[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [aymara] Aymara and Western Culture

Actually, David, while I certainly do not value my opinion over yours, my
Logos experience is that once you have thoroughly parsed the source
sentence, you can get pretty much whatever you want for the target sentence.
That is, if it is a "rule based" system.  In Vietnamese which was the first
language we did, the books were instructions for the military, and there
were a number of things in English that the Vietnamese did not like.  One
thing that I remember in particular was that they did not like all the
passive constructions in our manuals "the hatch must be locked" and other
such.  It was really no big problem to find these passives, since they were
"commands" they usually came at the beginning of a sentence or clause, and
we found an acceptable Vietnamese expression and used it.  I would make the
point that such translations are not going to be either perfectly nuanced or
poetic.  Logos sold its system with the caveat that some post-editing would
be needed.  And for in-house needs, when they just wanted to get
information, not compose letters to send to their most honored customers, a
number of Logos customers were satisfied with raw output.  In the very
beginning, we had to decide what kinds of texts we would translate, and our
first choice was newspaper articles, because we figured people would want to
be able to get the news from all around the world right away, so they would
be satisfied with unedited output.  But in the end because of the contracts
we were able to land, at first they were all for our government, we ended up
translating technical manuals.  Even there, however, there was quite a bit
of "customization."  We had, for instance, Siemens, the IBM of Germany for a
customer, also Deutsche Banke, a chemical company, BMW was also our
customer, and we devised a dictionary system whereby either they could get a
dictionary tailored to their vocabulary or else they could throw a switch
and get, I think we had a choice of 3 available.  I had my own dictionary
when I had my own French system, it had the name of my initials, and since I
know French fairly well, I did a number of customized things.  For one
thing, our customers for French were Canadian, not French, so for the
technical words, or food like hot dog, etc., I used the non-French term.  So
as I say, I appreciate your approach, and this is not an argument on my
part, but just an attempt to put my experience at your service so as to
suggest that once you really get into such a system, there is quite a lot
more that you can do than you might think.  We had only one competitor in
the beginning, a company called Systran that got all the BIG government
contracts, NASA, for instance.  And they had to report on the great big
government reports, and of course we read their report.  And every time they
got more money from the government, they added another module to their
already huge system.  Whereas whenever we had a  breakthrough, we would
*eliminate* lots of rules, which would be replaced by simplified rules that
"saw" it all from a higher viewpoint, so to speak, and could do more with
less.  So I hope you will find this somewhat helpful, because there does
seem a good possibility of getting some kind of a project going.  Much love
to all, Laura
----- Original Message -----
From: David Sanchez <davius_sanctex@terra.es>
To: <aymaralist@aymara.org>
Sent: Wednesday, February 21, 2001 4:33 PM
Subject: RE: [aymara] Aymara and Western Culture

> Dos preguntas, Jorge:
> >    Unfortunately, it has been demonstrated that
> >    the Aymara would greatly facilitate the translation of any other
> >    idiom into its own terms, but not the other way around. Thus, because
> >    of its perfection, Aymara can render every thought expressed in other
> >    mutually untranslatable languages, but the price to pay for it is
> >    (once the perfect language has resolved these thoughts into its own
> >    terms), they cannot be translated back into our natural native
> >    Aymara is a Black Hole.
> Admitamos que exista una función f: L----->Ay definida sobre el conjunto
> de proposiciones de la lengua L y le asigna un equivalente semántico en
> Aymara.
> En estos términos podríamos decir acaso que el hecho de que "Aymara can
> render
> every thought expressed in other mutually untranslatable languages" es
> equivalente
> a que existe una función f como la anterior tal que (es una pregunta no
> afirmación):
> (i) f es suprayectiva (one-one)
> (ii) f no es inyectiva (one-to-one) y por tanto no admite inversa.
> Podría ser esta una buena caracterización de los problemas de
> intraductibilidad
> del Aymara a otras lenguas? O los problemas no van por ahí?
> david sanchez