I was an observer in a debate recently on the question of whether a community of practice had to be self-organising, or if it could be directed. My own view is that communities can evolve, but cannot be designed top down. Of course you can stimulate and direct evolution. However a CoP "Roll out" plan always gets me worried. The original work on this was done by Wenger and in the debate he was being criticised for basing his theory on the endpoint of a process that had taken years to evolve.
I think it is correct to say that Wenger and others in their initial work study the outcome of an evolutionary process. The Boeing and other groups had arisen naturally and were self-organising. However it is also true to say that they started that way. The danger I think was when people could a description of the end point and tried to turn it into a consultancy recipe. Now this is a characteristic of management science, which bases itself on retrospective coherence. What most of the consultants, facilitators etc missed was that you can not replicate the end point of an evolutionary process, but you can stimulate similar starting conditions. That stimulation can be as simple as making the tools available, or providing some initial stimulation or sponsorship.
There are to my mind two disastrous approaches:
1 - Creating an organisational template for communities of practice, with a full roll out plan, dedicated staff etc. etc. This is the classic engineering approach which assumes that there must be a top down, designable RIGHT answer. In practice different communities work in different ways and in different ways at different times. A list serve may be good enough, maybe Grove or similar to get started. A point may arise where you need a taxonomy, formal roles, start up processes etc. but that is expensive and requires a lot of energy to give it any chance of working. Better to create the right ecology in which different types of collaboration can take place, and then consolidate successful experiments when (and only if) they or the organisation can benefit from formalism.
2 - Taking a paternalistic (or maternalistic approach) in which people are held to be children or kids needing help or assistance. This to my mind often goes with consultants to take up therapeutic techniches (in narrative Appreciative Inquiry is one) and move them sideways into organisations. Such approaches were designed for situations where the Therapist takes a dominant role (which is why some consultants like them) and assumes that the recipients are in need of therapy. Most of the time they are not - its the management and consultants who need it not their subjects/victims.
My own view, itself derived from complexity theory is that you need to create an environment in which people can play with multiple tools, moving some of the results to a formal environment, when and if needed. With the growth of social computing and familiarity with those tools this easy to achieve.
This is called clustering (putting together different informal communities) or swarming (creating an attractor mechanism and seeing what comes). Formal communities need to have a cycle of destruction and rebirth built into them otherwise they will be come a force of conservatism. I wrote this up in Complex Acts of Knowing and in a more popular form under the title Just in Time KM.
The first of those articles includes some original research in IBM (which we checked out elsewhere) which showed that the ratio of informal communities to formal communities was 1:10K or put another way if you half the number of staff then you can assume that is the capacity for self-organising spontaneous CoPs that you can expect.
Comments (5)
I agree that COPs are most effective when they can self-organize (or self-dissolve), but it can be necessary in a large organization to jump-start this activity with a few well-grounded groups that have a clear understanding of why they are a COP and their objective in working in such a fashion. They provide an example and hopefully a stimulation to others in the organization that a COP is a useful why to share and exploit knowledge. This is not so much to provide for a "centrally-planned" state in a KM initiative as it is to demonstrate the strength of a knowledge democracy.
Posted by Joel James | November 6, 2006 4:29 PM
Posted on November 6, 2006 16:29
Designing a community is not nearly as precise a science as designing an aircraft. Thus, any “consultatory recipe” worth its salt must consider the variety of conditions that reflect the ecology and thus influence success. Those conditions tend to be cultural as well as resource-oriented, e.g., employee interest in knowledge sharing (vs. hoarding), organizational slack, organizational hunger for innovation, resources available to organize information and direct attention to employee and management concerns, and so forth.
Let’s consider a somewhat worst-case (but realistic) example. Let’s say that management needs to cut the costs to create products and services, and employees support that goal. Chances are, there is little organizational slack, thus employees have little time to engage in activities that are not perceived to be highly valuable. In a healthy CoP, "trained" leaders can drive the focus on employee needs, balancing with management concerns as necessary. If leaders can take the time to understand the knowledge that is critical to the practice, they can help link folks who “have”, with folks who “need.” If these leaders are opinion leaders with a “servant leadership” attitude and some knowledge of how to do their jobs, and management strongly supports this form of collaboration, the efforts can have a good chance to succeed. In this environment a successful “bottom up” approach is unlikely because the organizers will be hard pressed to compete for the attention of their target audience, and may also be ineffective at making knowledge easy to find.
Even this example is somewhat simplistic. I have yet to see a KM tool that has the versatility to help us design and implement a dependably-workable CoP strategy—with the lightest of touches or more invasive interventions (linear or non-linear), driven by what makes sense for the environment.
The foundation for this tool would be an organizational assessment, measuring the various factors that greatly influence success and using those metrics to generate the most healthful, recommended approach—that can be adjusted as necessary during implementation as the inevitable “unknowns” surface.
Yes, I do believe that “top down” can work—but that statement is a far cry from endorsing a one-size-fits-all “recipe” for building successful communities.
If such an assessment tool exists or is under development, perhaps the person who is aware can share.
Posted by Karen T. Danis | November 7, 2006 10:39 PM
Posted on November 7, 2006 22:39
Top down direction to establish CoP's has its limitations since CoPs' in many cases are self organizing systems with a "mind" of their own. If one uses the Missouri slogan: I'm from Missouri; show me the benefit and give folks an indication of what's in it for them, they might join a CoP. That said, I thnk you still need a champion who is quasi-neurotic about the benefits of participating in a CoP. Enlightened leadership can energize the champions and give them license to go forth but is also encumbent on the CoP members (most likely) to demonstrate an organizational benefit.
Complexity theory does apply to CoPs in that there is usually some business problem (strange attractor) that folks rally around for mutual problem solving benefit. Perhaps, a more relevant question is who has the power in a CoP to lead and who follows.
Posted by Robert E. Neilson | November 8, 2006 12:05 PM
Posted on November 8, 2006 12:05
Thanks, I blogged it here: http://joitskehulsebosch.blogspot.com/2006/11/communities-of-practice-two-disastrous.html
(with blogger trackback doesn't work)
Posted by joitske | November 9, 2006 9:29 AM
Posted on November 9, 2006 09:29
"My own view is that communities can evolve, but cannot be designed top down".
I'll go along with that. And we know that COP, and their COI, form around spots in cyberspace where people can have a natter and tag their learning objects. Trouble is, looking from the outside of any org, they tend to disappear in the noise, or can't be found. But what would happen if you gave them a fixed spot which didn't disappear after the funding ran out or project finished?
What happens if yu took one domain and classified the groups (COP) space by using an old taxonomy, sa the dewey decimal code (DDC). So if you were in the .edu.au domain, you would be thinking www.groups.edu.au. If you wanted a group who could allocate the rooms (online environments), regardless of what they consisted of, you'd know (or at least a DDC libraian would)that they could be found at www.025.431.edu.au http://ddc.typepad.com/025431/2007/01/making_dewey_vi.html
And their peers could be found at the same number in any other .edu domain.
The problem, as it appears to me, is that (as yu say) "management science bases itself on retrospective coherence". Our National institutions will always look backwards, which is OK. But in a world where their inhabitants increasingly communicate on a global basis, they're a bit irrelevant. The interesting thing in this change from one way-media built around old National institutions to a two-way media built around global working groups and their communities of interest, is how institutionalists will constantly talk about the size of "their database", but never measure how often a page from it is viewed.
I guess they're afraid of finding out just how irrelevant they really are.
Posted by simonfj | February 14, 2007 7:44 PM
Posted on February 14, 2007 19:44