Adaptive learning of a naturalistic bimanual task in virtual reality
Traditional motor adaptation studies often use constrained tasks that limit natural movement strategies. Using virtual reality, we studied people performing a realistic bimanual plate-lifting task while learning to account for a visual gain distortion applied to the right hand. We measured adaptation of early hand speed and the final plate position in three task conditions: bimanual lifting to a narrow target, unimanual lifting to a narrow target, and bimanual lifting to a wide target. As in previous studies, both hands initially adjusted to the distortion in bimanual conditions. But ultimately only the right hand adapted its speed and showed after-effects, contrasting prior reports. Contrary to our expectation, participants did not adapt early speed more when using one versus two hands. When we widened the target zone, participants achieved greater success in final plate position without reducing adaptation of early speed. Finally, both bimanual groups used a strategy of tilting the plate to be successful and showed no after-effects in final plate position when the distortion was removed. In contrast, the unimanual group did not tilt the plate and did show after-effects in final plate position. These findings reveal that in naturalistic tasks, people leverage multiple movement strategies to achieve goals. Overall, our findings support established principles of adaptation but also challenge expectations derived from more constrained motor learning paradigms, highlighting the importance of studying motor learning in more naturalistic contexts.