# Conflicting Motor Plans and Sensory Attenuation: Evidence From Event‐Related Potentials for Sounds Generated by Pro‐ and Antisaccades

## TL;DR

This study shows that conflicting motor plans during eye movements affect how the brain processes self-generated sounds, suggesting motor signals influence perception.

## Contribution

The study provides evidence that conflicting efference copies during antisaccades affect sensory attenuation differently than in prosaccades.

## Key findings

- N1 amplitude for prosaccade-generated sounds decreased over time compared to antisaccade-generated sounds.
- P2 amplitudes for prosaccade-generated sounds decreased earlier than for antisaccade-generated sounds.
- Conflicting motor signals during antisaccades disrupt sensory attenuation mechanisms.

## Abstract

The reduction of neural responses to self‐ versus externally generated stimuli has been ascribed to predictions based on an efference copy of motor commands. However, general predictive mechanisms not specific to movements may also play a role. For antisaccades, that is, eye movements in the opposite direction of a target stimulus, an automated prosaccade has to be suppressed, which may lead to conflicting efference copy signals, as an efference copy is likely created also for the prosaccade. If efference copies for the suppressed and executed saccade are in conflict with each other, prediction mechanisms based on their information are potentially disturbed, which may affect the processing of saccade‐generated stimuli. We compared the N1 and P2 components for pro‐ and antisaccade‐generated sounds with those for visually cued external sounds and found differing temporal dynamics of both components during the course of the experiment, depending on the saccade type. The N1 amplitude for pro‐ but not antisaccade‐generated sounds changed over the course of the experiment, with evidence of an attenuation relative to visually cued sounds at the end. The P2 for prosaccade‐generated sounds decreased already earlier than that for antisaccade‐generated sounds, which only decreased toward the end of the experiment. These findings suggest that both early (N1) and late (P2) processing of saccade‐generated sounds is affected by conflicting efference copies, with the early effect probably reflecting forward model predictions and the later effect indicating agency perception based on these predictions.

In this study, we provide evidence for the involvement of motor information in the attenuation of self‐generated compared to visually cued sounds, beyond the involvement of general temporal predictability. We show that only event‐related potential amplitudes for sounds generated by prosaccades, but not antisaccades, decreased significantly over time, likely stemming from conflicting motor signals during antisaccade execution, which demands the suppression of a reflexive prosaccade.