TOPICS IN THE FORMAL SEMANTICS OF THE NOUN PHRASE WINTER

[Jump to Bibliography]

COURSE DESCRIPTION

Traditional logic developed a fine-grained account of a small number of noun phrases like those given in (1) below. However, many phenomena that are of central importance for the semantics of natural language were by and large ignored by traditional logical frameworks. For instance: plural, bare, mass, partitive, and coordinate noun phrases as in (2) are highly common in all languages, but logical semantics had little to say about these constructions. (1) some painter, every painter, Picasso, the painter (2) no painters, painters, much paint, most of the painters, neither the painters nor the musicians Montague's pioneering work changed this situation dramatically. Much research in the semantics of natural language since the early 70s has been devoted for extending systems of quantification so to treat phenomena in the interpretation of the noun phrase that were not substantially addressed by traditional logic.

The aim of this course is to give an overview of current work on NP semantics and to discuss the implications of this work for the overall architecture of semantic theory. We will especially concentrate on three empirical domains: coordination, plurality and scope. The course will open by a general discussion of the objectives and methodologies of natural language semantics. I will argue for a conservative truth-conditional theory that is empirically motivated mainly by robust phenomena of entailment, and less by weaker linguistic intuitions like ambiguity or presupposition, for instance. The best methodology known for natural language semantics is argued to be modeltheoretic (=based on abstract algebraic constructs rather than the syntax of logical formalisms) and compositional (=directly interpreting well-motivated syntactic representations). The Boolean Semantics of Keenan and Faltz (1985) and its sub-field of Generalized Quantifier Theory (Keenan 1996) will be introduced as the foundations of such a theoretical framework.

Working within this framework, we will attempt to answer three questions about the semantics of the noun phrase. I will give an overview of previous approaches to these questions and will introduce new solutions based on recent work.

Question 1:

What is the origin for the apparent double nature of "and"? For instance, why does the predicate in a sentence like "Mary and John smiled" distribute to the conjuncts in the subject while in the sentence "Mary and John met" the subject behaves as one semantic unit?

Proposed answer:

There is only one use of "and", the Boolean one. Ambiguity is generated by a flexible interpretation procedure following Partee and Rooth (1983) and Partee (1987) that is responsible for the "collective" reading of "and".

Question 2:

What does the scope of indefinites like "some student" appear to be so free when compared to other NPs like "every student"?

Proposed answer:

The flexible interpretation of the noun phrase involves a choice function mechanism following Reinhart (1997), which enriches the simpler modes of quantification used in generalized quantifier theory.

Question 3:

What are the properties of the procedure that generates distributivity effects with plurals?

Proposed answer:

Two distributivity operators, at the NP level and at the VP level, are responsible for these effects. These operators are atomic (range over singular individuals) and unary (apply to only one argument at a time).

The course will end with a general overview of the system of Flexible Boolean Semantics that emerges from the proposed solutions.

PREREQUISITES

Winter, Yoad (1996)  [PDF]

"A Unified Semantic Treatment of Singular NP Coordination." Linguistics & Philosophy, No. 19.

Winter, Yoad (1999)  [PDF]

"Dependency and Distributivity of Plural Definites." De Dag, proceedings of the workshop on definites.

Winter, Yoad (1997)  [PDF]

"Choice Functions and the Scopal Semantics of Indefinites." Linguistics & Philosophy, No. 20.