LINGUIST List 4.676

Thu 09 Sep 1993

Disc: ARPA's Human Language Technology Program

Editor for this issue: <>

Directory

1. David Adger, DARPA, Science etc
2. Swann Philip, 4.667 ARPA's Human Language Technology Program
3. robert n shull, Re: 4.659 ARPA's Human Language Technology (HLT) Program

Re Vicki Fromkin's recent posting, similar attitudes seems to obtain
here in the UK. I know of two recent project proposals to the Science
and Engineering Research Council that were rejected on the grounds
that they would be unlikely to lead to the production of ``products'',
and that the basic research paradigms (formal models of discourse, and
situation semantics respectively) were unscientific. It is very
worrying that research councils who should be funding with a long term
view to constructing a base of scientific knowledge, are rejecting
proposals because they prejudge that certain enterprises are unlikely
to be of foreseeable _commercial_ benefit; and especially that problems
which are essentially linguistic in nature are not seen to be amenable
to scientific investigation and are therefore not funded.

I agree with Michael Covington that direct complaints from individuals
are unlikely to have much effect. What needs to be challenged is the
underlying attitudes of politicians who have a responsibility for
science to linguistic research, and this may be best done
by our professional organisations. In the Netherlands, the Dutch
linguistic community seem to do very well from the ZWO, maybe there's
something to be learned from this?

David

There are a bunch of education researchers who've started
calling themselves "learning scientists" working in a whole
new discipline called the "learning sciences". This is as
big a joke as "political science". To the man in the street,
"science" implies the ability to combine some mixture of
general laws of nature, the use of use of measurement,
the experimental method and precise predictions...

There can be no "fundamental science" of language because
language (viewed as a semiotic system) is not fundamental,
it's an historical and social process that cannot be
isolated from its matrix in any interesting or useful
way. You can use science to illuminate many aspects of
language (as you can apply science to archaeology) but
that doesn't mean you have a "language science".
Acoustics is a science, phonetics is mostly science but
phonology is not and never will be...

Science is no big deal. It's just one of the tools that
humans have stumbled across in their drive to dominate
Nature. To say that the study of language is not a science
is to say something *positive* about language and linguists.
I think this is why many real scientists (who are less than
starry eyed about what they do) are perplexed by the
desire others have to label anything and everything as
science...

Philip Swann
University of Geneva

Message 3: Re: 4.659 ARPA's Human Language Technology (HLT) Program

I'm a relative newcomer to the linguistics and cognitive science
scene, but I find the HLT message clear and reasonable. It doesn't
make use of our jargon, but if you are looking for the intended
meaning, the message (Vicki Fromm's excerpt) is not something that
warrants a protest - an interpretation perhaps, but not a protest.

The message says in too many words, -We want functional results, not
theories. We want to be able to use language, not just characterize it.-
The use of the term -fundamental science- can be read as -fundamental
research (the kind that answers questions that only linguists are
interested in).

I think a ground swell of protest would be all out of proportion to
what is basically an unfortunately worded paragraph. A more
appropriate response might be to reqest a clarification.

Yes Vicki, I would like to see the whole document.

Robert Shull
Grad Student
(Linguistics, Speech&Hearing, CogSci)
Indiana University
rshullsilver.ucs.indiana.edu