A cross-language study of compensatory response to formant-shifted feedback

Learning new sounds in a second language requires the acquisition of new motor routines and new sensorimotor planning systems needed to ensure coordination. Auditory feedback is an important part of the planning and control system required for fluent speech production. ESL vowel production was studied using a real-time formant perturbation technique to modify auditory feedback. Three groups of subjects (Native English, Japanese ESL, and Korean ESL) produced tokens of the English word “Head” with the first formant (F1) shifted either up or down in frequency. When F1 was shifted up, compensations by Native English speakers were larger than either ESL group. The F1 lowering perturbations produced more similar compensations by all three groups. This direction asymmetry in magnitude of compensation is discussed in relation to differences in native vowel inventories and the nature of auditory feedback processing.