Timing at peak force may be the hidden target controlled in continuation and synchronization tapping

Timing control, such as producing movements at a given rate or synchronizing movements to an external event, has been studied through a finger-tapping task where timing is measured at the initial contact between finger and tapping surface or the point when a key is pressed. However, the point of peak force is after the time registered at the tapping surface and thus is a less obvious but still an important event during finger tapping. Here, we compared the time at initial contact with the time at peak force as participants tapped their finger on a force sensor at a given rate after the metronome was turned off (continuation task) or in synchrony with the metronome (sensorimotor synchronization task). We found that, in the continuation task, timing was comparably accurate between initial contact and peak force. These two timing events also exhibited similar trial-by-trial statistical dependence (i.e., lag-one autocorrelation). However, the central clock variability was lower at the peak force than the initial contact. In the synchronization task, timing control at peak force appeared to be less variable and more accurate than that at initial contact. In addition to lower central clock variability, the mean SE magnitude at peak force (SEP) was around zero while SE at initial contact (SEC) was negative. Although SEC and SEP demonstrated the same trial-by-trial statistical dependence, we found that participants adjusted the time of tapping to correct SEP, but not SEC, toward zero. These results suggest that timing at peak force is a meaningful target of timing control, particularly in synchronization tapping. This result may explain the fact that SE at initial contact is typically negative as widely observed in the preexisting literature.

Timing accuracy in self-timed movements related to neural indicators of movement initiation

Timely movement initiation is crucial in quick reactions or when a series of movements has to be strung together in a timed fashion to create a coordinated sequence. Stochastic neural variability can lead to misinitiation errors as reaction time studies suggest. Higher reaction times occur when preparatory neural activity reaches an initiation threshold later relative to shorter reaction times. Whether this also applies to self-timed movements is harder to scrutinize because they lack an external event that could serve as a reference for timing accuracy estimations. By example of a self-timed goal-oriented throwing task, we used a method that synchronizes the throwing movements by their kinematic profiles to assess relative timing differences in throwing release. We determined neural preparatory processes of the release using the movement-related electrophysiological Bereitschaftspotential (BP). By analyzing differences in shape and timing of the BP in delayed and non-delayed throws, two variables could be extracted that are related to timing differences on the kinematic level. First, temporal deviations in BP curves partly meet the kinematic deviations. Second, delayed releases were preceded by a short flattening of the BP curves prior to release. Thus, temporal and shape deviations in the neural movement initiation are assumed to delay self-timed movements.