| Abstrakt: |
In the context of Borer's (2005) theory of nominal classification, the aim of this paper is to explain why measure words in some languages (English, French, Hebrew) necessarily take an -s (two bottles of milk versus *two bottle of milk) while in other languages (Azeri, Persian, Ojibwe) measure words can surface without plural marking (the equivalent of two bottle of milk is grammatical). If we assume -s in English-type languages is responsible for division in measure constructions (as in Borer 2005), we face the following puzzle: What is responsible for division in Azeri-type languages in the absence of the plural? We argue that, for a number of reasons, it cannot be the numeral (two) and propose that division is performed, in the absence of a plural, by measure words themselves (as in Chierchia 1998; Stavrou 2003; Acquaviva 2008, among others). We argue that whether or not plural marking appears on the measure word depends on a higher projection that expresses the counting function (distinct from the classifying/measuring function, Rothstein 2010b). Measure constructions thus provide evidence for the idea that, in addition to the dividing plural, we need a higher, counting plural, bolstering the hypothesis that the plural comes in many flavours (Acquaviva 2008; Harbour 2008; Wiltschko 2008, 2012; Butler 2012; Mathieu 2012, 2013, 2014). [ABSTRACT FROM AUTHOR] |
