Past research has shown that musical training induces changes in the processing of supra-segmental aspects of speech, such as pitch and prosody. The aim of the present study was to determine whether musical expertise also leads to an altered neurophysiological processing of sub-segmental information available in the speech signal, in particular the voice-onset-time. Using high-density EEG-recordings we analyzed the neurophysiological responses to voiced and unvoiced consonant-vowel-syllables and noise-analogs in 26 German speaking adult musicians and non-musicians. From the EEG the N1 amplitude of the event-related potential and two microstates from the topographical EEG analysis (one around the N1 amplitude and one immediately preceding the N1 microstate) were calculated to the different stimuli. Similar to earlier studies the N1 amplitude was different to voiced and unvoiced stimuli in non-musicians with larger amplitudes to voiced stimuli. The more refined microstate analysis revealed that the microstate within the N1 time window was shorter to unvoiced stimuli in non-musicians. For musicians there was no difference for the N1 amplitudes and the corresponding microstates between voiced and unvoiced stimuli. In addition, there was a longer very early microstate preceding the microstate at the N1 time window to non-speech stimuli only in musicians. Taken together, our findings suggest that musicians process unvoiced stimuli (irrespective whether these stimuli are speech or non-speech stimuli) differently than controls. We propose that musicians utilize the same network to analyze unvoiced stimuli as for the analysis of voiced stimuli. As a further explanation it is also possible that musicians devote more neurophysiological resources into the analysis of unvoiced segments.

Musicians and musically untrained individuals have been shown to differ in a variety of functional brain processes such as auditory analysis and sensorimotor interaction. At the same time, internally operating forward models are assumed to enable the organism to discriminate the sensory outcomes of self-initiated actions from other sensory events by deriving predictions from efference copies of motor commands about forthcoming sensory consequences. As a consequence, sensory responses to stimuli that are triggered by a self-initiated motor act are suppressed relative to the same but externally initiated stimuli, a phenomenon referred to as motor-induced suppression (MIS) of sensory cortical feedback. Moreover, MIS in the auditory domain has been shown to be modulated by the predictability of certain properties such as frequency or stimulus onset. The present study compares auditory processing of predictable and unpredictable self-initiated 0-delay speech sounds and piano tones between musicians and musical laymen by means of an event-related potential (ERP) and topographic pattern analysis (TPA) [microstate analysis or evoked potential (EP) mapping] approach. As in previous research on the topic of MIS, the amplitudes of the auditory event-related potential (AEP) N1 component were significantly attenuated for predictable and unpredictable speech sounds in both experimental groups to a comparable extent. On the other hand, AEP N1 amplitudes were enhanced for unpredictable self-initiated piano tones in both experimental groups similarly and MIS did not develop for predictable self-initiated piano tones at all. The more refined EP mapping revealed that the microstate exhibiting a typical auditory N1-like topography was significantly shorter in musicians when speech sounds and piano tones were self-initiated and predictable. In contrast, non-musicians only exhibited shorter auditory N1-like microstate durations in response to self-initiated and predictable piano tones. Taken together, our findings suggest that besides the known effect of MIS, internally operating forward models also facilitate early acoustic analysis of complex tones by means of faster processing time as indicated by shorter auditory N1-like microstate durations in the first ~200 ms after stimulus onset. In addition, musicians seem to profit from this facilitation also during the analysis of speech sounds as indicated by comparable auditory N1-like microstate duration patterns between speech and piano conditions. In contrast, non-musicians did not show such an effect.

Musical expertise has been shown to induce widespread structural and functional alterations in the brain, even-handedly affecting top-down and bottom-up factors. At the same time, it is known that the early evoked gamma-band response (GBR) can be modulated by top-down as well as bottom-up factors such as attention and sound intensity. In this study, we examined the effects of musicianship and attention on the intensity modulation of the auditory-evoked GBR. We compared the electroencephalogram of 17 professional musicians with that of 17 musical laymen obtained during either a forced-choice discrimination task (active) or a passive listening condition. Pure 1000 Hz sine tones were presented at three systematically varied sound intensities (40, 60, and 80 dB sound pressure levels). The results of auditory-evoked potentials and evoked GBRs obtained in the active condition predominantly corresponded to the findings of previous studies. Besides the already known augmentation of the early evoked GBR because of enhanced intertrial phase coherence with increasing sound intensity, we also observed stronger GBRs and enhanced phase locking under the active condition compared with passive listening, whereas the general shape of intensity modulation was comparable between the two conditions. In addition, phase locking to stimulus onset was increased for stimuli of all three intensities when attended, whereas in musicians, only stimuli of the highest intensity (80 dB) induced significantly increased phase locking under the active condition. Taken together, our results suggest that musical expertise influences attention effects on the intensity-modulated early auditory-evoked GBR with respect to intertrial phase coherence.

No abstract