I found a new blog through Andrew McAfee that I am rather liking: Transparent Office. The blogger, Michael Idinopulos, works at Socialtext, and the blog is focused on the use of social software in the enterprise. One of his very first posts addressed the concept of in-the flow versus above-the-flow kind of work.
In the old world of emails and knowledge management systems, our tools and processes force a rigid distinction between "doing your job" (i.e., in-the-flow activities, usually in email) and "giving back to the organization" (above-the-flow contributions to a knowledge management system). That framing of the issue ensures that people will spend almost all their time in email and very little time contributing knowledge--hence the "culture and incentives" problem that has bedeviled Knowledge Management since the very start.
Too often we approach knowledge management tasks from a naive perspective. We approach it as a Field of Dreams type of problem, if we build it they will come... The problem is, it doesn't matter what you build (i.e. what tools you use), because it's a far more complex problem than just providing the means to manage knowledge. You have to first address the will, the incentives, the really big questions.
Reading Michael reminded me of the Dave Snowden interview that I had posted about previously: The Impact of Web 2.0 on Knowledge Work and "Knowledge Management"
One of the things I've always argued very strongly is that you shouldn't provide any incentives to knowledge work because, for example, to reward people for contributing to a knowledge management database fails to understand the basic trust implications of the knowledge interaction. If you ask me for something that I need in the context of genuine need, very few people if anybody other than the obdurant is going to refuse to give it to you... but if you ask me to share my knowledge in anticipation of possible need in the future by somebody I don't know, then you're never going to get it. It's the immediacy of the context that matters.
Here we have three people (McAfee, Idinopulos, Snowden) with expertise in enterprise knowledge management telling us it's an almost impossible task to accomplish. It's a very complex issue that defies overly simplistic attempts to try to explain away failure. What makes it more difficult is trying to find some places where it has worked in order to try to replicate success. It's almost impossible to find above-the-flow success stories beyond the large scale open example of Wikipedia.