Too modest to puff his own horn, Thomas Stewart takes on (the failures of) knowledge management.
Wed, 16 Jan 2002 08:50:45 EST

In a message dated 1/15/2002 5:43:35 PM, writes:

>ontologies are brittle and categorization ephemeral;  too much structure
>is a bad thing;  raw search beats structured queries;  views, not 
>Here's the problem with knowledge management:  most organizational and
>individual knowledge is unstructured, poorly organized, difficult or 
>to search *today*, and balkanized with respect to access.  Today's KM 
>try to impose additional structure and relatively rigid organizational
>schemes on this information, which requires changes to the way people create
>that knowledge in the first place ---

Well put.

You know how people sometimes make a distinction between data, information, 
knowledge and wisdom? I find that a difficult one to make, because the 
categorization depends on who is doing the categorizing. I.e. what is data to 
one person might be wisdom to another: I might be wise in the ways of the 
duck-billed platypus; but my wisdom is a data point to a telemarketer of 
marsupial collectibles. A lot of knowledge management "solutions" have the 
same problem. They pigeonhole things in ways that make sense to pigeons, but 
not to odd ducks. (I'd better stop with the metaphors.) 
(Yes, I do think it's possible in a rough way to say that some kinds of 
knowledge material are data/information, and other kinds are 
knowledge/wisdom; but these distinctions become untenable as the size of a 
group increases or as point of view changes. The accoutning department does 
not have the same view of your sales-call information as the sales department 
Similar situation: A few years ago I was giving a talk to some people at the 
CIA, and afterwards fell into a discussion with one of them about the 
difficulties of classifying information--which is another way of saying, 
structuring it. In the old days, he said, wistful for the Cold War, we knew 
what the problem was: Russian troops through the Fulda Gap. We could 
therefore parse and classify information on that basis. We could do "need to 
know" because we could figure out who needed to know what. But today, when 
threats can come from all over, we do not know what we need to know, and we 
certianly do not know who needs to know it, and we do not know whether agentn 
007 does or does not need to know agent 006. How can you classify information 
in a secrecy sense, he asked, when you can't classify it in a Linnean or 
Dewey decimal system sense?

Particularly as you get out of routine work, the knowledge that needs to be 
managed tends to become unstructured. KM, therefore, has to be a suite of 
activities. Knowledge management is not an application, any more than 
financial management or customer relations management or Management Itself 
is. Some of these activities involve software--of several kinds. Some involve 
wetware. Librarians are as important as libraries, for example.