Plural sentences with conjunctive predicates often receive a ‘non-boolean’ interpretation. E.g. the men are swimming and crawling is interpreted as true when only some of the men are swimming, and the rest of them are crawling (Krifka 1990). However, whether this kind of interpretation indicates a general non-boolean reading of conjunction remains unclear, since as often as not, the interpretation of such a sentence is strictly ‘boolean’ (cf. the men are swimming and smiling). A challenge for the semantics of plurals is to account for these different interpretations, in which properties are distributed differently over individual men. Winter (2001) extends the Strongest Meaning Hypothesis of Dalrymple et al. (1998), which was initially proposed to account for the varying logical interpretations of reciprocal sentences. According to Winter’s proposal, a plural sentence with predicate conjunction receives a strong, boolean interpretation unless this contradicts properties of the complex predicate, which results in a weaker interpretation.
I challenge this prediction by showing experimentally that for a large set of sentences, there is a continuum of acceptability values for non-boolean interpretations. I account for this continuum with a principle that predicts how language users apply predicates to plural subjects, resolving vagueness systematically. For any given pair of predicates, the principle predicts the divergence from a boolean interpretation based on the typicality structure of the complex predicate: the less typical the complex predicate is considered to be in a boolean interpretation, the weaker the interpretation of the sentence becomes.
I will discuss this work in parallel to previous experimental work on reciprocal sentences, to which a similar reasoning applies. These results combined imply sensitivity to the conceptual structure of predicates as a basic mechanism of resolving the vagueness of distributive quantification with plural predication.