Jakub Dotlačil, UC Santa Cruz

Distributivity: Its Sources and Incremental Interpretation

Many plural expressions can be interpreted in several ways. Consider, for example, the plural expression `my friends', which can refer to the individuals forming the plurality (the distributive reading), as in the sentence `My friends are tall', or which can refer to the group and not the individuals (the collective reading), as in the sentence `My friends outnumber my enemies'.In this talk I discuss my recent joint work with Adrian Brasoveanu, which studies the incremental interpretation of sentences with plural expressions and the on-line resolution of distributive and collective readings. I will report the results of three eye-tracking experiments which investigated: (i) whether plural definites are immediately interpreted as collective or as distributive; (ii) whether the choice of interpretation is solely driven by pragmatics; (iii) whether there is a difference between the processing of quantificational / scopal distributivity (e.g., introduced by the floating quantifier `each') and lexical distributivity (e.g., introduced by `tall' as a lexically distributive predicate). I will argue that plural definites are by default processed as collective and that this preference is not caused by pragmatic plausibility or semantic complexity. Furthermore I will show that quantificational and lexical distributivity are processed differently, and argue that the difference in processing should be reflected in the formal semantics analysis. Finally, building on the work of Frazier et al. (1999), I will claim that the results show us that the distinction between collective and distributive readings is a case of ambiguity but only when we limit our attention to quantificational distributivity.