In this talk I show (based on joint cross-linguistic work with various collaborators) that distributive  — i.e., `classical’ — meanings for the universal part of the language (connectives and DP-internal quantifiers) correspond to bigger morpho-syntactic structures while the corresponding smaller structures have a weak (plural) meaning. This, I will argue, raises a problem for a number of current theories of scalar implicatures, as the `classical’ universal expressions required as alternatives are structurally too complex to be accessible. To tackle this problem, I will tentatively suggest an adjustment to the algorithm via which scalar implicatures are derived.

Location – Royce Hall 362