
Alain