The word "knowledgebase" is evidence of an assumption, rapant in the
computer community, that everything in life can be equalled or bettered if
repackaged in an existing computing environment. The assumption is only
half right: the destructive effects of poor thought can be reliably
enhanced by being placed in a computing environment, while the
enlightening effects of good thought are hard to reproduce in any
environment. I'd like to comment on the sensitivity of knowledge to its
representation.
The primary symbolization of knowledge is that which attaches to the
original experience in which the knowledge was engendered. An account of
experience is a story, and an important experience is accounted for in a
'true story' in Plato's sense. A true story is one that reliably
reproduces the experience, and creates the knowledge, in one who is
prepared to hear it. Gadamer and many others refer to the "story
structure of experience". Jerome Bruner argues against the turn in
cognitive science that misses the story structure as the basis of human
knowledge.
Accounts of experience are a primary symbolization. These are difficult
to deal with and unreliable in their effects. Those who want to preserve
the experience resort to secondary sybolizations, such as dogma and
systematics. Drilling in the formulas preserves the experience in a way,
and can be used to steer around some errors. But reenactment of the
knowledge-generating experience would be preferable.
Voegelin wrote a huge multi-volume work on intellectual history, then
concluded that he was working in secondary symbolizations and threw the
whole thing away. Gadamer was being catty when he pointed out to me that
Voegelin had actually read secondary sources -- can you imagine that! This
suggests how sensitive the knowledge experts can be when it comes to
representation. I once offered to hyperindex the Voegelin oevre, and his
legatee asked me pointedly, "Why?" In a seminar paper I extracted a
typology from Gadamer, and the seminar leader was astonished at my
'success' because Gadamer normally defies this treatment.
Let me reword three of your four propositions. (Here I go again, using a
form that contradicts the point.)
* A story can be a means of re-engendering knowledge in another.
* A true story is one that is redolent, inducing response to reality, not
to an ideological or doctrinal formulation.
* The experience of the hearer is drawn forth by the telling, but is also
offered up by the prepared hearer.
A conclusion one could reach from this is that knowlege is embodied but
not stored. Similarly, if information is data with a meaning, and if it
always take 'a someone' for meaning to exist, then information isn't
stored either, although it can be "explicated", and the explication
stored. Perhaps, then, knowledge can be explicated as well as engendered.
The explication will be a secondary representation, and a primary
representation will be a story by which knowledge is engendered. Neither
an explication nor an engendering story "store" knowledge. Knowledge is
an attribute rather than a substance, and it is an attribute of a human. A
human has knowledge if he can apply it. Both a human and a computer are
able to recount both primary and secondary symbolizations, but the act of
recounting is not evidence of having the knowledge.
-- Kent Myers kent_myers@smtplink.sra.com