Listener-Speaker Perceived Distance Predicts the Degree of Motor Contribution to Speech Perception

Abstract

Listening speech sounds activates motor and premotor areas in addition to temporal and parietal brain regions. These activations are somatotopically localized according to the effectors recruited in the production of particular phonemes. Previous work demonstrated that transcranial magnetic stimulation (TMS) of speech motor centers somatotopically altered speech perception, suggesting a role for the motor system. However, these effects seemed to occur only under adverse listening conditions, suggesting that degraded speech may stimulate listeners to adopt unnatural neural strategies relying on motor centers. Here, we investigated whether naturally occurring interspeaker variability, which did not affect task difficulty, made a speech discrimination task sensitive to TMS interference. In this paradigm, TMS over tongue and lips motor representations somatotopically altered the discrimination time of speech. Furthermore, the TMS-induced effect correlated with listeners' similarity judgments between listeners' and speakers' speech productions. Thus, the degree of motor recruitment depends on the perceived distance between listener and speaker. This result supports the claim that discriminating others' speech pattern requires the contribution of the listener's own motor repertoire. We conclude that motor recruitment in speech perception can be a natural product of discriminating speech in a normally variable and unpredictable environment, not merely related to task difficulty.

The effect of speech distortion on the excitability of articulatory motor cortex

“…Importantly, however, we also present data that go beyond the observation of group effects, as we highlight how the relationship between motor activation and perception may operate on an individual level, complementing previously observed relationships between behaviour and the speech motor system (Bartoli et al, 2015;D'Ausilio et al, 2014). Here, we found that individuals were highly varied in their ability to identify distorted speech sounds and that this corresponded to the level of activity in their speech motor system during speech perception.…”

“…In addition, the presence of somatotopic facilitation during distorted speech findings that have shown somatotopic TMS effects reflected in differences in behavioural performance when using non-distorted speech (Bartoli et al, 2015;Möttönen and Watkins, 2009). The high degree of stimulus predictability and certainty, and/or lack of an active listening task to moderate stimulus engagement may have resulted in the absence of somatotopic effects during perception of natural speech, and modulation of the motor system more generally (Hickok et al, 2011;Möttönen and Watkins, 2009;Sato et al, 2009).…”

“…The authors interpreted these results to suggest that activity in articulatory motor cortex is enhanced when perceiving speech movements that the perceiver is already experienced in producing and perceiving themselves. Similarly, Bartoli and colleagues observed that the effect of TMS to speech motor areas was related to the listener-speaker perceived acoustic distance, such that response times were facilitated for smaller acoustic distances (Bartoli et al, 2015). By this account, activity in M1 speech areas should be greatest when listening to familiar, natural speech, relative to less familiar, motor-perturbed distorted speech, which is difficult to understand, suggesting somatotopic differences would also be most distinct during perception of natural, unperturbed, speech.…”

Brain Networks Engaged in Audiovisual Integration During Speech Perception Revealed by Persistent Homology-Based Network Filtration

“…The motor cortex, especially the left motor cortex, plays an active role in speech perception, as it is involved in the discrimination of speech sounds (Bartoli et al, 2013;Osnes et al, 2011). This left motor system seems to contribute to speech perception under adverse listening conditions, such as under conditions of degraded or incomplete speech but with recognizable phonetic signals (Osnes et al, 2011).…”

The neural oscillatory markers of phonetic convergence during verbal interaction

“…In this sense, it has been proposed that the rolandic alpha ERD reflects the unspecific activation of sensorimotor areas, while the beta ERD signals a relatively more focal motor recruitment (Pfurtscheller & Da Silva, 1999;Pfurtscheller, Pregenzer, & Neuper, 1994). In fact, specific sensorimotor regions recruited during speech production are also activated during speech listening (Fadiga, Craighero, Buccino, & Rizzolatti, 2002;Watkins et al, 2003;D'Ausilio et al, 2014) and the perturbation of these sensorimotor centers affects speech discrimination performance (Meister, Wilson, Deblieck, Wu, & Iacoboni, 2007;D'Ausilio et al, 2009;D'Ausilio, Bufalari, Salmas, & Fadiga, 2012;Bartoli et al, 2015;Möttönen & Watkins, 2009). In fact, specific sensorimotor regions recruited during speech production are also activated during speech listening (Fadiga, Craighero, Buccino, & Rizzolatti, 2002;Watkins et al, 2003;D'Ausilio et al, 2014) and the perturbation of these sensorimotor centers affects speech discrimination performance (Meister, Wilson, Deblieck, Wu, & Iacoboni, 2007;D'Ausilio et al, 2009;D'Ausilio, Bufalari, Salmas, & Fadiga, 2012;Bartoli et al, 2015;Möttönen & Watkins, 2009).…”