Learning Motor Sequences with and without Knowledge of Governing Rules

Abstract

Objective. To investigate the behavioral and neural effects of rule-based knowledge on motor sequence learning. Methods. The authors developed a novel 2-dimensional variant of the serial reaction time (SRT) task to test the effect of prior, verbalizable rule knowledge on motor learning behavior. To examine neurophysiological effects, they also performed functional magnetic resonance imaging on a small cohort of subjects while performing the same task. Results. Behavioral data demonstrated that instruction on sequence-governing rules enhanced behavioral performance in both learning magnitudes and rates. The neuroimaging data revealed substantially different, but partially overlapping, learning-related activation patterns with and without prior rule instruction. Direct comparison of these 2 conditions revealed significantly different involvement of bilateral superior and anterior prefrontal cortex (Brodmann areas 8 and 10, respectively), right superior temporal cortex (BA 38/21), and left cerebellum. Conclusions. These behavioral findings demonstrate an advantage of teaching governing rules prior to 2D-SRT task performance. While these neuroimaging findings remain to be replicated in a larger cohort of subjects, results suggest that substantially different—though partially overlapping—brain regions subserve learning in these 2 rehabilitation-relevant conditions. Thus, appropriate choice of pretraining may benefit, for example, rehabilitation populations, at least in motor skill acquisition that requires sequencing.