20.536 synthetic work & its possible fallout

               Humanist Discussion Group, Vol. 20, No. 536.
       Centre for Computing in the Humanities, King's College London
  www.kcl.ac.uk/schools/humanities/cch/research/publications/humanist.html
                        www.princeton.edu/humanist/
                     Submit to: humanist_at_princeton.edu

         Date: Tue, 27 Mar 2007 08:37:19 +0100
         From: Willard McCarty <willard.mccarty_at_kcl.ac.uk>
         Subject: synthetic work & its possible fallout

Allow me to try out a less than half-cooked idea on the assembled
company to see where and how it goes. All replies welcome.

Let's say we divide up work in the digital humanities into analytic
and synthetic kinds. Numerous times, received by no howls of protest
that I know of, I've argued that for the analytic kind what's most
interesting is what computing cannot do, since (presuming much
perfective iteration) the failure illumines those aspects of an
artefact we do not know how to describe or perhaps cannot describe
analytically. That's what Clifford Geertz has called "modelling of",
since it attempts to mimic selected aspects of an artefact. But for
the synthetic kind, Geertz's "modelling for", the original is absent
or does not exist. In that case, failures of description cease to be
important. What is important is what one can project into the world
and experience or operate on. Digitally recorded music is a trivial
example for which the original is absent but the intent is to
reproduce, for normal hearing, with absolute fidelity. A VR
reconstruction from the fragmentary remains of an ancient theatre is
a non-trivial example for which the original is absent, but its
absence requires, let us say, creative inference as well as selective
choice as to what to reconstruct. Ok so far?

Now let us say that the original is not merely absent or has
disappeared from the world but never existed at all. What are the
possibilities? One I can think of involves literary allusion. Let us
say for purposes of argument that we have in digital form all
relevant literature, reasonable representations of historical events
and whatever else might be considered possibly relevant. Let us
suppose we have a theory of how allusion actually works and are able
to write this theory into software that then uses the available
materials to construct some intelligible representations of possible
readings. In this fanciful example we're not specifying the readings
and working back to the mechanism (which would be analytic, yes?),
rather we're working from the poem, according to a theory of how
allusion does its thing, outward to results. Since it's now easily
imaginable that we'd have in digital form more literature than any
individual could hope to have read in a lifetime, there would be no
"original" readings to compare our results to. Then there's the tough
part. Since allusive connections would themselves affect the
possibilities of further such connections, my imaginary "model for"
would be evolving, not simply playing out what had already been in
some sense foreseen. This "model for" would then have the status of
an almost primary artefact.

Comments?

Yours,
WM

Dr Willard McCarty | Reader in Humanities Computing | Centre for
Computing in the Humanities | King's College London |
http://staff.cch.kcl.ac.uk/~wmccarty/.
Received on Tue Mar 27 2007 - 03:04:36 EST