3.29 concording, cont. (29)

Sat, 13 May 89 16:46:05 EDT

Humanist Discussion Group, Vol. 3, No. 29. Saturday, 13 May 1989.

Date: 12 May 1989
From: Willard McCarty <MCCARTY@vm.epas.utoronto.ca>
Subject: concording subsumed

Jack Abercrombie tantalizes us with a hint of a process
liberated from the imitation of old models in the new technology. Now, I
think, we may be getting somewhere. Concording is a primitive, to be
subsumed and transformed as part of a more complex process that exploits
the potential of the machine, right? Let us think about what that larger
process is. Again, we ask, what do we want computers to do for us?

An ordinary concordance, printed or otherwise, is not terribly good for
finding all the inflected forms of a word when all you've got, or want
to specify, is the lemma. So, the concording function needs a
morphological component. Even then, such a tool is exceedingly clumsy
when you want to find ideas, themes, and structures but have only got or
can specify words. So, we need a component than can act on the basis of
synonyms, antonyms, and conventional associations. What else?

Willard McCarty