8.0176 Importance of Algorithms; Review of EPFTD (1/50)

Fri, 9 Sep 1994 10:28:43 EDT

Humanist Discussion Group, Vol. 8, No. 0176. Friday, 9 Sep 1994.

Date: Thu, 8 Sep 94 21:40:59 -0400
From: mccarty@epas.utoronto.ca (Willard McCarty)
Subject: review of the EPFTD

HUMANISTs may be interested to note the review of The English Full-Text
Poetry Database by John Sutherland, "When in Rom", in the London Review of
Books for 9 June 1994, pp. 7-8. Most of it is a mild celebration of the
commercial venture, "the kind of alliance between profit-driven
entrepreneurs and institutions of learning that Tory theorists foresee as
the salvation of Britain's higher-education sector". Whether salvation is
the likely result, and whether we are actually better off for various
scholarly concerns having been overlooked, are questions some here may wish
to debate. My eyebrows were raised, however, by a prediction at the very
end of Sutherland's review:

The whole idea of critical brilliance will come more and more to
depend on the ability to devise elegant algorithms as the databases
become larger, more interactive, with subtler tagging and search

This is an interesting claim. Let us assume for the purposes of argument
that devising these algorithms can "by the end of the century" -- the
millenary moment of Sutherland's prediction -- be done using some computing
language congenial to literary scholars. (In less than 6 years?) Let us
further assume that these databases have continued to grow beyond the size
at which no scholar can have read through the resource he or she is using.
Will we be wanting to define the elegance of literary algorithms as our
colleagues in the sciences do, "pleasing by ingenious simplicity and
effectiveness" (OED)? How will effectiveness be determined? The realm of
literary data is rather different from the realm of physical data.

I don't mean to dismiss Sutherland's notion that algorithms will be
important, or more generally, that empirical modeling of literary theories
will. In a lecture Northrop Frye delivered here in Toronto a few years ago
he noted that were he writing Anatomy of Criticism now he would pay a great
deal of attention to computational modeling, as some of us are doing now.
What seems to me to need more thought is the interaction between elegant
algorithms and unaided human insight into imaginative literature -- or more
accurately, between algorithmic and intuitive thinking, both equally human
but profoundly different. Only a believer in the myth of progress will fail
to see that the latter is not finite and thus doomed to be conquered by the
former. Shouldn't our attention be directed to the moving intersection of
these two rather than to the silly contest between academic generations,
"the old buffers pitted against the new virtuosi of the virtual library"?

Willard McCarty

Willard McCarty
Centre for Computing in the Humanities
University of Toronto