Mike Gotta points to this article by Dennis Howlett called "The poverty of enterprise 2.0 and social computing," in which Howlett observes:

In the context of ’social’ anything, these are incredibly important concepts because what we’re really talking about are power relationships. In any business, power relationships are what provide the hidden glue that makes organizations develop hierarchies and structures. We see this reflected in almost every major form of software you care to examine. From process workflows that mange order to cash, through problem resolution in the call center and out to procurement. We have baked those relationships into the structure and organization of everything we see as providing the means of operating successful businesses. Then all of a sudden, business leaders are asked to forget everything they know, accept that structures can and will be subverted but that it will all be OK because people will naturally want to collaborate to get things done. This is a fundamentally incorrect assumption. [emphasis added]

In my view, this is exactly correct. Social computing applications are not like most other IT, in that they actually transform social relationships in the workplace rather than merely expedite or automate existing processes. You can't buy the technology and receive the hoped-for benefits (spread of tacit knowledge and best practices, improved communication and responsiveness, faster organizational learning, etc.) without also experiencing at least some redistribution of power away from knowledge and process owners.

This is, as I write in Generation Blend, especially problematic when clout correlates with seniority, as it typically does in most organizations. Older power holders got where they are by using pre-collaborative tactics of monopolizing knowledge and using it as a competitive advantage in their own career development, either consciously or not. Perhaps not everyone in an organization is as careerist and attention-focused, but certainly many star performers are, and it is not an irrational response in a labor market characterized by frequent bouts of downsizing and outsourcing.

The novelty of social computing technology and its similarity to consumer software of conspicuously questionable business value provides senior power-holders with a ready-made argument. They claim the software is not read for prime-time, when in fact, what they themselves are not ready for are the power-implications of having to share knowledge, distribute decision-making, and make their work more transparent.

IT departments are not paid to think about stuff like that, and if the business value of social computing is pitched to IT decision-makers in terms of optimizing processes, reducing costs, and all the familiar arguments, even good solutions may not make a dent if they neglect the disruptive effects on the underlying system of incentives and prerogatives supporting the power-holders within the organization.