Chinese nouns are mass nouns: An information-theoretic computational proof

The mass–count distinction in Mandarin Chinese has been heatedly debated in linguistics. Previous research investigated the mass–count distinction primarily through qualitative methods. Our study applied mutual information and created three variation conditions to explore the relationship between Chinese nouns and the individuation function of classifiers. Furthermore, we examined the mass–count distinction by quantitatively analyzing 1,000 instances of Num–CL–N (numeral–classifier–noun) structures. The computational results indicate that the individuation function of classifiers is influenced by noun homogeneity. Moreover, we argue that Chinese nouns exhibit an inclination for homogeneity and the deep semantic processing of nouns is similar to classes or sets, providing evidence for the broad mass noun hypothesis and the collective-name hypothesis in the philosophy of language.