Newsgroups: sci.lang
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!uunet!zib-berlin.de!math.fu-berlin.de!news
From: brose@inf.fu-berlin.de (Gerald Brose)
Subject: Re: Syntax in the Lexicon???
Message-ID: <0IPWBHZB@math.fu-berlin.de>
Sender: news@math.fu-berlin.de (Math Department)
Nntp-Posting-Host: schubert.inf.fu-berlin.de
Reply-To: brose@inf.fu-berlin.de
Organization: FU-Berlin, Informatik
References: <3g2sqoEnb5@uni-erlangen.de>
Date: Fri, 27 Jan 1995 15:44:01 GMT
Lines: 70


>>>>>> "Player" == Player  <smryan@netcom.com> writes:
>In article <smryanD2LtDE.DHH@netcom.com> smryan@netcom.com (Player) writes:
>
>
>Player> I'm thinking about trying to do a grammar where the syntax and
>Player> lexicon are the same. For example, you start with
>Player> +V+independent
>

It all depends whether you are aiming at a 100% integration of a language's 
syntactic description into the lexicon (not very likely if natural languages
are concerned) or whether you are just proposing a shift in emphasis, viz.
to describe a good proportion of the syntactic 'behaviour' of a lexical item 
in its entry in the lexicon. There have always been linguistic theories which
depend on a more 'elaborate' lexicon than others because of the amount of
syntactic information included there.

>[...]
>Player> What I was wonderring is if anybody has seen this done before
>Player> and do you have a reference if so? (And to be pedantic, please
>Player> give me the reference if you have it.)
>
>
>What you are proposing is similiar to what is currently done
>within the framework of unification-based grammar. 

Not only in unification-based grammar.

If you are interested in those kind of linguistic theories you might also want to
look at what is commonly subsumed under the term 'valence/valency'. Lucien 
Tesni`ere is generally credited with the introduction of the term 'Valence'
into linguistics (as borrowed from chemistry), and there have been a number
of valence theories since his 'El'ements de syntax structurale' appeared in 
1959. IMHO, an especially interesting, modern valence theory is being developed by
Joachim Jacobs in Germany where a lambda-calculus-like formal notation is
employed to capture syntactic as well as semantic features of lexical
items in what he terms 'Valenzrahmen' (valency frames). These valency frames
are then listed in the lexicon.

Just a few references for those who might be interested:

	J. Jacobs, Valenz im Lexikon, 1993.

	J. Jacobs, Syntax und Valenz, in L. Hoffman (ed.), Deutsche Syntax, 1991.

	Th. Herbst, Adjective Complementation, in: Applied Linguistics, Vol. 5, No. 1,
		1984. 

	(D. J. Allerton, Valency and the English Verb, 1982.)

For a short introduction to valency see:

	D. J. Allerton, Valency and Valency Grammar, in the Encyclopedia of Language
		and Linguistics, 1994.



Regards, Gerald.

____________________________________________________________________________

Gerald Brose                         Ph-one:  ++49-30-838-75112	
Freie Universitaet Berlin            Ph-ax:   ++49-30-838-75109		
Institut fuer Informatik             Mail:    brose@inf.fu-berlin.de
Takustrasse 9                        http://www.inf.fu-berlin.de/~brose/
14195 Berlin, Germany
____________________________________________________________________________


