[Corpora-List] corpus syntax (and how we can use it to code meaning)

[rodolfo delmonte]

Yorick,
the impression one gets from reading this thread and the ones related 
to it is not just that people in corpus linguistics - and perhaps also 
a good share of CL/NLP - are convinced that meaning of words can be 
recovered by means of KWIC, bigrams, and other similar lexically 
related measures.
My "hunch" is that semantics of words can only be drawn from semantics 
of propositions in which words are contained, not just a topographical 
notion, but a syntax related notion if you like. That is syntax and 
semantics - in that sense, not just lexical - is what determines 
meaning at last.
The question is that this is hard to come by with good approximation 
with - just - statistically based methods. We all look forward to see 
this new generation of CL/Corpus linguists who will be able to couple 
St.Methods, corpus measures and symbolic computation.
After all, I cannot give up the idea that it is our intuition as 
theoreticians that must guide us...
Best
Rodolfo

P.S.: I know this has already been discussed between us, but the list 
deserves more fuel...


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