New approaches to lexical ambiguity: theory, methods, models, and measures
Abstract
Lexical semantics sits at the core of cognitive science, yet word meaning is not static: it is shaped by ambiguity, contextual constraints, and learning history (Rodd, 2020; Piantadosi et al., 2012; Trott & Bergen, 2023). A central challenge is to connect experimental evidence of how meanings are learned, accessed, and selected in real time, with computational accounts that scale to the lexicon and capture how senses emerge from language use (Haro & Ferre, 2018; Beekhuizen et al., 2021). This symposium brings together diverse researchers from the Global North and South who use complementary experimental paradigms and computational approaches to study meaning in context, including pupillometry, large-scale word association graphs, contextual embedding models of developmental change, and large language-model analyses of experimental stimulus variability. Together, the talks bridge computational and experimental perspectives on word meaning processing and word sense learning, offering converging insights into how lexical knowledge is represented, updated, and deployed across timescales. In so doing, it pushes back against classic, relatively siloed approaches to studying lexical semantics, highlighting how an interdisciplinary approach to these issues truly is more than the sum of its parts.