Processing of self-initiated sounds

En savoir plus sur le livre

The dissertation investigates the processing of self-initiated versus externally-initiated sounds, proposing an internal prediction mechanism for sound discrimination. This mechanism likely involves an efference copy of motor commands and sensory feedback loops. Previous research indicated attenuated N1 or N1m components for self-initiated sounds compared to external ones, but the underlying mechanisms remain less understood. This thesis identifies three essential features of the internal prediction mechanism: specificity, adaptivity, and predictability, explored through five EEG studies. Participants pressed a button to trigger sounds, with conditions including replayed sounds. Experiment 1 used unpredictable sounds to test the internal prediction mechanism under uncertainty, finding N1 responses still attenuated for self-initiated sounds. Experiment 2 confirmed that these attenuations were linked to self-initiation, as only self-initiated sounds showed reductions in an auditory stream containing both types. Experiment 3 examined the impact of foreknowledge on stimulus timing, revealing N1 reductions for self-initiated sounds even when participants were cued about impending sounds. Experiment 4 focused on middle latency responses, showing early attenuation for self-initiated sounds starting at 35 ms post-onset. Experiment 5 introduced unexpected sounds after button presses, yet N1 reductions persisted for self-initiated so