University of Virginia Library

Computers Inept

I would like to correct some of
the basic assumptions behind the
editorial "Faculty Fiefdoms (sic)"
(Cavalier Daily, December 5, 1969).

One of the basic arguments in
this article claims that "Undergraduates
are intelligent enough to
know that not everyone in this day
of computer translation needs to
know a foreign language. They are
also intelligent enough to realize
that they aren't going to master the
language in three semesters they're
required to take."

In the first instance, the day of
computer translation is far from
being here. Many prominent linguists
consider the task fruitless and
that it will eventually prove impossible,
or at the very least, impractical.
Even those who are presently
working on this problem hold little
hope for effective "machine translation"
in the near future.

As of this past summer, when I
left one such project, the capability
to "parse" (analyse and translate)
passages up to one hundred words
in length had broken the 50% mark,
in some cases reaching 75%. Anything
longer could be treated with
only the possibility of retrieving a
maximum of 25% of the text
through translation.

The point is that even a student
with only three semesters of a
foreign language can do at least as
well as present computer programs
are capable of doing, and in most
cases much better. The computer is
only capable of doing what it is
programmed to do. Therefore it
cannot engage in free conversation
requiring independent thought.
Fortunately, undergraduates can
think independently, and are not
thoroughly programmed - yet.