As to the discussion on whether a linguist should be able to program, I
would like to point out the following. When one needs to solve a problem
that needs both computational and linguistic knowledge, there is no harm
in having skills in both domains. It can only increase comprehension of
the problem involved. That's also one of the reasons why there is a
research domain called "Computational Linguistics". Of course, not all
linguistic (NLP) issues demand an equal computing or theoretical effort,
but in general, the discussion between a computer scientist and a
linguist gets much more fruitful if both of them know what kind of work
the other is doing.
I state this from my own experience in the domain, starting out by
language studies, and learning how to program, mostly autodidactically.
Learning to program is certainly a hard job, but it is mostly a matter
of motivation. If you really want, you can even get to implement fairly
complicated NLP tasks (A modest remark: as an exercise, I have
implemented an approximate string search approach). The more experienced
you get in programming, the more things you can try out yourself, and
the better you'll understand how other tools work and how they may be
improved. The corpus linguistics field is highly interesting as to that,
with all its diverging (and overlapping) tools.
I hope this may contribute to the discussion,
Tom.
-- LANT nv/sa, Research Park Haasrode, Interleuvenlaan 21, B-3001 Leuven mailto:Tom.Vanallemeersch@lant.be Phone: ++32 16 405140 http://www.lant.be/ Fax: ++32 16 404961