Kind Reference in Classifier Languages
This project investigates how kind reference is encoded in classifier languages within the Indo-Aryan family, focusing on Bangla, Odia, and Assamese—a small but striking areal subgroup that contrasts with most Indo-Aryan languages in possessing classifier systems. My research examines the nominal structure of these languages, with special attention to the behavior of bare nouns and the role of classifiers in licensing definite, indefinite, and kind-level interpretations. Although bare nouns in these systems share features with canonical kind-oriented languages like Mandarin, their full distribution diverges in systematic ways, revealing a more nuanced typology of kind reference.
Focusing on Bangla, I argue that bare nouns are best analyzed as singular kind terms, an account supported by the behavior of ra, an animacy-restricted classifier that functions as a lexicalized type-shifter from singular kinds to properties. This analysis captures the language’s extensive use of singular kind reference while explaining its restricted distribution. Ongoing work extends this approach to other classifiers in Bangla (ta and gulo) and integrates new empirical data from fieldwork on Odia and Assamese. Together, these findings broaden the empirical foundations for formal models of kind reference and highlight the theoretical significance of South Asian linguistic variation, which remains underrepresented in semantic typology.