Proceedings of the Annual Meeting of the Cognitive Science Society

This study investigates processing of conventional and novel metaphors in Marathi, an SOV language, with behavioural and electrophysiological measures, to extend graded models of metaphor comprehension established with SVO languages. Participants performed literal truth value judgment tasks with reaction time (RT), accuracy, and event-related potentials (ERPs) recorded. Behavioural results showed that literal sentences elicited the fastest RTs and the highest accuracy, contrasted by conventional metaphors' longer RTs and lower accuracy, reflecting decision-level conflict. Novel metaphors showed intermediate performance, indicating increased semantic processing demands with reduced conflict. ERP analyses of the N400 revealed greater centro-parietal negativity for novel metaphors relative to literal and conventional sentences, consistent with increased semantic integration difficulty, although ROI-level effects were not statistically significant. Overall, the findings support cross-linguistic robustness of graded model of metaphor comprehension, and how semantic processing varies as a function of familiarity by extending N400-based accounts of metaphor processing to Marathi.