Wu and Bodomo (2009) dispute Cheng and Sybesma's (1999, 2005) analysis of Chinese numeral classifiers as being able to function as definite articles. While I agree with Wu and Bodomo's overall conclusion, here I focus on parts of their argument that do not stand up to scrutiny and distract from their main point. In particular, I argue that Wu and Bodomo's conclusion that numeral classifiers are lexical items is incorrect. Also, I question their apparent conclusion that the availability of numeral classifiers in a language indicates that that language's nouns are inherently mass.