Categories and Calculus (re: Applied Evolution)

John McCreery (JLM@TWICS.COM)
Tue, 26 Sep 1995 08:22:11 +0900

First, a quick thank-you to Nick Corduan for the
Moche/Chimu example. Being neither an archeologist
nor a Latin Americanist, I wait eagerly for those better
qualified to comment on the specifics. I did spend
some time, however (a long, long time ago) doing
European medieval history, and what I observe in
Nick's remarks is a variant of the problems we faced
in trying to decide, for example, when the Roman
Empire ends and the Middle Ages begin. I see a classic
civilization, a time of troubles with natural and
political disasters, then the emergence of a new
pattern that, when we examine it closely, has
numerous continuities with the old. When does one
become the other? Then I recall a lot of Stephen
Gould's writing that addresses the same kind of issue
in biological speciation and the physical
anthropologists' search for fossils to close the gap
between, say, Australopithecus and Homo Erectus.
Stepping back and recalling the discussion we had
recently of Berlin and Kay on color categories, I do the
philosopher's trick and suggest that in all the cases we
are having problems imposing categorical boundaries
on continuous streams of phenomena that are only
"known" by sampling at intervals. Then we attempt to
use our samples as prototypes for a system of
categories that divides the stream into static lumps.
We turn the streams into strings of beads, then are
worried about how we get from one bead to another,
forgetting that the beads are artifacts and the
underlying reality continuous and constantly changing.
My mind drifts back to a math class, and I suddenly
have a vision in which I see us reversing the evolution
of the integral calculus. Integrals were created by
imaging that the space under a curve can be
represented by rectangles which almost, but not quite,
fill the space in question. As the rectangles are made
arbitrarily smaller and smaller they fill the space
more and more completely. In our everyday
"commonsense" way of thinking, we pretend that
whatever categories our prototypes suggest exhaust
the space we are looking at. They are, inevitably, only
approximations. Thus, at the boundaries, we are
always to some extent mistaken about where the
boundaries should fall. I wonder how differently we'd
think about things if we saw the world in prototypes,
with distributions spread out around them, instead of
in nasty categories with boundaries to be defended?
What then, for example, would become of "race" and
"ethnic groups and boundaries"? The moral of the
story? Yes, Virginia, there is an anthropology. It, and
we, are all connected.

Comments?

John McCreery
Yokohama
Tuesday, September 26, 1995