Sensory Attenuation related to Action-effect Prediction and the Sense of Agency

by Elisabeth Lindner
Doctoral thesis
Date of Examination:2023-09-07
Date of issue:2024-09-03
Advisor:Prof. Dr. Alexander Gail
Referee:Prof. Dr. Alexander Gail
Referee:Andrea Desantis
Persistent Address: http://dx.doi.org/10.53846/goediss-10709

 

 

Files in this item

Name:Dissertation_publ_ohCV.pdf
Size:4.62Mb
Format:PDF

The following license files are associated with this item:

Abstract

English

Sensory consequences caused by a self-generated action (action-effect) are assumed to lead to less intense perception compared to external stimuli. This reduced perception referred to as sensory attenuation is suggested to be driven by internal predictions of action-effects and has been linked to the sensation of being in control of one’s own actions and its sensory outcomes (sense of agency). In this thesis I question whether sensory attenuation as represented perceptually and neuronally (in electroencephalography, EEG) is indeed related to the sense of agency. In Study 1 we approached this question by investigating whether and how relevance of the action-effect for a subsequent action affects sensory attenuation as measured in EEG. By analyzing event related potentials in response to actively generated prediction-congruent and prediction-incongruent sounds we demonstrate that sensory attenuation in the auditory N1 component is not a necessary correlate of auditory action-effect prediction. Study 2 sheds light on perceptual auditory sensory attenuation related to temporal action-effect prediction. Further, we assessed self-reported sense of agency. We found no evidence of sensory attenuation for temporal action-effect prediction, while the judgments of agency decreased with increasing delay, indicating a dissociation between the two phenomena in the context of auditory temporal action-effect prediction. In study 3 we examined whether and how different levels of active control influence visual perception. While we observed that reaction times on a detection task increased with decreasing levels of control, there was no difference compared to a passive condition, in which the stimulus in question was replayed from the active condition. This suggests that not the action and therefore active control itself had an effect on perception, but rather prediction per se. Summarizing, none of our experiments revealed sensory attenuation for sensory effects of self-generated movement, as assumed in standard motor theory. This challenges the validity of sensory attenuation as an implicit measure of agency and suggests a dissociation between sensory attenuation and the sense of agency. However, predictive processes related to the effects of controlled action themselves may still hold an important role in the formation of agency, potentially in conjunction with more general gating mechanisms.
Keywords: Sensory attenuation; Action-effect prediction; Sense of agency; EEG; Pupil; Cancellation; Psychophysics; N1; P2
 
Statistik