Knowledge management: the latest battle between the neats and the scruffies

I’m off to participate in a panel session (B105 – Changing/Resetting the Enterprise With PKM & Social Software Tools) this week at the KM World 2009 conference in San Jose. I thought I would rerun this post from a couple of years ago, as it reflects some of the thinking I plan on sharing. If you’re there, let’s hope we manage to connect.


"There are two groups of people, those who divide people into two groups and those who don’t." – Robert Benchley

Years ago, when I was doing work in the field of AI, I came across one of those binary splits that continues to be useful for my thinking; the split between "neats" and "scruffies." In the field of AI, the split differentiated between those favoring highly structured, logically precise approaches and those who preferred something more along the lines of "whatever works." Wikipedia offers a nice summary of the debate from that field.

Back in my school days, I think I was a neat (philosophically, not in terms of my room or study skills). When I first delve into new areas I am drawn to those who argue the neat case. As I get older and, I hope, more experienced, however, I find myself increasingly scruffy.

Much of the recent debate in the narrow field of knowledge management can be interpreted as one more recapitulation of the neats vs. scruffies argument. The technologies of blogs, wikis, and social media that collectively comprise the emerging notion of Enterprise 2.0 celebrate scruffiness as the essence of success in knowledge-intensive enterprises. The claim, backed by appropriately messy and sketchy anecdotal evidence, is that a loose set of simple technologies made available to the knowledge workers of an organization can provide an environment in which the organization and its knowledge workers can make more effective use of their collective and individual knowledge capital. Grass roots efforts will yield value where large-scale, centralized, knowledge management initiatives have failed.

Several implications flow from adopting a scruffy point of view. For one, "management" becomes a suspect term. If you can manage at all, you must do so at another level of abstraction. You aren’t managing knowledge; instead you are trying to manage the conditions under which knowledge work takes place and within which valuable knowledge might be created or put to use. At that point, it becomes more productive to think in terms of leadership rather than management; particularly if you subscribe to Colin Powell’s characterization of a leader as someone you’ll follow to discover where they’re going.

Second, you will need to deal with the problems that the neats have created in previous runs at knowledge management without alienating them at the same time. In most large organizations, knowledge management has been characterized as a technology problem or as a analog to financial management; placing it squarely within the purview of the organization’s neatest neats. This is a recipe for disappointment, if not outright failure.

It might possibly be an open question whether knowledge management can be eventually reduced to something as structured as accounting or library science. But it is a lousy place to start. Most organizations aren’t yet mature or sophisticated enough about knowledge work issues and questions to be obsessing about taxonomies or measurement and reward systems for knowledge work. But those are activities that are neat and specifiable and only superficially relevant. They lead to complex efforts to get to the right answer when we would be better served by simpler efforts to focus on productive questions.

C-words of knowledge

I’m working on a report for a client about knowledge management and knowledge sharing and I am deliberately avoiding the question of defining “knowledge.” I’ve learned that it’s a rat hole of interesting coffee shop conversation that ultimately produces little of value. On the other hand, I started playing with the idea of words that you might use during the conversation. Combine that with Twitter, constrain the problem, and see what results. I posted the following Tweet yesterday to start things off:


So far that’s led to contributions from @shifted, @hjarche, @hylton, @coyenator, and @rsukumar. Here’s the list as of this morning, which I’ve split between verbs and nouns (we seem to be a bit short on the noun side):


What would you like to add to the list?

JP Rangaswami on KM … “Clear, Transparent, Searchable, Archivable, Retrievable”

This has been lurking in my RSS aggregator for quite a while, courtesy of Jon Husband. There is really a lot of highly condensed insight in this post.

Thanks to JP Rangaswami for distilling social computing (in the context of work) to an essence.

From his post “Facebook and the Enterprise, Part 5: Knowledge Management“.


“More and more, knowledge management is going to be about reducing the cost of, and simplifying the process for, letting someone watch what you do. Nonintrusively. Time-shifted. Place-shifted. Searchable. Archivable. Retrievable.”


Via Dave Pollard via Nancy White

Powered by Qumana

JP Rangaswami on KM Clear, Transparent, Searchable, Archivable, Retrievable
Tue, 02 Sep 2008 13:08:06 GMT

Some links on social media applications within organizations

As part of my talk yesterday at the Social Media Strategies conference, I made passing reference to a number of stories, blog posts, bloggers, thinkers, and writers. It’s an occupational hazard of being a former professor.

I’ve written about different elements of yesterday’s talk over the course of various blog posts over time. Here are links to some of the most directly relevant together with links to other items I referenced:

Finally, I drew on a number of smarter people than I on the topics of expertise and organizational change. Here are some good entry points for further reading.

Seeing Organizational Patterns : A New Theory and Language of Organizational Design, Keidel, Robert W.
Keidel is an organizational theorist/designer who builds a very practical way of thinking and talking about organizations around the simple observation that all interactions in organizations can be understood in terms of the blend of control, cooperative, and autonomous ways of relating that organizational members can engage in. For the sports-minded, Keidel maps these three basic relating choices to the sports of American football, basketball, and baseball. He builds a nice case that organizational design choices can all be understood in terms of how these three fundamental relationship choices are mixed and blended.

Situated Learning: Legitimate Peripheral Participation, Lave, Jean
Jean Lave is an ethnographer working as part of Xerox’s Palo Alto Research Center. In this volume, Lave explores learning as primarily a social phenomenon and builds a very practical theory of how apprenticeship forms of skill acquisition and learning work in the real world.




Pragmatic Thinking and Learning: Refactor Your Wetware, Hunt, Andy
I’ve become a general fan of Andy Hunt’s pragmatic programming series of books. They are useful well beyond the narrow area of software development. In this new book, Hunt offers a useful introduction to the Dreyfus model of expertise and how it applies in the general context of knowledge work.




Quality Software Management (V1) : Systems Thinking, Weinberg, Gerald M.
The first of a four-volume exploration of the particular and peculiar challenges of managing the development and implementation of software. The first volume introduces fundamental notions of how to model and think about complex systems and how they respond to change. Weinberg adapts Virginia Satir’s family therapy theories to the environment of complex organizational environments.




Quality Software Management (V3): Congruent Action, Weinberg, Gerald M.

While all four volumes of Weinberg’s work are valuable, this volume on what Weinberg describes as “Congruent Action” is the most useful for understanding organizational change in concert with Volume 1.

Dueling philosophies: social media vs. knowledge management

Venkat Rao of Xerox recently introduced an important argument about the underlying differences between social media and knowledge management approaches inside the Enterprise. Here’s the way I described them at delicious. Both are worth a look, a read, and some thought.

Ray Sims is collecting definitions of knowledge management

Combine two slippery but important words and it’s little wonder that you can find such a proliferation of definitions. For some reason, this reminds me of Danny Kaye’s Choreography number from White Christmas.

43 knowledge management definitions – and counting

Before I really get going on my day, here is an entertaining (or sobering) list of 43 knowledge management definitions – and counting from Ray Sims, who is heading back into the world of being an official KM’er as I head out to do product management.  It might have been funnier had he stopped at 37.

For many years I’ve been saying that I didn’t like the term “knowledge management” as (a) it was fundamentally an oxymoron, (b) there was no consensus within the industry as to what the term meant, and (c) in many companies the term carries negative connotations due to a perceived lack of value from earlier so-called knowledge management efforts and/or belief that knowledge management was a fad that we have moved on past or has been absorbed into other disciplines.

I like a number of these and have used variations of them in the past.  As someone on the Act-KM mailing list noted, there are easily as many definitions of knowledge.  Ray or another enterprising individual might want to stack these definitions into buckets about how “knowledge” is perceived by the people using the definition.  Process-centric definitions would look at knowledge-as-verb.  Storage-centric definitions might think of knowledge as a thing to be controlled.  People-connection definitions might think of knowledge as appearing via interaction. etc.  Ray has already created a couple tag clouds of the definitions.

The problem of emergence

Andrew McAfee’s Sloan Management article defining Enterprise 2.0 is available for download, so I took the opportunity to reread it, after a recent chat over coffee with Jordan Frank of Traction Software.

Enterprise 2.0 is Now Free

The article, at least.  MIT Sloan Management Review, with support from IBM, is making a set of ‘classic’ (thanks!) articles freely available to all comers. So the full text of my original SMR article “Enterprise 2.0: The Dawn of Emergent Collaboration” can be downloaded here.

I don’t know if this is a temporary or permanent arrangement, so I’d suggest acting quickly.

One of McAfee’s central arguments is on the importance of emergence in successful Enterprise 2.0 initiatives. Here’s the way he puts it:

Second, the technologists of Enterprise 2.0 are trying hard not to impose on users any preconceived notions about how work should proceed or how output should be categorized or structured. Instead, they’re building tools that let these aspects of knowledge work emerge.

This is a profound shift. Most current platforms, such as knowledge management systems, information portals, intranets and workflow applications, are highly structured from the start, and users have little opportunity to influence this structure. Wiki inventor Cunningham highlights an important shortcoming of this approach: “For questions like ‘What’s going on in the project?’ we could design a database. But whatever fields we put in the database would turn out to be what’s not important about what’s going on in the project. What’s important about the project is the stuff you don’t anticipate.” [p.25]

In an accounting or ERP system, the system’s designers specify all aspects of workflow, database design, and information structure in advance. Users are expected to select from among pre-defined choices and enter only such data as the designers have provided for. In designing a system for emergence, the designers leave a number of these decisions open; waiting for users to fill in the blanks. So, for example, instead of locking down a taxonomy for categorizing documents, the designers might provide a tagging system to allow a folksonomy to emerge from the idiosyncratic choices of each user.

The attraction of emergence is twofold. One is the realization that conventionally structured approaches have generally failed when tackling knowledge intensive problems. Knowledge work and knowledge workers don’t mesh well with the structuring techniques appropriate to industrial work.

The second is the perceived success of emergent approaches behind current Web 2.0 success stories on the Internet. It’s easy to see the power of emergence in such examples as flickr, facebook, and technorati.

Transplanting those experiences inside the boundaries of the organization is no simple task. What works at the scale of the public internet may not generate sufficient momentum within the confines of a single organization. Moreover, Internet success stories ignore or gloss over the failures and also-rans. Failure in the market is tolerated in ways that don’t translate well inside organizations. 

You want the energy and creative outcomes that can come from a successful emergent approach, but you can’t simply rely on unaided market forces to fuel the process. “Unaided” is the key notion. Emergent successes in the market benefit from scale and viral strategies, but they don’t happen by accident. For starters, there is a marketing strategy and plan that exists in parallel with a technology implementation plan.

Enterprise 2.0 efforts within organizations also need a marketing plan to accompany their implementation plans. Like any marketing plan, this plan must identify and characterize its target market of potential users. In particular, the plan needs to identify those potential users who are most likely to benefit from the new capabilities and whose successful use of the technology will be interpreted as an endorsement to be emulated.

Is a marketing plan, by itself, sufficient to allow the other aspects of an Enterprise 2.0 implementation to emerge from use? Appropriate scaffolding and careful seeding of content will prove more useful. A complete taxonomy, for example, may overwhelm a small set of potential early adopters. On the other hand, an empty tagging system will prove too much of a blank slate for users more accustomed to the structures of conventional systems. Providing a sample of suggested tags or categories coupled with some live content can point users in the right direction.

Supporters and early adopters will also benefit from coaching and mentoring on how to use selected technologies to accomplish their goals. This coaching would focus on working out strategies for how to use the technology to accomplish specific business and organizational goals. This requires a different kind of engagement between the implementation team and the target user group. In particular, it entails introducing the user population to key design questions and issues that would typically have been dealt with by the implementation team.

In some respects, “emergence” is a fancy organizational development word for “messy.” The more our systems must deal with the complexities of the real world, the messier they must be to accommodate that messiness. Large scale organizations in general, and IT organizations in particular are not generally comfortable with messiness. Calling it emergence helps. But the fundamental need is to acknowledge that it is more useful to learn as we go and build our systems accordingly, than it is to force fit these systems into structures that cannot contain them.



Pushing for simpler knowledge management

Dave Snowden’s blog Cognitive Edge has rapidly become one of my top sources for insight into the subtleties of managing knowledge in the organization. Earlier this summer, he packaged up some of his insights and advice on the connections between learning and knowledge that reflect his bias that “narrative material, anecdotes, pictures, fragments of stories is more valuable than structured documents and closer to the way we naturally share and create knowledge and learning.” [Learning lessons or lessons learnt? ]

For example, here is one of his action recommendations:

Learning lessons or lessons learnt?

  • Make capture continuous and a part of the job, not a post job after action review. The more capturing stories is part of the way we do things around here, the better it is. The other advantage is that you can then see trends emerging in the way that people index the story material which allows early intervention. This sort of switch is key to moving from a retrospective and codified set of documents, to a dynamic narrative based learning ecology.

We need to be learning lessons continuously, not documenting lessons learnt.

If many earlier approaches to knowledge management were technologically over-engineered, Snowden is essentially arguing for the power of simplicity.

More on knowledge management as learning support

Greg Lloyd at Traction Software also picks up on the same JP Rangaswami post that I did yesterday. He offers several additional examples of the value of making knowledge work visible as a simple tool for supporting on the job learning. Here’s one of his many useful insights. Go read the rest.

Learn by watching – Then do

Learn by watching – Then do
Blog446:  August 14, 2007 7:22:00 PM EST, Posted by Greg Lloyd

Each project’s serial file was nothing fancy. Usually it was a few file drawers with incoming and outgoing correspondence, briefing slides, q&a memos, contract actions and meeting notes, all top bound in chronological order – full contracts, formal specs and other deliverables were filed separately. In pre-email days, the project serial file was a pretty accurate snapshot of our interactions with the outside world interleaved with internal notes and memos. We all kept our own date stamped lab notebooks for private jottings.
A day or so of close reading and the chance to ask a few pointed questions to the original project engineer (“You said WHAT to Captain K??”) usually got us up to speed on the pulse of each project – not just the formal status and deliverables. We learned to use the project file to refresh our memory on details before and important meeting or decision – or just to reflect and review the bidding. We learned to use each other’s project files to keep track of dependencies and learn how to handle problems. …
I know that an electronic form of serial file can replace the old paper trail, since that’s what I use every day. The TeamPage blog + wiki tool lets everyone look over my shoulder – and vice versa – as we tear off in different directions and do our work as individuals or teams.
I rarely need to read any one project in real time, but I know that I can come up to speed quickly, search across all projects, and dive in if I need to. If someone asks for help or sees an opportunity, they can post it if it’s not urgent; add a tag to anything that needs quick action; or IM a permalink if they need me to look at something now. What I can do, all of Traction’s employees can do – only the “Board of Directors” project is private. Board pages or posts – including monthly financials – are cross-tagged to make them visible to all hands when the dust settles.
There are days when I wonder whether one of the fundamental impediments to the take up of blogging and wikis within organizations is, in fact, their utter simplicity.

Knowledge management = creating environments for learning

One of the recent additions to my feed subscriptions is Confused of Calcutta by JP Rangaswami. Recently, he’s been thinking about Facebook and its potential role in Enterprise settings. Today’s installment has an interesting riff on the nature of knowledge management. It dovetails nicely with some of the things I’ve had to say about visibility and knowledge work.

Facebook and the Enterprise: Part 5: Knowledge Management

Knowledge management is not really about the content, it is about creating an environment where learning takes place. Maybe we spend too much time trying to create an environment where teaching takes place, rather than focus on the learning.

Since people want to learn by watching others, what we need to do is to improve the toolsets and the environment that allows people to watch others. It could be as simple as: What does my boss do? Whom does she talk to? What are her surfing habits like? Whom does she treat as high priority in terms of communications received? What applications does she use? Which ones does she not use? When she has a particular Ghost to deal with, which particular Ghostbuster does she call?