Measurement and optional classifiers in Mandarin Chinese

This paper presents a compositional analysis of the fact that Mandarin individuating classifiers are systematically optional in various degree constructions (see also Lin & Schaeffer 2018 for experimental evidence), by taking a mixed approach incorporating the insights from Chierchia (1998; 2010) that Mandarin nouns denote kind terms and individuating classifiers offer the level of individuation and those from Krifka (1995) that (bare) numerals do not encode the cardinality function. By considering (bare) numerals as degree terms (e.g., Hackl 2001; Nouwen 2010; Rett 2014; Kennedy 2015, among many others), the mixed approach advocated here embraces the hypothesis that the locus of variation between English and Mandarin lies in neither the semantics of nouns nor that of numerals, but in the measure operators: these linguistic elements (including sortal/individuating classifiers) are necessary to mediate between numerals and nouns to avoid the semantic type-mismatch. The proposed analysis of individuating classifiers not only explains the role of Mandarin individuating classifiers in degree constructions (i.e., their syntactic optionality, along with a semantic variation in the dimension of comparison), but also closely connects with Bale & Barner’s (2009) idea about quantity judgments that comparative constructions can be used as a reliable diagnostic of the mass-count distinction in natural languages beyond English. Specifically, the fact that Mandarin unclassified nouns allow both cardinality and non-cardinality monotonic dimensions in a variety of degree constructions based on quantity judgments indicates that they are mass-count neutral; a tentative semantics of Mandarin nouns for their mass-count neutrality is thus suggested. Some factors leading to the individuation of nouns are also discussed.