Motor-based prediction mediates implicit vocal imitation

Area of Science:

• Psycholinguistics
• Neuroscience
• Speech Communication

Background:

• Phonetic convergence, the unconscious speech adaptation to an interlocutor, is linked to predictive mechanisms.
• Two prediction types exist: memory-based (reducing sensitivity to vocal identity) and motor-based (listener's vocal characteristics).
• The precise role of motor-based predictions in phonetic convergence remains unclear.

Purpose of the Study:

• To investigate whether motor-based predictions suppress or enhance sensitivity to listener-matched features.
• To determine the contribution of motor-based predictions to phonetic convergence.
• To explore the dynamic interplay between predictive signals and external input in speech adaptation.

Main Methods:

• A novel speaking oddball task with auditory mismatch negativity (MMN) as a neural index.
• Participants were divided into a shadow group (promoting vocal convergence) and a non-shadow group.
• Electroencephalography (EEG) was used to analyze neural responses.

Main Results:

• Motor-based predictions were found to enhance sensitivity to listener-matched predicted features after vocal convergence behavior.
• This enhancement correlated positively with the degree of vocal convergence observed.
• The findings challenge the notion that motor-based predictions solely suppress neural responses.

Conclusions:

• Motor-based predictions can enhance perceptual learning and guide vocal adjustments, not just suppress responses.
• These predictions orchestrate sensorimotor interaction and memory-based operations for implicit learning in social contexts.
• The study highlights a more dynamic role for motor-based predictions in mediating phonetic convergence.