satisfiability of (f FEATURE) =/= VALUE in coordination?

Helge Dyvik helge.dyvik at LLE.UIB.NO
Wed Oct 17 11:21:44 UTC 2012


We came across the same problem in our Norwegian XLE-implemented grammar, and noticed that a consequence of this property of XLE is that a seemingly exhaustive disjunction { P | ~P } is not necessarily exhaustive when sets are involved. I.e., if ! is a set, then the constraint

{ (! F)=c v
  …
| ~(! F)=v
 … }

does not cover the cases where only some of the set members satisfy (! F)=c v. Worth bearing in mind for the grammar writer.

Helge


Den 17. okt. 2012 kl. 13:04 skrev Mary Dalrymple:

> You could think of this as an instance of Wasow's Generalization, which Sadler and Nordlinger state as follows:
> 
> Wasow’s Generalization: An element in construction with a coordinate constituent must be syntactically construable with each conjunct
> 
> If an element is constrained by the grammar in some negative or positive way, then if that element is a coordinate structure, each of the conjuncts must be constrained in the same negative or positive way.  
> 
> - Mary
> 
> 
> 
> ________________________________________
> From: Lexical-Functional Grammar List [lfg at listserv.linguistlist.org] on behalf of Adam Przepiorkowski [adamp at IPIPAN.WAW.PL]
> Sent: 16 October 2012 17:50
> To: LFG at LISTSERV.LINGUISTLIST.ORG
> Subject: Re: [LFG] satisfiability of (f FEATURE) =/= VALUE in coordination?
> 
> Dear John,
> 
> Many thanks for your quick reply!
> 
>> (f CASE) ~= ACC is NOT SATISFIED in this case.  In effect, the
>> distribution gets wide-scope relative to the negation.  This seems to fit
>> the linguistic facts better than giving the negation wide scope.
> 
> We understand that this is the decision taken in the XLE *implementation*,
> but I've been wondering if there is any *theoretical* justification for
> or discussion of such scoping in the LFG literature?
> 
> All best,
> 
> Adam P.
> 
>> 
>> Cheers,
>> 
>> John
>> 
>> On 10/16/12 9:14 AM, "Adam Przepiorkowski" <adamp at IPIPAN.WAW.PL> wrote:
>> 
>>> Dear All,
>>> 
>>> It's probably a newcomer's question – apologies if so.
>>> 
>>> Imagine a coordination of two phrases, accusative and genitive (LFG
>>> literature knows such examples from Russian).  Assume – as usual – that
>>> CASE is a distributive feature.  What is the satisfiability of the
>>> following trivial statement for such a coordinate phrase f?
>>> 
>>> 
>>> (f CASE) =/= ACC
>>> 
>>> 
>>> Two answers:
>>> 
>>> 1. SATISFIED because 1) it's the same thing as "NOT ((f CASE) =c ACC)",
>>>  and 2) "(f CASE) =c ACC" is not satisfied (it is not satisfied for
>>>  one of the conjuncts),
>>> 
>>> 2. NOT SATISFIED because "(f CASE) =/= ACC" is not satisfied for one of
>>>  the conjuncts.
>>> 
>>> XLE seems to favour the latter answer, but – given how negation is
>>> defined in LFG (e.g., Dalrymple 2001, pp. 111–112) – we would expect the
>>> former.
>>> 
>>> It seems that we are essentially asking about the relative scope of
>>> negation and distribution.
>>> 
>>> All best,
>>> 
>>> Adam Przepiórkowski
>>> Agnieszka Patejuk
>>> 
>>> --
>>> Adam Przepiórkowski                          ˈadam ˌpʃɛpjurˈkɔfskʲi
>>> http://clip.ipipan.waw.pl/ ____ Computational Linguistics in Poland
>>> http://jlm.ipipan.waw.pl/ ___________ Journal of Language Modelling
>>> http://zil.ipipan.waw.pl/ ____________ Linguistic Engineering Group
>>> http://nkjp.pl/ _________________________ National Corpus of Polish
>> 
>> 
> 
> --
> Adam Przepiórkowski                   ˈadam ˌpʃɛpjurˈkɔfskʲi
> Zespół Inżynierii Lingwistycznej _____ http://zil.ipipan.waw.pl/
> Narodowy Korpus Języka Polskiego _____ http://nkjp.pl/
> Journal of Language Modelling ________ http://jlm.ipipan.waw.pl/
> Polska Lista Językoznawcza ___________ http://tnij.org/pling

--------------------------------------------------------------------
Professor Helge J. Jakhelln Dyvik
Institutt for lingvistiske, litterære og estetiske studier
Faggruppe for lingvistiske fag
Universitetet i Bergen 
Sydnesplassen 7  	Tel.: +47 55582261
5007 Bergen              	E-post: helge.dyvik at lle.uib.no
--------------------------------------------------------------------



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/lfg/attachments/20121017/32f61cf4/attachment.htm>


More information about the LFG mailing list