Proceedings of the Annual Meeting of the Cognitive Science Society

Experimental paradigms that perturb auditory feedback induce two distinct but related behavioral responses: (1) compensation, a real-time, within-token, adjustment to speech, and (2) adaptation, an adjustment to speech that persists even after the removal of perturbation, evidencing learning (e.g., Houde & Jordan, 1998; Purcell & Munhall, 2006a, 2006b; Tourville et al., 2008). In this paper, we present a fully dynamic model of one-shot adaptation, i.e., learning after one trial of perturbation (Hantzsch et al., 2022), in the framework of Dynamic Field Theory (DFT; Schöner & Spencer, 2016). The model triggers compensation and adaptation via prediction-perception mismatch driven by lexical-item-based predictions. Compensation derives from resolving multiple inputs into a neural field and via update to the mapping between phoneme representations and production targets, while adaptation relies only on the latter. We discuss how the model, which integrates speech motor control with linguistic representations, accounts for established results while generating novel predictions.