The 27th
PENN LINGUISTICS COLLOQUIUM  
KEYNOTE LECTURE A Compositional Characterization of the Edward Keenan, UCLA As a descriptive fact, the DPs which occur most naturally in ThereSs are (boolean compounds of) those built from intersective (and preferably cardinal) Determiners (notions I define in the talk): Cardinal
Int not Card
Nonint
In this talk I show how to derive the intersectivity restriction from a compositional interpretation of ThereSs which, in a sense I define, treats the postDP coda (in the class, etc.) as a domain restricting the interpretation of the DP. The compositionality problem I claim to solve is that of imposing the restriction on the N denotation in Det+N just given the Det+N denotation. Det+N might denote the same generalized quantifier (DP denotation) as some other Det'+N' even though N and N' may denote different sets. Time permitting I'll contrast my solution with Zucchi's noncompositional proposal (NALS 1995); more interestingly we'll address the question as to why the cardinal Dets and the DPs they build are more natural than the intersective noncardinal ones.
