By David Weinberger
When Isaac Asimov wrote his Foundation trilogy about how a galactic empire rose and fell, he came up with a way to chart large-scale events so that they'd have the inevitability of history: He read history. He took the general outline of the events from Gibbon's "Rise and Fall of the Roman Empire." (Of course, if he did that now, he'd be exposed as a plagiarist and then sued for violating Gibbon's copyright, but that's not my point.) We don't do a very good job of predicting history because it's way too complex. History emerges from the welter of events, interests and resources of us humans, with the occasional volcanic eruption and drought just to keep us honest.
Your business isn't as complex as all of human history, but it's complex enough that you can't predict much beyond the revenues for the coming quarter. And your business is embedded in a market and economy that are even more complex and unpredictable. You can't be in business for more than a month without finding that out. For example, the first company I worked at launched a product that required what was then an excessive amount of RAM—4 MB—on a Mac II. Could we have predicted that a Korean factory would burn down, helping to cause a dramatic rise in chip prices? In fact, even if RAM had remained cheap, our product's success depended on how quickly the Mac II was adopted by corporations, the plans of competitors and hundreds of other variables in an equation we couldn't have formulated.
So, we all do the best we can, assessing the risks and hedging our bets. But there's a limit. Business is just too complex. On the other hand, it's complexity that enables the emergence of the unpredictable that moves us forward. Complexity isn't the whole story; emergence is the other half.
Emergent properties are, by their nature, surprising; they can't be predicted. There are good, solid reasons why what emerges happens, but you can't know the causes ahead of time and predict what will happen: There's no short-circuiting the process as we can with simpler events like figuring out what the phase of the moon will be exactly five years from now.
After a property or event emerges, we can analyze it. "Oh, sure, Netscape was doomed because Microsoft's operating system monopoly ... " Or was it because Microsoft's massive cash reserves ... Or was it because Microsoft's experience in building mass-market software ... The fact is that all of these factors are relevant, but they weren't enough, even taken together, to let us predict the outcome of the browser wars any more than we were able to predict the day the Vietnam War would end. There are just too many factors. Besides, pinpointing the cause of death isn't the same thing as understanding life.
So does this mean that business is so unpredictable that KM is without value? Obviously not. The smaller and closer in the decision, the more reliable are the predictions of outcomes. Further, much of KM is concerned with the transfer of skills, not the predicting of futures.
But even in the transfer of skills, emergence plays a role. In fact, it may be at the heart of the problem with tacit knowledge.
The most important tacit knowledge isn't simply explicit knowledge that hasn't yet been uttered. Humans aren't databases just waiting for the right queries to be run. The tacit knowledge that lets a senior technician diagnose problems faster and more accurately than others, or the tacit knowledge that lets a CFO see in 10 seconds why a proposed merger is unlikely to work, is emergent knowledge. It comes from a rich, confusing context of memories, heuristics and associations that the technician or the CFO may not be able to explain. In many cases, she or he will form a judgment incredibly quickly and only afterward will try to analyze why.
So, the attempt to make explicit the tacit knowledge in an organization may in fact be an attempt to short-circuit the chaotic process of emergence. But that's exactly what emergence doesn't allow. In such cases, a KM system can nourish the intelligence from which wisdom emerges but neither replace it nor make it explicit.
David Weinberger edits "The Journal of the Hyperlinked Organization" (hyperorg.com), e-mail firstname.lastname@example.org