Fig. 1.

Experiment 1–3 task design. a Experiment 1: subjects performed a standard visuomotor rotation task, attempting to land a rotated visual cursor on a target. A between-subjects, 2 × 2 design was used, crossing the factors Rotation Magnitude (the size of the rotation in the rotation block; left) and Set Size (the number of possible target locations in the task; right). b Experiments 2–3: FREE task: subjects performed trial pairs consisting of learning trials (left) and execution trials (right). Thirteen rotation sizes were pseudorandomly presented (−90˚:15˚:90˚). FORCED task: subjects were required to initiate their movement after target presentation and <100 ms after the fourth tone. Targets could appear in either 1 of 12 locations (Exp. 2) or 1 of 2 locations (Exp. 3). The time of target appearance was titrated to induce subjects to react with a distribution of RTs