All Life Is Problem Solving

Joe Firestone’s Blog on Knowledge and Knowledge Management

All Life Is Problem Solving header image 2

KM 2.0 and Knowledge Management: Part Five, More on Dave Snowden’s Take

August 14th, 2008 · No Comments


4) Another important point made by Dave in the podcast is that those looking to create a knowledge sharing culture are thinking about knowledge sharing from the wrong perspective. The problem is not to create such a culture, but rather to increase the connectivity of people, whereupon they will naturally share because the increased connectivity increases “the immediacy of KM requests.” Moreover, this sharing will happen without using explicit incentives, which don’t work anyway. In relation to this formulation, I very much agree with Dave on some, but not all of its aspects.

I agree that those looking to create a knowledge sharing culture are thinking about this issue from the wrong perspective. I think the right perspective is to focus on changing behavior in such a way that the changes are consistent with existing cultural norms, while the changed behavior is such that it creates great pressure for changes in culture to co-evolve along with increasing changes in behavior. Further, I agree with Dave that incentives don’t work in getting people to share. I also agree that the way to change behavior is to enable greater connectivity, which will give people the opportunity to naturally share. I disagree, however, that increased connectivity will necessarily be followed by a stable increase in the level and quality of knowledge sharing, since there is a considerable possibility that increased connectivity will lead to conflict, mistrust, and eventually a decline in sharing. Which way things will go, depends in part on the social context of the introduction of Web 2.0 tools, as well as on how people work through that social context to collaborate in a cooperative way. Finally, Dave refers to KM requests for knowledge sharing. However, even though there may be some of those, requests for knowledge sharing most frequently come from people who are looking for and who may need help, and therefore they are primarily knowledge processing rather than KM requests.

5) In developing his view of the Impact of Web 2.0 tools on KM, Dave draws the conclusion that the use of such tools would lead to a situation in which there would be no management of knowledge sharing. That is, there would be no formal KM roles, but that “knowledge work is just the way we do things around here.” It would not be managed at all.

This view, of course, is based on an interpretation of complexity theory suggesting that the self-organization of knowledge sharing activities enabled by using Web 2.0 tools would result in stable and beneficial knowledge sharing in an enterprise that, in turn, would remove the need for formal KM activities. I don’t agree with this line of reasoning. I think the tools do enable self-organization in knowledge sharing and that the resulting patterns may well be beneficial. However, I don’t believe in permanently effective and stable “invisible hands” in human CASs, especially when humans participating in CASs may want to manipulate knowledge sharing patterns to their own advantage.

Moreover, Web 2.0 tools are only one wave of technology. The next wave, probably focusing on applications using intelligent agents, will present us with new possibilities and also new possible interventions for enhancing various aspects of knowledge life cycles, including knowledge sharing. Still further, KM is not concerned with enhancing knowledge sharing alone, but also with problem seeking, recognition, and formulation, and with knowledge production. While Web 2.0 tools impact these areas as well, I don’t think their impact on self-organization, distributed cognition, and enhanced functioning in these areas is as clear as their impact on knowledge sharing.

Yet another reason why KM won’t disappear as a result of success with Web 2.0, is that the (KM) goal of enhancing knowledge processes is an open-ended one. Even if the introduction of Web 2.0 proves to be a success for KM in knowledge sharing, that won’t make formal KM disappear because the need for formal KM will be gone, Instead, it will merely emphasize that KM can be successful given the right conditions. That will only whet the appetite for formal KM activities and will serve to increase the credibility of KM practitioners.

6) Jon Husband also asked Dave about the impact that Web 2.0 tools are likely to have on the idea and practice of knowledge audits of knowledge assets. Dave answered by saying that such audits implied that knowledge was static and a thing, and that most effective knowledge exists in flows, and is contextually created in a time of need, and he provided the previous line I quoted above about knowledge being a “. . . real time assembly of multiple fragmented memories . . .” He then went on to say a) that a problem with knowledge audits is that “you can’t audit fragments.” b) Web 2.0 results are too unstructured to view them as knowledge assets in themselves. You can do a lot with wikis. But the issue with blogs is that you need a way to recall blogs that are some years old. c) A big “issue is categories and keywords,” and tagging. Blogs can be viewed as narratives that create a semi-structured system, and Cognitive Edge is doing this in some of its work. But to do this well, you need some type of common tagging system in each enterprise, and that system has to handle text, pictures, and you tubes. d) In creating that kind of system, we’re asking people to create levels of meaning for objects like blogs and narratives by human-based tagging. Keywords one associates with a blog or a story may not be in the story, human-based tagging is the key to creating additional levels of meaning. It is not just the text that counts in blogs; it is the context of their creation including the people links.

Dave and I have an ongoing disagreement about the issues of whether knowledge is a thing, or both a thing and a flow. The latest installment of this disagreement is in this post. Here, I don’t need to review that. I did like Dave’s emphasis on how knowledge is created in context and in response to need and also his criticism of knowledge audits which, I too, have objected to for two reasons. First, because lacking a basis for distinguishing between information and knowledge in the hands of KM practitioners they are really information audits, and second, because from my point of view, a) knowledge audits can’t tell you where knowledge gaps are, because such gaps are discovered in specific contexts where problems are arising in instrumental activity, while “knowledge audits” are an abstracted activity that catalogs knowledge claim object holdings, without reference to problem contexts, and b) audits of knowledge asset holdings are not nearly so significant for KM as audits of various sub-processes of the knowledge life cycle and how well they are performed in various business domains in the organization.

But apart from our agreement on the very limited value of knowledge audits, I was also very pleased to see Dave’s emphasis on the importance of generating categories and tags, of the problems of creating common tagging systems, and of the importance of doing this because of the need to create additional levels of meaning for texts and text fragments, to express the richness of context. In some of my earlier work, I’ve emphasized that cultural knowledge is comprised not only of knowledge claims that have survived our criticisms, tests, and evaluations; but also meta-claims about these, recording the performance track record of such claims. While I know that Dave did not have knowledge claim evaluation or error elimination in mind when he talked of placing additional levels of meaning on fragments in Web 2.0 contexts, I’m quite excited about the potential for such human-based tagging to begin to record such track records in the course of adding additional levels of meaning to text.

7) In discussing the problem of merging more traditional sources with Web 2.0 creations, Dave makes the very good point that what matters is “creating a map of the dependency of core business processes on knowledge objects.” He then points out that knowledge objects need to be indexed, clustered, and connected to decisions, and that coherent clusters need to created that can then be managed. He also points out that this is a “bottom-up” design. Jon Husband commented that this approach ‘tracked well” to the use of tags and asked whether that was the same as “bottom-up?” To which Dave replied, that it is similar, but that one needs to put some structure on tagging so that people will use tags in the same way.

I like this way of putting things very much. For business process analysis and management, I think this is a very good view of things, and for knowledge asset audits this is also a way of looking at things that is very useful. Such a map is also a very good construct for sharing knowledge, since it places knowledge objects in the context of their use in business process activity and decision making. Finally, it’s also a very good framework for taking a functional view of both Web 2.0 and older sources in a common context.

8) To end the interview, Jon elicited Dave’s ideas about what will happen to KM in the context of the spread of Web 2.0. Dave’s view is that the KM function will become more important, while at the same time, formal KM will disappear. This will happen in the context of the collapse of the dominance of IT in the enterprise. IT departments will need to place more security around data and content, but reduce security on collaboration to a minimum.

I didn’t find Dave’s reasoning in this part of the interview to be very cogent. Specifically, the connections between the various parts of his view of what will happen are not very clear to me. Perhaps he thinks that the decentralizing and self-organizing effects of social computing will undermine both centralized IT and the formal version of KM which has made use of applications that have been enabled by centralized IT. If this is right, then it might follow that both centralized IT and KM will both decline as Web 2.0 continues to spread. However, I don’t think this line of reasoning works.

For one thing, I don’t think centralized IT is collapsing, since Service-Oriented Architecture (SOA) seems to be giving it a boost right now, which will take at least a few years to subside. For another, I don’t think that formal KM is really linked to the centralized IT function, or that it is focused entirely on knowledge sharing. Nor do I believe that the success of self-organization and distributed cognition in the enterprise removes the need for formal KM. As long as formal KM supports self-organization and distributed cognition it can flourish in a Web 2.0 environment, and especially in a Web 2.0 environment characterized by SOA and mash-ups.

Finally, it is interesting that Dave’s take on Web 2.0’s impact on KM is not heavily oriented toward claims that there is a KM 2.0. Many associated with the KM 2.0 meme, including Jon Husband, John Tropea, and Luis Suarez, apparently view Dave’s interview as supporting the idea that there is a KM 2.0. But I think that Dave himself, views the impact of Web 2.0 as leading to the death of formal KM, by removing its rationale. Perhaps this difference is just semantics. But somehow, I don’t think so. Dave has never viewed generations or ages of KM as defined by changes in technology. Instead he has contended that change is a function of change in the conceptual point of view characteristic of KM. However, I think supporters of the idea of KM 2.0 seem to be taken with a certain technological determinism and seem to think that a fundamental change in KM was at hand when Web 2.0 tools began to move into the enterprise. For myself, even though I don’t agree with Dave’s specific view of the conceptual change that defines a new generation of KM, I certainly agree with him that the issue here is conceptual, and not technological. The fundamental point is that KM is a human activity and not a class of software tools. Such tools are only enablers for KM practice. As such, changes in software tools that have implications for more successfully enabling certain areas of KM practice are very significant. But much more fundamental change in KM is the sort of change that expands the very scope of activities that we identify as KM. In the period before 1995, KM was mostly focused on sharing knowledge that was assumed to exist. But during the period from 1995-2000, it became increasingly well recognized that KM was also about enabling knowledge production in various ways. This was a fundamental change in KM, because it identified KM with innovation and organizational learning and not simply with spreading existing knowledge around. This fundamental change in KM is still in “early days,” because its implementation still has a long way to go. It is the real KM 2.0. Changes in conceptual orientation from top down KM to complexity, and changes in KM from Web 1.0 to Web 2.0 tools are very important, but they don’t add anything new to the scope of KM, and it is that scope that is the true fundamental dimension of the field. I’ll return to this theme in future posts, as I finish discussing the progress of the KM 2.0 debate in the remainder of 2007 and in 2008 up to the present.

To be Continued

Tags: Complexity · Epistemology/Ontology/Value Theory · KM 2.0 · KM Software Tools · Knowledge Integration · Knowledge Making · Knowledge Management