Abstract
Background
Sensory brain areas typically reduce their activity when we speak, allowing us to differentiate our own from someone else’s speech. Similarly, the amplitude of the N100 component of the EEG event-related potential in response to own speech is smaller than for passive listening to own or someone else’s speech. This amplitude suppression effect seems to be altered in voice hearers, which in turn could result in source misattribution (e.g., self-produced voice attributed to an external source). Emotion in speech can have a comparable effect, altering not only self-voice processing but also differentiation of the quality of auditory hallucinations in clinical and non-clinical voice hearers. For example, unlike in non-clinical voice hearers, auditory hallucinations in clinical voice hearers are usually derogatory in content and negatively affect daily functioning. Recent research strongly suggests that clinical and non-clinical voice hearers lie on a continuum ranging from low to high hallucinatory proneness. Based on this notion, the present study used EEG to investigate the effects of manipulations of self-voice quality in self-generated and passively listened-to self-voice as a function of hallucinatory proneness (HP) in healthy young adults. This is the first EEG study that examined the interplay of sensory suppression, emotion, and HP in a non-clinical population.
Methods
Participants varying in HP (according to the Launay Slade Hallucination Scale) participated in a standardized button-press task to elicit their own voice (compared to passively listening to it) in which the self-voice changed stepwise from fully neutral to fully emotional. The experimental task comprised three conditions: motor-to-auditory (MA), where the button-press generated the voice, auditory only (AO), where the voice was presented without the button press, and motor only (MO-a control condition to remove the motor related artifacts from the MA condition), where the button press did not generate the voice. Neutral and angry self-voice (single syllable ‘ah’ and ‘oh’ vocalizations of 500 ms duration) were recorded for each participant before the EEG acquisition. These voices were morphed to generate a neutral to angry continuum consisting of five stimuli ranging from fully neutral to fully angry: 100% neutral, 60-40% neutral-angry, 50-50% neutral-angry and 40–60% neutral-angry and 100% angry.
Results
Preliminary results with 17 participants show a significant effect of emotional self-voice quality on N1 suppression effect, with a larger suppression effect for the 100% angry as compared to 100% neutral self-voice. On the other hand, 60-40% neutral-angry, 50-50% neutral-angry and 40–60% neutral-angry self-voice show an enhancement effect. Furthermore, the results show a significant interaction of HP and voice quality on N1 suppression effect such that high HP showed no N1 suppression effect for the 100% neutral self-voice and an enhanced N1 effect when emotional quality of the self-voice increased.
Discussion
These data suggest that participants perceive the manipulations in the self-voice quality such that they recognize their own fully neutral and angry voice depicted by N100 suppression effect. Similarly, an N100 enhancement effect for 50-50% neutral-angry voice suggest that it is perceived as the most uncertain or peculiar of all the stimuli. Further, low and high HP show difference in N100 suppression effect for different voices, suggesting that HP may alter self-voice processing and these alterations are enhanced for emotional self-voice. This further supports the fact that abnormal perceptual experiences in voice hearers are higher when auditory hallucinations are emotional in nature.
Voice symptoms and handicap in adults with cochlear implants
