Consolidation during sleep stabilizes non-native speech sound learning

Abstract

Learning to perceive non-native speech sound categories is a challenge for many adult listeners. Even after successful learning, individuals vary in the degree to which they can successfully retain and consolidate learned perceptual categories in memory. In this talk, I describe a recent line of work from our lab showing that sleep plays a greater role in perceptual learning of non-native speech sounds than previously supposed. Specifically, we find that consolidation processes during overnight sleep protect learned phonetic information from interference from similar sounds from one's native language. I will discuss the potential sources of native language interference by reviewing a series of studies that test the low-level acoustic and abstract phonological contributions to interference. Finally, overnight gains in performance can be specifically linked to the duration and quality of sleep, suggesting a key role for sleep, per se, in the retention of non-native speech categories over multiple sessions. Taken together, this line of research suggests that individual variability in non-native speech sound learning may in part be linked to individual differences in sleep behavior and differences in the degree of pre-sleep exposure to competing native-language phonetic contrasts.