Vasconcelos, M. F. 1 & Pinheiro, A. P. 2
1 Psychological Neuroscience Laboratory, Research Center in Psychology (CIPsi), School of Psychology, University of Minho, Braga, Portugal
2 Faculdade de Psicologia, Universidade de Lisboa, Lisboa, Portugal
The current study probed whether and how musical training is associated with enhanced auditory statistical learning (ASL), irrespective of stimulus type. Behavioral (i.e. explicit) and electrophysiological (i.e., implicit) measures were combined to examine online SL and its outcomes. Learning outcomes considered both the online responses to violations in the streams' statistical structure and the offline responses to recognition tasks. Eighteen musicians and 22 non-musicians performed three experiments examining the processing of statistical regularities in prosodic (i.e., words with pitch contour), non-prosodic (i.e., words with flat contour) and musical streams. Compared to non-musicians, musicians showed an enhanced negativity in the 250-300 ms latency window when learning prosodic words and an enhanced positivity in the first 100 ms post-stimulus onset when learning melodies. Only structural violations to prosodic words elicited different ERP responses between groups. Behaviorally, only musicians demonstrated successful learning of the items contained in the auditory streams. Overall, these findings suggest that the neurofunctional mechanisms underlying ASL are modulated both by the musical background of participants and by stimulus type. Additionally, they shed light on the processes underlying positive transfer effects from music to speech, also providing evidence for cross-domain neuroplasticity effects of training.