ABSTRACT
0. Introduction. The fragment given at the end of the last chapter is unsatisfactory in a number of respects. Although it does show how collective and distributive predicates can be conjoined without requiring a closure-condition theory of distributivity, it does not address at all the other sorts of evidence which have been adduced in favor of the closure-condition approach. For example, any account of distributivity which pretends to completeness would have to address the complex anaphora facts which Roberts (1987a, 1987b) has covered in her detailed closure-condition account of distributivity; Hoeksema's (1983, 1989) argument from subject- verb agreement and Scha's (1981) argument from predicates like contain and run parallel would also have to be addressed.
