[Corpora-List] Moving Lexical Semantics from Alchemy to Science

[Ken Litkowski]

On 1/29/2011 9:28 AM, Adam Kilgarriff wrote:
> Ken says:
> > It would be nice if we could apply John's rules to Ted's compounds 
> and *put those findings into a dictionary*
> > (lexicographers have only barely done so, while lexicologists need 
> that information).
>
> Good news is, Dante (http://webdante.com) has done something very like 
> this in a v large exercise on corpus-driven lexicography for English. 
>  It contains 20,844 compounds, all identified automatically and 
> confirmed and classified by lexicographers..  Of these, 2,989 have 
> more than one meaning.

I think Adam has missed the point and Dante won't really fill need. In 
the dark ages, particularly Ray Jackendoff, a notion of lexical 
redundancy rules arose. When we look at John Sowa's analysis of "sensor" 
with a lambda function, I think we're seeing the operation of such a 
rule. These rules are not encoded in a dictionary and I believe they 
should be. Looking at definitions shows that they may be such rules, 
e.g., a gauge is "an instrument that measures and gives a visual display 
of the amount, level, or contents of something". In this definition, 
"something" plausibly can be John's "n" in a noun compound, so that we 
have entries for fuel gauge, rain gauge, wind gauge, etc. These rules 
should be part of the lexicon. I believe Dan Flickinger did this in 1987.

In addition to the sheer interest in having such rules, I think they may 
also help in achieving the goal of this thread. With such rules, we can 
*predict* the meaning of a noun compound. How well is precisely the 
stuff of science.

I do not mean to diminish Dante. We can justifiably look forward to its 
availability, both for research and for commercial applications.

     Ken
<http://localhost/cgi-bin/onlineOde/print_entry.cgi?id=049971632&match_point=8,13&lemma=gauge&right_column_mode=labels&caption_style=short> 


-- 
Ken Litkowski                     TEL.: 301-482-0237
CL Research                       EMAIL: ken at clres.com
9208 Gue Road                     Home Page: http://www.clres.com
Damascus, MD 20872-1025 USA       Blog: http://www.clres.com/blog

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/corpora/attachments/20110130/10f6e2a8/attachment.htm>
-------------- next part --------------
_______________________________________________
Corpora mailing list
Corpora at uib.no
http://mailman.uib.no/listinfo/corpora