Auditory N1 Suppression and Omission N1 Do Not Share a Common Underlying Mechanism.

Abstract

Recent theories describe perception as an inferential process based on internal predictive models that are adjusted by prediction violations (prediction error). Two modulations of the auditory N1 event-related brain potential component have been interpreted as reduced or enhanced prediction error for predictable sensory input: The sound-related N1 component is attenuated for self-generated sounds compared to the N1 elicited by externally generated sounds (N1 suppression). An omission-related component in the N1 time-range is elicited when the self-generated sounds are occasionally omitted (omission N1). We wanted to confirm that both N1 suppression and omission N1 are sensitive to the predictability of sound identity, as reported in the literature. We manipulated the predictability of sound identity in a self-generation paradigm in which button presses in one condition always produced the same sound or in another condition produced a sound randomly selected from a large set of sounds. Omission N1 was modulated by manipulating the predictability of sound identity but surprisingly N1 suppression was not. This contradicts previous reports, challenges prediction-related interpretations of the N1 suppression, and supports alternative explanations for N1 suppression like action-related unspecific sensory gating.