Open Journal Systems

Counting in the verbal domain

Susan Rothstein


This paper argues that the telic/atelic distinction cannot be reduced to a distinction between countable and non-countable predicates in the verbal domain. I show that telic and atelic predicates behave alike with respect to most countability tests and that both are cumulative. No verbal predicates, whether telic or atelic, can be directly modified by cardinal numerals, but all can be modified by a numeral together with the classifier time(s). I suggest that VPs do not denote countable sets. However, VPs, and in particular telic VPs, may denote sets of individuable events, making telic VPs similar to object mass nouns since the entities in the denotations of these VPs are individuable but not countable. Two general conclusions can be drawn from this. First, the contrast between atelic and telic predicates should be formulated in terms of constraints on individuability, not countability. Second, contrary to, e.g. Bach (1986), the mass/count distinction does not structure the verbal domain, and thus, in languages such as English, the nominal and verbal domains are very different.

Texto completo:

PDF (English)