Newsgroups: rec.arts.int-fiction
Path: gmd.de!ira.uka.de!sol.ctr.columbia.edu!spool.mu.edu!uwm.edu!linac!att!mcdchg!chinet!jorn
From: jorn@chinet.chi.il.us (Jorn Barger)
Subject: Re: IF Theory Article
Message-ID: <C0qt34.Bq5@chinet.chi.il.us>
Organization: Chinet - Public Access UNIX
References: <C0I8Iv.172@chinet.chi.il.us> <10040881@MVB.SAIC.COM>
Date: Tue, 12 Jan 1993 13:21:03 GMT
Lines: 32

(That *was* the whole article, Dave.  But I'd love to spin it out
a bit here.)
The structure I'm proposing starts with a simple abstraction hierarchy
that has person-place-thing-motive as its top-level discriminations,
but each of these are of course further subdivided (and one great
challenge is to find other concepts at this level of generality).
The topological trick is to do a cross-product of the whole hierarchy
*with itself*, recognizing that the stories suggested by the cross-
product of, eg, *survival* (a motive) with *tool* (a thing), *will
inherit structure* from the simpler crossproduct of motive and
thing.  So we postpone the combinatorial explosion by carefully
exhausting the toplevel relationships before exploring
1) their specializations (the crossproduct of more-specialized
elements)
2) their *complexifications* (3- to N- dimensional crossproducts
of top level nodes)
There's a profound difference between these two sorts of ramification,
in that the menus for type 2 *will always offer the whole abst.  hier.
over again*, while type 1's menus will be peculiar to the point
you're moving from.

An example:
Imagine you've navigated from the zero-point of this 'content
network' to the relationship "person uses tool".  Now, you may
choose to specialize person or tool ( --> person uses crowbar),
or you may prefer to add an element to the "person-tool repertory"
like motive-survival, suggesting the obvious story "person uses
tool to satisfy survival motive".

More later....

jorn@chinet.chi.il.us (Jorn Barger)
