On the origin of the effect of silent articulation on speech perception

Silent articulation influences the identification of ambiguous speech sounds. This effect has often been attributed to articulatory suppression – the idea that concurrent motor activity prevents the recruitment of motor resources that would normally support speech perception. However, silent articulation also generates motor, somatosensory, auditory, and phonological signals, any of which could potentially contribute to the effect. The objective of the three experiments reported here was to discriminate between articulatory suppression and these other potential origins of the silent-articulation effect on speech perception. To do so, we determined the minimal signal-to-noise ratio at which participants could discriminate between two pairs of syllables (labial/lingual and lingual/lingual) while repetitively moving the finger or performing a lip movement, either opening and closing the lips (Experiment 1), protruding and retracting the lips (Experiment 2) or both (Experiment 3). Categorization performance for the two pairs of syllables during finger movement provided the baseline against which a putative effect of the lip movements on the labial/lingual pair was tested. The two lip movements engaged similar motor resources (contraction of the orbicularis oris). Only lip opening engaged motor planning and somatosensory feedback similar to those associated with one of the syllables used in the task. Neither of them engaged auditory or phonological factors. Only lip opening differently influenced the two types of syllables in comparison to the baseline. These findings suggest that our own mouth movements shape what we hear only when they resemble articulatory gestures of possible speech sounds. This challenges explanations based solely on motor-resource depletion or on phonological or auditory imagery and helps clarify how the motor system interacts with, but is not necessarily required for, speech perception.