Two procedures, one mechanism? Recent findings on the automation of voluntary movements

Acquiring automaticity in voluntary movement production is usually thought of as a gradual shift in motor control from declarative to non-declarative memory structures, requiring extensive practice. Interference with a cognitive demanding secondary task should be high at the beginning of skill acquisition, but low once the automatic stage has been reached. Following this line of reasoning, three experiments were carried out. After a short initial learning phase, a criterion movement was practiced, along with provision of KR, either in a single-task condition (Exp.1, 1600 trials; Exp.2, 3200 trials), or in a dual-task condition (Exp.3, 560 trials). Performance was tested prior to and after practice. Tests included both a single- and a dual-task condition, with no KR provided. In the dual-task condition, an additional choice reaction time task had to be carried out concurrently to the criterion movement. Exp.1 shows clear evidence for dual-task interference in practice and control subjects, but no reduction of dual-task costs in the course of practice. In Exp.2, only at the very end of (single task) practice (i.e. after about 3000 trials), dual-task costs are completely reduced. Dual-task costs were also nearly eliminated in Exp.3, surprisingly after no more than 240 practice trials, when practice was carried out under dual-task conditions. Thus, there might be no gradual shift from the attentional to non-attentional control. Rather, two systems (attentional resp. nonattentional) seem to be working independently from each other.