Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Prosodic Pitch Perception in Right Premotor Cortex: A Transcranial Magnetic Stimulation Study
Poster Session B, Friday, October 25, 10:00 - 11:30 am, Great Hall 3 and 4
Daniela Sammler1,2, Tilo Zotschew2,3, Gesa Hartwigsen2,4; 1Max Planck Institute for Empirical Aesthetics, Frankfurt am Main, Germany, 2Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig, Germany, 3University of Bremen, Bremen, Germany, 4Leipzig University, Leipzig, Germany
Speech perception is believed to recruit (pre)motor cortex (PMC) in the left dorsal auditory stream. Articulator-specific involvement of left PMC has been found for the perception of phonemes and lexical tones. However, PMC’s role for prosody—the melody of speech—still lacks investigation. Motor control of prosodic pitch contours is anchored in the right (and left) PMC. This kindles questions of hemispheric lateralization, reminiscent of auditory asymmetries. Moreover, the role of PMC in supporting either perceptual or response-related processes is still a matter of debate. The present study used repetitive transcranial magnetic stimulation (rTMS) and behavioral drift diffusion modelling to (i) assess the causal role of right PMC in prosodic pitch categorization, compared to left PMC and a non-prosodic pitch control task (speaker gender categorization), and to (ii) disentangle the weighting of perceptual vs. response-related biases in PMC contributions. In two separate sessions, 24 listeners (M ± SD = 27 ± 3.4 years) categorized prosodic pitch contours (question vs. statement) and speaker gender (male vs. female) of monosyllabic words while receiving inhibitory or sham rTMS of left or right PMC (MNI: ±45, 5, 40 mm). A 5-pulse train of 10 Hz rTMS was delivered in each trial, time-locked to word onset. The order of sessions (left/right PMC) and blocks (prosody/gender task, sham/effective rTMS) was counterbalanced across subjects. Single trial response times and accuracies were analyzed with the drift diffusion model (DDM; fast-dm) to decompose latent processes that lead to a decision. Parameters v (drift rate; rate of evidence accumulation), a (threshold separation; indicating response biases), and t0 (non-decision time; decision-independent time for stimulus encoding and button press) were compared between tasks (prosody/gender), stimulation (effective/sham), and hemispheres (left/right) using rmANOVAs. Multiple regressions were used to predict ∆v (effective–sham) in both tasks from individual ratings of (i) perceptual focus on prosodic contour, (ii) on mean pitch, (iii) amount of subvocal rehearsal, and (iv) stimulated hemisphere, controlled for (v) years of musical training and (vi) perceived task difficulty. Drift rate v in the prosody task dropped after effective compared to sham stimulation of right (but not left) PMC, and more strongly in participants who were biased to focus on the prosodic contours. No such effects were found for parameters a and t0 or the categorization of speaker gender. Neither amount of subvocal rehearsal, nor years of musical training or perceived task difficulty influenced ∆v. The combined findings extend the modulatory role of PMC in speech perception from phonemes and tones to speech prosody. In particular, the results highlight the causal role of right PMC in pitch contour categorization, in line with auditory cortical asymmetries. The absence of rTMS effects on parameter a denoting response biases, and the stronger modulation of performance in case of a stronger perceptual bias towards pitch contour suggests a role of PMC in perceptual rather than solely response-related processes. Future studies should investigate the time-course of PMC involvement during prosody processing using chronometric protocols to substantiate this conclusion.
Topic Areas: Prosody, Speech Perception