From: Andrew Kulikovsky (email@example.com)
Date: Fri Sep 19 1997 - 11:45:07 EDT
On Fri, 19 Sep 1997, Jonathan Robie wrote:
> At 08:46 AM 9/19/97 -0400, Randy Leedy wrote:
> >I will also take issue with Clayton's statement that Greek is not an
> >L-R language. I am sure that skilled readers of Koine Greek did not
> >jump around in their sentences any more than we jump around in
> >sentences in our own language.
> Reading a sentence (which humans do) is very different from parsing a
> sentence (which computers do). Humans understand the semantics and
> pragmatics of what they read. Computers don't. Many aspects of sentence
> meaning may not be derived purely from the form of the sentence. Consider
> Chomsky's classic example:
> 1. Time flies like an arrow.
> 2. Fruit flies like a banana.
> Unless a parser knows about fruit flies, it can't know that "flies" in
> example 2 is not a verb. What about this example:
> 3. Fruit flies like an arrow.
> What does this mean? To a linguist, it obviously refers to the classic
> examples 1 and 2, which are well known - but only to linguists. To the rest
> of us, it is nonsense. But how is a computer to know that 1 and 2 are not
> nonsense, but 3 is?
I'm back on the list but a long way from home (in Sweden).
I know that SIL/Wycliffe Bible translators are using computer tools to do
translation work. As Jonathan and Randy have pointed out there are lots
of problems but the computer can generate an excellent first cut
translation, which would take months or years for a human to do. Of
course the computer translation is then thoroughly worked over, with
corrections to style and meaning etc,
Computer have their place but they are only as good as their programs
(and their programmers) and since programmers are fallible humans - well
that says it all.
Andrew S. Kulikovsky B.App.Sc(Hons)
This archive was generated by hypermail 2.1.4 : Sat Apr 20 2002 - 15:38:29 EDT