Over the last years, I've been following research in KM and TEL around CoPs. CoPs have been widely suggested as a possible way to implement the people-driven approach to KM and deal with tacit knowledge (especially between 1998 and 2004). However, I still did not come across a study that reflects a successful application of the CoP concept (I would appreciate if someone could point me to such a study). I guess the reason is that most of the CoPs are dead before people start evaluating them. Actually, the studies that I found, including this one by Richard McDermott, who co-authored the book "Cultivating Communities of Practice" together with Etienne Wenger and Bill Snyder, show that something must be wrong with CoPs and that we do need a rethinking of this concept.
In my eyes, the concept of CoP has no future. This is mainly due to the fact that CoPs are organized from the top down and thus can only work in a stable and controlled environment. CoPs cannot work within today's increasingly complex and fast-changing knowledge/learning environments. As with complex adaptive systems, self-organization and emergence should be the solution. Over the past couple of years, with the democratization of the tools of production and distribution on the Web, there is a clear shift from CoPs to networked individualism. It becomes everyday obvious that closed, bounded, structured, and hierarchical CoPs are facing slow death, giving place to open, distributed, diverse, and self-organized knowledge ecologies, that emerge naturally from the bottom-up connections of personal knowledge networks (PKN). Facebook and Twitter are good examples of this inevitable shift.
Wenger would say CoPs are everywhere and we are living at the intersection of many of them. I would say I disagree. We are not merely members of CoPs nor are we living at the intersection of many of them. Each of us is rather at the center of his or her very own PKN.
I've read about the new role of "community manager", which is supposed to be widely adopted by organizations in the coming years. In my opinion, the only thing that a "community manager" can achieve is to manage his or her PKN and, even then, only imperfectly. "Communities" or the open version of the same; i.e. networks can never be managed. Organizations do not need a "community manager", what they do actually need is to provide an open and freeform environment, conducive to networking at both external and internal/conceptual levels. That is, an environment where people can freely make connections, see patterns, reflect, (self)-criticize, detect/correct errors, inquire, and test, and thus build, extend, and restructure their PKNs.
Are CoPs really dead? I'm very interested to hear your thoughts on this.