On March 17 and 18th John Tropea, one of the most active bloggers on KM 2.0 and social computing issues made two very interesting contributions to discussion of this issue. On March 17, in a blog entitled “Why KM 1.0 Failed in a Nutshell,” John put his finger on a point very essential to this whole debate, when he says the following and then quotes extensively from an interview with Ross Mayfield, the CEO of Socialtext.
“. . . this article has some of the best quotes on why KM 1.0 has failed.
This is explained so perfectly from the workers point of view…before you get into KM 2.0, if you want to begin to explain to someone what’s wrong with KM 1.0, these quotes will do the job:”
“The way organizations adapt, survive and be productive is through the social interaction that happens outside the lines that we draw by hierarchy, process and organizational structure. The first form of social software to really take off to facilitate these discussions was email.”
“Most employees don’t spend their time executing business process. That’s a myth. They spend most of their time handling exceptions to business process. That’s what they’re doing in their [e-mail] inbox for four hours a day. Email has become the great exception handler.”
“Unfortunately, what it means is all the learning disappears because it’s hidden away in people’s inbox. It’s not searchable and discoverable…”
“So at the edge of your organization, there are all kinds of exceptions that are happening. If you handle them appropriately, you can adapt to where the market is going. You can adapt to the problems you have in your existing structures.”
“…the greatest source of sustainable innovation is how you’re handling these exceptions to business process.”
Here John, and Ross Mayfield, too, put their collective fingers on the difference between operational business processing and knowledge processing in the three-tier model. “Exception handling” is “knowledge processing” and “creative learning” in that model. Ordinary business processing doesn’t involve creative or non-routine knowledge processing/learning. That gets done in exception handling.
John then adds to this quote by offering a statement of Mike Gotta’s” which underlines Mayfield’s point. What John doesn’t do here is to explicitly connect his title: “Why KM 1.0 Failed in a Nutshell,” to these quotes. So, we are left to infer the proposition that KM 1.0 failed because it didn’t deal with “exception handling,” creative learning, and non-routine knowledge processing. And to infer further that e-mail is the primary tool used for this purpose in the enterprise and that social computing software is now creating new capabilities for exception handling that never existed before and that forms the focus of KM 2.0.
There are other things one can infer from John’s entry also however. If one views things from my definition of KM as activity intended to enhance non-routine knowledge processing, implementing Web 2.0 software in the enterprise with the intent of supporting better “exception handling” (i.e. problem solving/knowledge processing) is KM activity. Whether it should be called KM 2.0 is another issue. But, that aside, the actual use of “2.0” tools is part of knowledge processing and problem solving. Further, if KM 1.0 really wasn’t about enhancing “exception handling” by implementing interventions that were intended to facilitate knowledge processing, then the problem with it wasn’t just that it failed in its intent to do KM, but, instead, was that it was never KM in the first place, because its thrust was to enhance routine business processing by enhancing information sharing, rather than to enhance “exception handling.”
Now, this is a pretty radical conclusion about KM efforts that used tools that weren’t of the “2.0” social computing variety, and I think, perhaps, that my projection of the implications of John’s view is probably overly harsh. Yet I think there are two points here worth making.
First, the partisans of KM 2.0 often describe KM 1.0 in an oversimplified way. Thus, it is doubtful whether there ever was a KM 1.0 in the way that writers such as Dave Pollard characterize it, except perhaps very early in the history of KM Best Practices systems. Certainly, such an image of KM hasn’t been dominant since well before Web 2.0 were first widely used a few years back. So, the image of KM 1.0 that is often contrasted with the new KM 2.0 is an idealized story about what KM 1.0 was about, designed to facilitate the image of KM 2.0 that its partisans are trying to convey. However, such a distortion of reality can be dangerous, just because its flawed view can lead to explanations of why older styles of KM failed that may be in error. Thus, John seems to be suggesting that older KM efforts failed because they were only about enhancing routine business processing rather than exception handling. But, if, in fact, many KM projects since the mid-90s have been concerned with enhancing problem solving, exception handling, and innovation, then a) it may be that quite a bit of KM intended to enhance knowledge processing may, in fact, have been going on before “KM 2.0,” and b) we may have to explain why these projects failed (when they did) in another way, which perhaps may not be so friendly to the idea that using “2.0” tools represents an entirely new departure for KM, that will work this time around.
But, second, having said the above, it is also true that many “KM” projects, right up to the present, are about enhancing “knowledge sharing” in support of operational business processing and are not about “exception handling,” problem solving, or innovation. Such projects, however, are often not KM projects, but, in fact, are information management projects using the KM label. This is so because these projects have no way of distinguishing “knowledge” from “just information,” so what they are enhancing is processes of information flow and they include no capability for measuring whether or not they are enhancing actual knowledge sharing.
On March 18th, John added to his views in an entry called “Knowledge Management as an ecosystem.” This post talks about a presentation of John’s. It’s in the presentation that we really get John’s much more detailed view of KM 2.0. The presentation, also entitled “Knowledge Management as an ecosystem” begins with a discussion of current enterprise issues emphasizing rigidity, closedness and siloing, rigid tools, overuse of email, limited connectivity, and overload in communications. He then presents Mayfield’s views, and then explains that the enterprise issues are happening because KM hasn’t been fluid enough, people don’t participate enough in contributing their tacit knowledge, hierarchy, control, and micro-managing dominates decision making, and sharing feels like a task completely apart from one’s work. He then presents the view that the solution to these problems is a “conversation market” relying on “the wisdom of crowds,” self-organization, distributed knowledge production, autonomous flows, network effects, and emergence, and then moves on to outlining the characteristics associated with “the New KM,” as an ecological system. Here’s his list of ‘New KM” characteristics, interleaved with comments of mine on each of them.
JT: — Participation culture (Flatten participation barriers)
JMF: This is not a KM characteristic, but at best an outcome of an associated characteristic of organizational culture if the New KM is successful in its impact on knowledge processing and its enabling conditions
JT: — Social computing (participating, networking, initiated by the people (but they don’t call it KM))
JMF: This is is very loosely expressed, and is therefore deceptive in its implications. Participating and networking initiated by people is a) not necessarily computing, social or otherwise, b) an aspect of such participating and networking may be called social computing if its uses social computing tools, but what these are may be hard to distinguish, c) there’s no reason why anyone should call this KM since it is at best knowledge processing, provided that the participating and networking is about solving problems, and d) the widespread use of social computing doesn’t preclude KM. In fact, KM may be involved in enabling such computing at any time. If not today, then at some time in the future.
JT: — Knowledge Flow rather than manage (Publish/Subscribe model, Increase interactions (get knowledge moving), Communities/Networks, Come to me web, Contextual in time of need (just in time), Create “conditions” for knowledge rather than “manage”)
JMF: I agree that the general thrust of this is desirable, but its framing is deceptive. “Creating conditions for knowledge” is not, “not managing.” Unless the view of KM that is assumed is a “command-and-control” view. But I don’t know too many in KM (if any) who ever advocated or practiced that. Further “knowledge flow” is neither KM nor incompatible with it. So “knowledge flow” rather than “manage,” makes very little sense to me. I can see that KM ought be concerned with enhancing knowledge flow; but this certainly is not a special characteristic of any “New KM.”
JT: — Connect and Context vs Content and Collect (Knowledge/Findability via people (Trusted social filter), Emergence/Autonomy)
JMF: I guess I don’t see the “vs.” here. Nor do I see why this is not part of “knowledge processing” rather than “KM.” KM may be about enhancing these things. But they are not KM itself. Moreover, the importance of enhancing connectivity and context have been emphasized for many years in KM. Thoughts like these have been around for as long as I’ve been participating in “formal KM.” That is, since 1998. So, I don’t think this is any kind of “New KM” orientation.
JT: — Distributed (bottom-up) vs Command and Control (top-down) (Power of many vs a few (wisdom of crowds), Visibility and Opportunity (everyone is a potential innovator), Social Capital (a culture change in working))
JMF: I understand and agree with the idea that distributed knowledge processing produces more innovation and greater adaptability. But this is not KM. It is knowledge processing. Social capital development is also about enhancing the ways in which we process knowledge. KM is what we do to enhance distributed knowledge processing in human systems and to enhance the social capital we use in distributed knowledge processing (See Chapters 4, 6, 9 and 10 of Key Issues . . .)
JT: — Unstructured free-form tools (Easier to collaborate, Easier to get things done)
JMF: I agree that such tools enable “exception handling.” But, again, these are knowledge processing tools. KM may be involved in implementing their availability in organizations. But by themselves, they don’t involve KM activity uniquely.
JT: — “A way to work” rather than a task (Not trying to create a knowledge sharing culture, it just happens)
JMF: This one is a bit ambiguous. Is distributed knowledge processing “a new way to work?” Is “the New KM” “a new way to work?” Or both. I do agree that the development of a “knowledge sharing” culture follows enhanced KM and knowledge processing. “It just happens” as a result of effectiveness in these areas.
JT: — Sense-making (How do I make sense in the enterprise so I can act in it)
JMF: Sense-making is not KM. Knowledge Managers engage in sensemaking of course, and KM should enhance sensemaking, but better sensemaking is not part of a pattern that identifies “the New KM.” It is part of a pattern that identifies enhanced knowledge processing.
JT: — Aim Dave Snowden (Improve decision making, Conditions for innovation)
JMF: KM is for improved decision making. This has been one of its major value propositions from its beginning, i.e. enhanced knowledge capture and sharing were supposed to help decision making. Conditions for innovation were not a concern of First Generation Knowledge Management. But they’ve been a concern of Second Generation Knowledge Management since the 1990s.
In specifying his view of “the New KM” John doesn’t claim in the .ppt itself that “the New KM” and “KM 2.0” are one and the same. In fact, the .ppt implies that they are different, and states specifically, that “KM 2.0 is social computing.” Also, in using the term “the New KM”,” John doesn’t mention the previous use of the term. And I’d be remiss not to call attention to the frequent previous use of the term (10,500 hits on Google for “the New KM,” and 53.000 for “The New Knowledge Management.”)
My own use of the term, in “The New Knowledge Management: A Paradigm and its Problems” characterizes it as a paradigmatic approach to KM and says: “TNKM is a complex of epistemology, ontology, conceptual frameworks, methodological frameworks,methods, and a normative model called The Open Enterprise.” Altogether, my treatment of “the New KM,” written in 2003, employs 16 rather specific criteria to identify this paradigmatic approach to KM. Of course, neither my collaborator, Mark McElroy nor I have a monopoly on the use of this obvious term. But we have developed a very detailed view of the term, and I think that John might perhaps have noted our previous use and the way we had specified the idea.
His idea, above, is different, of course, and one might counter my various comments on it by pointing out that John intended to specify KM in the sense of an ecological system, whereas in my comments above, I seem to be thinking of KM as an activity or set of activities. All this is true. In previous writings I have referred to KM as an activity, a process of set of activities, a discipline or a paradigmatic approach to a discipline, but never to KM as an ecology. However. My response to such a reply is, that KM can never be an ecological system. This is not because there isn’t an ecology that enables to some degree or other knowledge processing in systems. On the contrary, there is such an ecology and if KM is about enhancing knowledge processing it should be centrally focused on making sure that the ecological system impacting knowledge processing is an “ecology of rationality,” rather than an ecology that undermines problem solving. But such an ecology is a normative vision for a state of the enterprise providing a high level of support for knowledge processing, that KM ought to aim at, rather than KM itself. In specifying the New KM, Marl McElroy and I have defined such a normative vision, and called it the Open Enterprise. It is not a KM system, but it is the vision of enterprise knowledge processing that we think that KM ought to aim at; and, incidentally, it has a number of commonalities with John’s “KM Ecological System.”
In my next blog, I’ll continue the discussion of John’s presentation, and will comment both on his treatment of “knowledge” and on his specification of “KM 2.0.”
To Be Continued