Title
Universality and individuality: the interaction of noun phrase determiners in copular clauses
Abstract
This paper presents an implemented theory for quantifying noun phrases in clauses containing copular verbs (e.g., 'be' and 'become'). Proceeding from recent theoretical work by Jackendoff [1983], this computational theory recognizes the dependence of the quantification decision on the definiteness, indefiniteness, or classness of both the subject and object of copular verbs in English. Jackendoffs intuition about the quantificational interdependence of subject and object has been imported from his broader cognitive theory and reformulated within a constraint propagation framework. Extensions reported here include the addition of more active determiners, the expansion of determiner categories, and the treatment of displaced objects. A further finding is that quantificational constraints may propagate across some clausal boundaries. The algorithm is used by the RELATUS Natural Language Understanding System during a phase of analysis that posts constraints to produce a 'constraint tree.' This phase comes after creation of syntactic deep structure and before sentential reference in a semantic-network model. Incorporation of the quantification algorithm in a larger system that parses sentences and builds semantic models from them makes RELATUS able to acquire taxonomic and identity information from text.
Year
DOI
Venue
1985
10.3115/981210.981215
ACL
Keywords
Field
DocType
noun phrase determiner,constraint tree,copular clause,posts constraint,copular verb,constraint propagation framework,quantificational constraint,quantification algorithm,relatus natural language understanding,broader cognitive theory,quantification decision,computational theory,semantic model,noun phrase,semantic network,computability theory,constraint propagation
Definiteness,Noun phrase,Determiner,Local consistency,Computer science,Natural language understanding,Artificial intelligence,Natural language processing,Universality (philosophy),Linguistics,Syntax,Theory of computation
Conference
Volume
Citations 
PageRank 
P85-1
0
0.34
References 
Authors
4
1
Name
Order
Citations
PageRank
John C. Mallery1182.85