All Life Is Problem Solving

A few days back Neil Olonoff was kind enough to send me a note on Linkedin alerting me to a webinar presentation he had given to members of the Federal Knowledge Management Working Group. Neil says: “It has resulted in an amazing amount of new energy and action in this group. Our Initiative to implement knowledge management in the Federal government is now well underway!” I downloaded the presentation from slideshare and have been thinking about it on and off since.

I was glad to see a proposal for a National KM Center, especially since I proposed one myself some time ago in some earlier posts, and I’m also happy that the Federal KM Working Group is pushing to get something to happen. However, I’m afraid I was disappointed in the argument of the presentation for a number of reasons. These include: a) using too narrow a conception of KM and therefore understating its importance for Federal activities; b) failing to commit to an account of “knowledge,” thus creating a lack of clarity in one of the primary arguments in the presentation; c) proposing that the Federal CKO be located in the Federal CTOs office; d) failing to make a more positive case for KM; e) a variety of loose formulations raising conceptual questions and lowering the credibility of the argument and f) problems with the model at the core of the presentation. I’ll expand on each of these points below and in a future post.

Using too narrow a conception of KM: Neil’s presentation seems to avoid actually defining KM, perhaps in the hope that by avoiding this very contentious issue in the field, his proposal could aggregate the interests of many practitioners who might all agree on what he’s proposing even if they can’t agree on how to define KM. However, the problem of definition can’t be avoided quite that easily, since his argument seems to suggest that KM, for him, is activity intended to enhance the performance of knowledge sharing in the Federal Government.

This conception of KM was characterized by Mark McElroy, nearly ten years ago now, as First Generation Knowledge Management, and it may be contrasted with the more expansive Second Generation KM view he and I have articulated in many pIaces, including in discussing National Governmental Knowledge Management last summer. In that discussion, I characterized KM as activity intended to enhance problem seeking, recognition, and formulation, knowledge production, and knowledge integration (which includes knowledge sharing, as well as other types of knowledge integration).

Of course, if the Second Generation conception of KM is used in proposals for a Federal KM Center, then, suddenly we’re talking about a KM which is much more broadly relevant to the issue of Governmental adaptiveness to challenges, since we’re not just talking about enhancing “knowledge sharing,” but are also talking about seeing problems in Government operations and processes and developing solutions to those problems, as well as just sharing these solutions with those who need them. Since KM is a manifestly much more significant discipline to decision making when construed in the Second rather than the First Generation manner, why would anyone prefer the First Generation conception?

This question is even more important to ask now that the Obama Administration has taken office. The new administration has a surfeit of problems and challenges it must meet. It does need better capabilities for sharing already existing knowledge; but it also needs better capabilities for seeing problems before their effects become unmanageable, and better capabilities for coming up with and severely evaluating and testing new ideas before they have to be applied. So, this Administration especially, since it is open-minded, pragmatic, and cares about reality, needs Second Generation KM and not just First Generation KM.

Second Generation KM is not only preferable to First Generation KM because its scope is broader; it is also preferable because First Generation KM has a foundational problem; specifically it’s inability to clearly distinguish knowledge from information. I’ll write more later about the problem of defining “knowledge” in Neil’s presentation, but the main point here is that a knowledge sharing approach assumes that we have knowledge, can identify it, can distinguish it from “just information,” and then can communicate it to others. However, there is nothing in the First Generation approach that allows us to identify “knowledge” when we see it. All the talk about “knowledge sharing” comes down to “information sharing” in the absence of the ability, inherent in this approach to distinguish knowledge from information.

In contrast to First Generation Second Generation KM includes knowledge production (creation, discovery, etc.). So, we know when we have made new knowledge. We can track knowledge as a cultural product if we wish and therefore we can tell when we are sharing knowledge and when we are sharing information. Thus paradoxically, it is the broader Second Generation approach to KM that allows one to actually track knowledge sharing activities and results, rather than simply tracking information sharing.

In proposing the Second Generation KM conception, I want to acknowledge that I’m not really suggesting anything original. Even before KMCI’s work developing the distinction between First and Second Generation KM nearly ten years ago, major works in Second Generation Km had already appeared. The foremost of these is Nonaka and Takeuchi’s, The Knowledge Creating Company, Oxford University Press, 1995, a book read widely in KM, that makes it quite clear that our discipline is about much more than knowledge sharing. And, increasingly, 21st Century KM practitioners are concerned about knowledge production as well as knowledge sharing. We see this in cases such as the Halliburton, and Partners Health Care cases, in Graduate programs such as the University of Technology, Sidney and The George Washington University, and in popular KM publications such as Inside Knowledge and KM World, in the Ark Group’s book series, and in a host of Elsevier book publications, including books by McElroy, myself, and Alex and David Bennet (Alex was a former head of the Federal KM Working Group). In light of all this, I have to ask why Neil and the current Federal KM Working Group apparently chose to represent KM in First, rather than Second Generation terms? Why diminish KM, when the transition to Second Generation is so clear?

Failing to commit to a clear account of “knowledge”: Again, the motive for this is probably to avoid those interminable arguments over definitions of “knowledge.” I empathize. However, we need to know what Neil means by “knowledge,” because he says: the Government has had “knowledge failures” and has “knowledge gaps.” He also says that “it” is an “enabler” and a “multiplier,” that it is “sticky,” and that sharing it leads to four benefits: “synergy and efficiency;” “quality and value enhancement;” “innovation;” and “morale.” How can we evaluate these claims without knowing what he means by “knowledge?” If he means one thing by “knowledge,” then “knowledge failures” would not be surprising and “knowledge” might not be such a good “enabler” and “multiplier,” and might not lead to “synergy” and ‘efficiency.” On the other hand, if he means something else by “knowledge,” it might be quite a good enabler and might fit his other descriptions as well, but it also might be very hard to identify when we encounter it; and also very hard to ensure that any KM activities can lead to its enhancement.

Nor do I mean to give the impression that there are only two possible construals of “knowledge.” Many different definitions are found in KM (See Ch. 1 of my Key Issues book). What I’m saying is that if Neil doesn’t state what conception he is using, it makes it much harder to evaluate the validity of his claims about the characteristics and benefits flowing from “knowledge.” Instead of clarity, his presentation relies on the good feelings we all have about “knowledge” to justify his claims about its utility and the significance of KM. That kind of approach may have been effective in the early days of KM in the 1990s, but now we’ve had a 20 year very mixed record of disputed successes and failures, and we have many current claims that KM is dying or dead. In this environment, we have to lay out arguments for KM that will survive moderately close analysis and that means being clear about what we mean when we use the terms “knowledge” and “KM.” We may have no agreement among us on such terms. But the need for survivable arguments and proposals demands that our readers be able to analyze what any one of us is proposing and to be clear about what our proposals really mean.

Proposing that the Federal KM center and the Federal CKO be located in the Federal CTOs office: This presents two problems. First, it supports the idea that KM is primarily about technology, and therefore should be subordinate to the CTO and the CTOs strategy. However, KM is not primarily about technology. It uses technology as a handmaiden certainly; but it’s focus is on enhancing knowledge processing, which even if we limit that to “knowledge sharing,” which I would not do for reasons given earlier, is a much broader focus than mere technology applications.

Second, and more importantly, locating the Federal KM Center in the CTO’s office subordinates it not only to the CTO’s strategy, but also to the President’s strategy, whatever that may be. Why shouldn’t The Federal KM Center be subordinate to strategy? Because there is a potential contradiction between the organizational function of enhancing a system’s adaptive capability and the operational strategy of either the CTO or the President. I’ve outlined the steps in my argument specifying this contradiction here, and I encourage readers to look at what I think is a careful development of this position. In a nutshell, however, the overall point of the argument is that adaptive functions of organizations, including problem solving and KM, are about more than just serving the variety of goals or the strategies of organizations. Rather, they are about change and the capacity to change themselves, and so they must transcend and check other executive functions of the organization, lest they freeze its operational pattern in a way that makes it too rigid to withstand the winds of change. This suggests that KM as a function should have autonomy relative to the Executive and therefore should be subordinate to the Legislative and not the Executive branch of Government, as is the Government Accountability Office (GAO). So, I envision a Knowledge Accountability Office (KAO) as the Federal KM Center, as I’ve proposed in an earlier post in this series.

Failing to make a more positive case for KM: Neil’s argument for a National KM Initiative uses the need for knowledge sharing in averting errors as the justification for the initiative. But what about the past performance of KM in the Federal Government? Why not use previous Federal KM activities and results to support a proposal for a National KM Center? In an earlier post, I’ve pointed out that a lot of KM work is done without using the label KM. Why not research previous efforts in the Federal Government of this kind? A lot of Quality work falls into this category, as does a lot of work in the intelligence arena, and in the Sciences. In general, any work in the Federal Government creating and then getting people to adopt new methodologies for helping to: see problems, solve them, and integrate them, is KM work. It’s done everyday in the Federal Government and some of it is successful. Wouldn’t the case for a Federal KM Center be that much stronger if one could point to the full range of successful KM work and talk about the need for a Center to coordinate that work, aggregate it, integrate it, encourage it, enhance it, and evaluate it as part of a systematic effort to enhance the adaptive capability of all agencies in the Federal Government?

A variety of loose formulations raising conceptual questions: On slide 2, Neil equates our “vast reservoir of information” with “these huge knowledge assets,” thus suggesting that “knowledge” and “information” are synonyms. This is something that no KM practitioner should ever do unless one is definitely committed to that view; simply because if one equates “knowledge” and “information,” one is also saying that “Information Management” is the same as “Knowledge Management,” and that there is no need for an autonomous field of Knowledge Management.

Slides 3 – 8 discuss a number of cases that illustrate the need for enhanced knowledge sharing including the: sinking of the titanic; 9 – 11 attacks; space shuttle Challenger disaster; and Hurricane Katrina. On slide 8 Neil claims that “a word to the wise” could have averted these disasters. However, this ignores the possibility that a word to the wise can be ignored by those in authority quite easily and that “knowledge sharing” may be ineffective in averting disasters much of the time. Thus, we know from the Rogers investigation that in the Challenger case when the joint that eventually was compromised by the O-rings failure behaved in unexpected ways on previous flights, NASA failed to test the joint even though it was aware of deviations from specifications. Also, NASA was warned of faulty seals, but saw the problem as “an acceptable flight risk,” and, according to the Rogers commission required proof that it was not safe to launch Challenger, rather than vice versa.

This NASA Challenger failure wasn’t just a failure of “knowledge sharing.” Rather, it was a failure of not looking for and recognizing problems where they existed. Enhancing “knowledge sharing” would not have averted the Challenger disaster. As Steven Spear has made clear in his book, only a more intense attitude toward seeking out, recognizing, and formulating problems would have done that. Further, NASA’s weakness in seeking, recognizing, and formulating problems persisted after Challenger, and was a major factor in the Columbia disaster of 2003. There, too, there was plenty of shared information, as well as shared knowledge, that might have averted the disaster, but also, there was a resistance to interpreting warnings and deviations of behavior from expectations as problems that had to be solved first, before Columbia was launched, and second, before it was allowed to re-enter the atmosphere.

While failures to share knowledge certainly existed in the Katrina and 9/11 cases, it is also clear that other factors may have been more important for averting both disasters. In Katrina, scientists had actually forecast the likelihood of a Katrina-like Hurricane hitting New Orleans, but there was a disposition to ignore these warnings, i.e. a failure of problem recognition. Further, much evidence was provided to the 9/11 commission showing that the possibility of a 9/11 type of event and the intention of al qaeda to attack the US were both there; and that had these been taken seriously by the Administration, there might have been much more disposition to “connect the dots” in time to avert the disaster.

Even the relatively unambiguous case of the Titanic may be problematic, as a case from which we can draw a “lesson learned” about “knowledge sharing.” First, when someone gets “a word to the wise” they may receive it with an open mind and evaluate it carefully. But what if they think that “they’re unsinkable,” or think that there can’t be icebergs where the “shared knowledge” says there are? My point here, is that to the recipient “shared knowledge” is always just information. The recipient will always need to evaluate it before accepting it as knowledge. If their gut reaction to the “shared knowledge” is to ignore, or to discount it, than sharing won’t avert disaster, even when the knowledge itself corresponds to reality. Of course, there’s no guarantee that “shared knowledge” will correspond to reality, and this is part of the reason why it is often ignored and doesn’t have the desired effect in averting disaster.

On slide 9 KM is characterized as “a ten year old discipline.” Actually, it’s at least a 20 year-old discipline if its dated from the time (1989) Karl Wiig used the term “Knowledge Management,” and Wiig, Karl-Erik Sveiby and Bob Buckman performed the first work identified by that name. Many in KM are aware that the discipline is much more than 10 years old, that KM work was going on all over the globe in 1999, and that KM Conferences bigger than any we have now were being held then, because KM was “the hot ticket.” In a post coming soon I’ll take up the question of problems in the model underlying the presentation.

To Be Continued