Presenter Name:Zehra Mussaji

School/Affiliation:McMaster University

Co-Authors:David Prete & Laurel J. Trainor

Abstract:

To understand language and music, the brain must make sense of incoming sounds. This can be explained by either predictive coding or neural adaptation. According to predictive coding, the brain forms hypotheses from incoming sound patterns to predict the next sound. If a sound violates prediction, it elicits a prediction error. According to neural adaptation, sounds activate feature neurons that habituate after repetition. When a deviant sound is presented, new feature neurons are activated. To disentangle these mechanisms, we measured EEG brain activity to silence, specifically looking at the mismatch negativity (MMN) and the P3a event related potentials, which are elicited by pitch, melody or rhythmic violations. We presented participants (n=27) with two silence conditions: an unexpected silence and an expected silence. The unexpected silence condition consisted of isochronous piano tones (onset-to-onsets 500 ms) with 10% pseudo-randomly omitted, creating unexpected silences. The expected silence condition consisted of isochronous piano tones with onset-to-onsets of 1000 ms, creating expected silences 500 ms after tone onsets. Comparisons between the unexpected and expected silences revealed significant MMN and P3a in response to unexpected silences compared to expected silences. Overall, our results provide strong evidence for the predictive coding framework, and suggest similar brain responses to unexpected silences as to unexpected changes in sound features.