Ayoub Daliri, Jonathan Dittman

To accurately produce speech, the central nervous system (CNS) predicts auditory consequences of its planned speech movements and compares its predictions with incoming auditory feedback. The CNS uses discrepancies between its predictions and incoming auditory information to modify its future output (adapt). While auditory errors are crucial for speech motor learning, not all perceived auditory errors are task-relevant. Therefore, the CNS needs to evaluate the relevance of perceived auditory errors. Here, we examined the relationship between relevance of auditory errors and extent of auditory-motor adaptation by systematically altering the relevance of auditory feedback. Participants (n=17) produced consonant-vowel-consonant words containing the vowel /ɛ/ (e.g., “head”) while their formant frequencies were shifted (80% shift) toward the vowel /æ/ and fed back to them in real-time. In one condition, auditory errors (i.e., received auditory feedback relative to planned production) depended on participants’ productions, and thus were relevant—participants could change their productions to reduce/change the perceived auditory errors. In another condition, auditory errors were pre-defined (regardless of the participant’s productions), and thus were irrelevant—changes in participant’s production did not result in change/reduction of the auditory errors. We found that participants adapted less and were less sensitive to auditory errors when the auditory errors were irrelevant. Additionally, we found that participants with smaller perceptual targets adapted more and had higher sensitivity to auditory errors. Together, our results highlight the intricate mechanisms—involving both perception and production systems—that the CNS uses to optimally integrate auditory errors for successful speech motor learning.