Looking for a sign forecasting failure in actions: reaching errors triggered by a slowdown of movement and specific brain activity in preceding trials

Summary Even experts can sometimes fail while performing fully learned movements. Do such failures suddenly arise, or are there any forecasting signs? It has been reported that the kinematics of the early phase of movements can predict the failure, and brain activity patterns specific to failures are observed just before the movement onset. The presence of abnormal brain activity patterns long before (> 30 s) a failure in a cognitive task leads us to question if signs of a failure in action could exist in trials preceding the failure. Here, we examined this question using a reaching movement adaptation paradigm conventionally used to test motor learning dynamics. Firstly, the presence of a behavioral sign that preceded failures was observed: the peak velocity of the reaching movement significantly decreased in the preceding two trials. Secondly, specific theta and alpha band activity of EEG were observed in the failure trials and the trials preceding the failure. These results suggest that a failure in actions does not occur suddenly, and some signs preceding failures can be observed in the prior trials. Our approach may pave the way to investigate how we prevent failures and improve motor performance.