I'd just like to take this opportunity to recommend a highly relevant
and really excellent paper that addresses the relation between
statistical methods and generative linguistics, namely
Steven Abney (1996). Statistical Methods and Linguistics. In: Judith
Klavans and Philip Resnik (eds.), The Balancing Act: Combining
Symbolic and Statistical Approaches to Language. The MIT Press,
Cambridge, MA.
In my opinion the paper really should be read by *anybody* who is
interested in language, and it should certainly be forwarded to any
linguists who question the relevance of statistical methods.
Steve's abstract of this paper follows. Note that his use of
the term 'apology' is as a synonym for 'apologia', viz. "a formal
written defense of something you believe in strongly."
This is an `apology' for statistical methods, written for a linguistic
audience. The contents: 1. Language acquisition, language variation,
and language change; 2. Adult monolingual speakers;
2.1. Grammaticality and ambiguity; 2.2. Is this linguistics? 3. How
statistics helps; 4. Objections; 4.1. Stochastic models are for
engineers? 4.2. Chomsky v. Shannon; 5. Conclusion.
Cheers,
Philip
P.S. Just so you don't think I'm trying to get y'all to buy the book,
here are some links to Steve's paper on the Web:
http://www.sfs.nphil.uni-tuebingen.de/~abney/Abney_95c.ps.gz
http://crl.nmsu.edu/~raz/Ling5801/papers/stat/Abney_95.ps
http://www.cogsci.ed.ac.uk/school/study/msc/course/dil/papers/abney.ps.gz
P.P.S. On the other hand, it really is a very good book and everyone
should own a copy. Isn't Christmas coming up sometime soon? :-)
This archive was generated by hypermail 2b29 : Thu Apr 05 2001 - 00:34:05 MET DST