[Corpora-List] (continued)

Tue Aug 19 12:29:46 UTC 2008

Thanks to Lou for his reminder that those who teach and translate are not to be
sneered at, after all the discipline owes much to the inspiration of teachers.
Palmer and Hornby, who did much for lexicographer, taught English, Firth was
originally a historian, who found himself teaching English. The Birmingham
group, and its off shoots, have been heavily involved in teaching for the
greater benefit of the world. The BNC with its accompanying tools has made
corpus linguistic methodology available to many teachers, who have improved
their skills as a result. I was introduced to corpora by Peter Roe  at Aston. He
changed my teaching outlook, and my career.

The fact of using corpora does not make a corpus linguist, corpora can be used
for many reasons. Like many I came in through teaching ESP. Dissatisfied with
methods and dictionaries I turned to corpora. I have become involved in the
world of lexicography, as although dictionary making works under tremendous
commercial constraints, it is a world full of people who are able to listen.

Translators may not necessarily use corpora in their daily lives,, but they
derive great benefit from the corpus studies on translation. One has only to
cite people as Mona Baker, Deborah Kenny and Jennifer Pearson to see the
importance of corpus linguistics to translation studies. A translator needs to
understand dictionaries, but that does not mean they have to actively do
lexicography. The same is true of corpus linguistics. The skills they know about
will serve them to be better translators. I work in a university where 'experts'
teach people to translate without corpora and with no understanding of
dictionaries, I also teach for another where corpora and dictionaries are
considered. Who produces the better translators? The job market will provide the
answer.

I used to teach ESP. I know longer do, but I do teach people who will manage
data, teach, and not only English. I and they believe that corpus linguistics,
and my research in the field, does bring something to them, even if they will
not become corpus linguists as such. I know many who teach ESP who research and
use corpora in their teaching. Corpus Assisted Language Teaching is a thriving
field, should we write it off as not sufficiently interested in the toolbox?

I drive a car, but I have a garage who looks after the machine. I just do the
clever bit of avoiding trees, boar, bores and anything else on my country roads.
It is the same with my computer. Hardy Heron gives me an environment, but
technicians and systems engineers will put my machine write when things go
wrong, or advise me when lost. Like most corpus linguists I use tools, I make
suggestions that help improve tools, and I respect those who dedicate themselves
to making the tools. We simply work together.

What is corpus linguistics FOR? One answer is to look at Susan Hunston's book on
Corpora in applied linguistics, unless of course you look down on applied
linguistics. She details some applications, but there are many more as corpus
linguistics is about language in society.

I am involved in lots of projects where meaning is essential. I have been
involved in medical and legal studies. My old speciality was scientific text,
teaching writing skills and adapting dictionaries. I currently work a lot with
political scientists, politicians, sociologists and historians. Meaning is
important to them. Funnily enough I work a lot with NLP as although I no longer
programme anything, I work with people who do and who actually see a need to
combine human applications with machine development? 

I shall also weigh in on behalf of Bill Louw. He also teaches of the benefit of
others. He gave us semantic prosody, an insight into language which is far from
being exhausted as a field of study. I could say that it is better to have one
big idea than none at all, but I won't as Bill has had many brilliant ideas on
collocation, which, as Firth and Sinclair showed is an area central to language
and not one to be simply reduced to formulae for the benefit of machine
translation tools and dictionaries. Understanding collocation is part of
understanding language, surely that is a useful exercise?

I also run conferences. We now do one every too years. The aim is to bring
together people who work with corpora whether they be teachers, translators,
linguists or NLP researchers. We believe in sharing ideas rather than claiming
that one approach is more useful than another. The proceedings are on-line
(http://web.univ-ubs.fr/corpus/publi.html) as we also believe in sharing
knowledge widely and freely. They are in French, but fortunately corpus
linguistics can be done in languages other than English.

Not watering down, or sneering at, corpus linguistics is the way forward. We
need to compare approaches and findings, not arbitrarily consider that one is
superior to another. Good fences make good neighbours, but I take a prototypical
view of borderlines where you can be central to a discipline, but still take an
interest in the periphery as one blends into the other on a continuum of knowledge. 

To sum up. This debate has/is airing very interesting views on what corpora are
for, and what corpus linguistics is about. This is after all corpora list and
not computational linguistics list, not that I would not be happy to subscribe
to that too.

Best

Geoffrey

Surlignage Andrea Mulloni <andrea2 at wlv.ac.uk>:

> Lou is theoretically correct, but...
> 
> There is a whole lot of translators and interpreters out there that  
> claim to be using corpus linguistics as a translation aid, but in  
> fact they just use concordancers as a fall back solution to missing  
> dictionary entries. They don't really know what corpus linguistics is  
> about.
> 
> The fact is - and being originally a translator myself, I feel  
> obliged to do a little bit of self-criticism here - that translators  
> and language teachers are not yet able to recognize the real  
> boundaries and potential of computational linguistics, often using  
> corpus linguistics terminology as a mere embellishment of their CVs.  
> So no, I don't think everyday translators really make (proper) use of  
> corpus linguistics. My BA thesis in translation and interpreting  
> dealt with a really really basic introduction to corpus linguistics  
> and its use in translation studies, and that was considered a quite  
> good job back then. Looking back at it now, I realize how little I  
> knew about the interaction of language and computers and how much  
> mathematics can actually contribute to the description of a language.
> 
> On the other side, I think that corpus linguistics can very well be  
> of help to computational linguists. Yorick mentioned the WAC  
> initiative. Any corpus linguist would probably ask questions about  
> the representativity of the corpus, which we computational linguists  
> often take as granted. We often think that numbers and source  
> compatibility suffice, but implementing representativity is much more  
> difficult than that. And we all know how important a well-balanced  
> corpus is to a computational linguist. So here you have it, language  
> engineering is like apples: you never throw away anything.
> 
> As far as Dan Melamed's assertion about absence of meaning: if it  
> were really so, what are we all doing here?
> 
> Just my two pennies.
> 
> 
> Andrea
> 
> Andrea Mulloni
> part-time PhD Student
> University of Wolverhampton
> 
> ---
> 
> That was a hard test Marcie. I didn't know if it was an essay test,  
> True or False, or multiple choice. I just put down "Not Guilty".
> 
> 
> 
> On 19 Aug 2008, at 12:19, Lou Burnard wrote:
> 
> > Lou (who really shouldn't post to public lists before having had his
> > second coffee of the day) says "Sadly, none of [the language-teaching]
> > community seems to have seen fit (yet) to contribute to the present
> > discussion. "
> >
> > Geoffrey Williams is very definitely a member of that community,  
> > and so
> > I believe is Dr Louw. Apologies to both for misrepresenting their
> > respective allegiances.
> >
> > I stand by my assertion that the debate hasn't taken fully into  
> > account
> > the transformative effects of corpora and corpus-methods in that
> > specific discipline though.
> >
> > _______________________________________________
> > Corpora mailing list
> > Corpora at uib.no
> > http://mailman.uib.no/listinfo/corpora
> 
> 

-- 
Geoffrey Williams, MSc, PhD
Professeur en sciences du langage
Directeur du département d'ingénierie du document
Université de Bretagne Sud - Faculté LSHS
4 rue Jean Zay
BP 92116
56321 LORIENT CEDEX
FRANCE

Tél: 33 (0) 2 97 87 29 20

--------------------------------------------------------------------------------
Université de Bretagne sud                               http://www.univ-ubs.fr/

_______________________________________________
Corpora mailing list
Corpora at uib.no
http://mailman.uib.no/listinfo/corpora