By some accounts, speakers of classifier languages such as Mandarin or Japanese, which lack count-mass syntax, require classifiers to specify individuated meanings of nouns. This paper examines this view by testing how Mandarin speakers interpret bare nouns and use classifier knowledge to guide quantification in four experiments. Using a quantity judgment task, Experiment 1 found that Mandarin speakers interpret nouns like English speakers, regardless of their syntactic status as mass or count in English. Experiment 2 showed that Mandarin speakers quantified broken objects like English adults, again suggesting that Mandarin nouns specify criteria of individuation. Experiments 3 and 4 together showed that classifiers are not typically required for individuation, except when the reference of nouns is semantically ambiguous (e.g., rock, string) and can denote either objects or substances. In sum, we argue that individuation can be specified lexically in classifier languages like Mandarin, and does not depend on classifier syntax.