Maximality and Definite Plurals - Experimental Evidence

Abstract

Statements with plural definite descriptions are often assumed to come with a maximal interpretation, requiring all the individuals meeting the description to have the property expressed by the predicate in the sentence. However, there clearly is some variability with respect to this requirement, as non-maximal interpretations seem to be possible in various circumstances as well. This paper presents experimental evidence informing the relationship between maximal and nonmaximal interpretations of definite plurals. I suggest that the results are best captured by a view that sees the semantics of plural definites as involving maximality, but which allows for non-maximal interpretations, even with distributive predicates, by some type of pragmatic mechanism. Various possible choices of such mechanisms are discussed, though the question of which one is best suited to account for the data is not conclusively resolved. However, the advantages over other accounts, which assume a non-maximal semantics combined with pragmatic strengthening when needed to derive maximal readings, with respect to the present results seem rather clear.