There are two major theories regarding the deep difference is between classifier (CL) and nonclassifier (non-CL) languages. Krifka (1995) suggests that CL and non-CL languages differ in the semantics of their numerals, while Chierchia (1998) argues that CL and non-CL languages differ in the semantics of their NPs. Bale and Coon (2014) have shown that intra-linguistic variation in Mi’gmaq and Chol bears out Krifka’s predictions, but they left it open whether intra-linguistic variation in other languages support Chierchia's approach. In the empirical contribution of this talk, I show that this is indeed the case in several languages, including but not limited to Colloquial Khmer, Bangla, Classical Chinese, Malay, Vietnamese, Nùng and Hungarian. In the theoretical contribution I will put forth a proposal as to now N or Num can block the appearance of a CL in a classifier language, and explore the repercussions for constituency in the NP, arguing that N and Cl form a constituent to the exclusion of Num.