Switching between task control modes incurs greater reprogramming costs than programming the response de novo.

• Simultaneous hand/foot lift exhibits different EMG timing depending on control mode. • We occasionally signaled a change from the default predictive to a reactive mode. • When signaled 250 ms prior to initiation, control mode switching did not occur. • When signaled 500 ms prior to initiation, switching caused large reaction time costs. • Task switching between control modes incurs greater costs than programming de novo. When performing synchronous hand and foot movements, the way the limbs are synchronized differs depending on the mode of control. When performed in a reaction time (RT) paradigm (reactive control), EMG onsets become synchronized resulting in asynchronous displacement onset. However, when the same movement is performed as an anticipation-timing task (predictive control), displacement onset is synchronized by unconsciously introducing a small delay between EMG onsets. The present experiment investigated the reprogramming costs associated with switching between predictive and reaction control modes. Participants (N = 12, 6F) were asked to simultaneously lift their right heel and right hand in an anticipation-timing task when a rotating clock hand reached a specified target. On some trials, an auditory stimulus was presented either 250 ms or 500 ms before the target and participants were instructed to execute the synchronous movement as quickly as possible after the signal (i.e., switch to reactive mode). Results showed that when the auditory stimulus was delivered 250 ms before the target, participants were unable to switch to a reactive control mode but did switch when the auditory stimulus was presented 500 ms before the target. As expected, the RT on switch trials was substantially longer (∼230 ms) than a simple RT control condition but was also significantly longer (∼130 ms) than a choice RT control condition. These results indicate that switching between control modes for a task involving the same musculature incurs reprogramming costs that are even greater than the time required to program the response de novo. [ABSTRACT FROM AUTHOR]