Associate Professor Savithry Namboodiripad recently gave an invited talk at the 6th South Asian Forum on Acquisition and Processing of Language (SAFAL), held at the University of Colorado Boulder from November 17–19, 2025. This was the first time that a conference dedicated to acquisition and processing of South Asian languages was held in the United States. The conference featured work from researchers around the world, including North America, Europe, and South Asia. Dr. Namboodiripad’s keynote presentation was titled “Characterizing multilingualism in global South Asia: Clustering methods as a tool to investigate variation in processing and use,” and an abstract is given below.
Abstract: Many theories of language processing and acquisition treat multilingualism as a problem or nuisance variable, with consequences for how participants are recruited, between-group comparisons are set up, and results are analyzed and interpreted. Large South Asian languages in particular are global, and patterns of migration and communication mean that the division between diasporic and in situ language users grows ever more blurred. I present collaborative work which employs agglomerative cluster analysis, a dimensionality reduction tool, to create emergent profiles of language experience based on participants' language background surveys. These clustering methods allow us to reflect the global reality by finding some order in heterogenous language use. With Hindi-Urdu and Malayalam speakers, we find that these emergent language experience profiles outperform top-down groupings of participants (e.g., diaspora vs in situ, or generation of immigration) in explaining variation in constituent order acceptability and word choice respectively. I end with comparisons to two non-South Asian languages which have implications for acquisition: findings across languages and empirical domains suggest that current use is more predictive than early use when it comes to explaining variation from acceptability judgments to phonological perception and heterogeneity in language ideology. Though the conclusion might be straightforward — measures of use are more predictive than proxy variables — this method of characterizing multilingualism allows us to expand our empirical and theoretical scope as scholars of processing and acquisition.
