Event Details

This event is co-organized by the NLT L2 Phonology Research Forum, the Multilingualism and Cognition Research Group, and the Heritage Language 2 Consortium (HL2C).

Is that really productive? Disruption to perceptual learning of a novel sound contrast

Alexandra Kapadia (University of Chicago)

Speech sound learning in a new language requires the learner to gain skill in both perceiving and producing novel sound contrasts. Theories of non-native speech sound learning propose that proficiency in the two modalities is acquired in tandem, supporting the notion that perception and production share representations that are refined through practice (Bradlow et al., 1997). A growing body of work, however, suggests that learning in both modalities is not necessarily coupled (e.g., Nagle & Baese-Berk, 2022). Further, perceptual learning of novel contrasts can even be disrupted when perception and production are trained concurrently (Baese-Berk, Kapnoula, & Samuel, 2025). In this study, we trained native English speakers on a Hindi-like dental-retroflex stop contrast. Participants were assigned to either a Perception-Only or a Perception+Production training condition, and all completed two days of training consisting of 8 blocks of ABX discrimination trials with feedback. Pre- and post-test discrimination blocks (no feedback) were administered each day, along with a separate production post-test (repeat the target sound). In addition to the discrimination task, participants in the Perception+Production group also produced the target sound after each discrimination trial during training. Perceptual learning of the novel contrast was demonstrated by participants in the Perception-Only training group, but not the Perception+Production training group, supporting previous reports of disruption to perceptual learning when perception and production are trained in tandem. Production across groups was compared to investigate whether acoustic correlates of the dental-retroflex distinction were present after training. Preliminary results do not suggest differences in distinction cues across the category boundary for either training group, indicating that production practice during perceptual training may not result in perception or production gains.

How to join this meeting:

We will circulate the Teams link via the usual mailing lists. To be added, please message p.rebuschat@lancaster.ac.uk.