To listen and to talk: Auditory M100 response shifts posteriorly when perceiving phonemes before speaking

International audienceOne of the most fundamental questions in speech perception research is how properties of the acoustic speech signal are mapped to linguistic elements such as phonemes. Distinct theories have been proposed to answer this question. A crucial distinction among these theories can be put in the form of a simple question: does the speech motor system have a role in speech perception? From this view, recent studies postulate that the anterior auditory cortex "what" processing pathway would be involved in acoustic-phonetic decoding while posterior auditory cortex "where/how" stream would underlie a sensorimotor mapping between auditory representations and articulatory motor representations. In return, this motor-related activity is hypothesized to constrain phonetic interpretation of the sensory inputs through the internal generation of candidate articulatory categorizations. Consistent with such perceptual-motor interactions in speech perception, we hypothesized that the "where/how" processing pathway would be more engaged when perceiving speech stimuli before producing them compared to passively listening to the same stimuli. Using magnetoencephalography we tested whether equivalent current dipole (ECD) source location estimate (which approximates the center of gravity of neural activity) of the so-called M100 response recorded about 100 ms from the auditory speech stimulus onset would shift posteriorly when subjects are perceiving phonemes and subsequently perform a speech production task, compared to a pure passive perception task. Ten healthy volunteers were presented the same syllables with two levels of ambiguity (presented with or without auditory noise) in four different conditions: passive perception, passive perception and overt repetition, passive perception and covert repetition, and passive perception and overt imitation. In the three last 'motor' conditions, the task of the subjects was to perceive the phoneme first, then wait for visual signal, and perform the speech production task. Compared to the passive speech perception condition, results showed a significant shift of the ECD-estimated location of M100 response to the phoneme sounds to a more posterior position in the left hemisphere during the motor tasks. This demonstrates that perceiving speech before speaking induces a stronger involvement of the "where/how" processing pathway and therefore suggests that sensorimotor interactions during speech perception are dependent on the exact content of the task