Eigen RadarScience
Analysis

Motor learning splits implicit aiming from movement recalibration

A single peer-reviewed study found that two implicit motor-learning processes behave differently when people relearn a reaching task or pause their training. One influences the direction selected to reach a goal; the other recalibrates execution against sensory error. Human experiments and computational comparisons challenge the assumption of one implicit system, while leaving the processes' interaction during natural movements unresolved.

Science··Morning
A participant in a mustard shirt uses a stylus on a digitizing tablet with his right hand; a single white target point appears on the monitor ahead.

Relearning separates two implicit responses

When people encountered a reaching task again, implicit aiming learned faster while implicit recalibration learned more slowly. A single peer-reviewed study in PLOS Biology compared these processes in controlled human experiments. Implicit aiming helps select a movement direction towards a goal; recalibration adjusts execution in response to the difference between predicted and observed sensory outcomes.[1]

Delayed feedback exposes persistent changes in direction

Laboratory experiments involved 168 undergraduates at the University of California, Berkeley, and online experiments involved 124 participants recruited through Prolific. All were right-handed. The team tracked laboratory movements with a digitizing tablet and online movements with laptop trackpads. Rotating an on-screen cursor let them examine adjustments in reaching direction.[1]

In the first experiment, visual feedback arrived 1.5 seconds late. The rotation rose gradually to 70 degrees and then fell to 25 degrees. Once feedback was removed and participants were told to move directly towards the target, deviations of roughly 10–14 degrees persisted. Instructions directing attention towards an external perturbation or an internal movement bias changed the size of this aftereffect.[1]

Short breaks reveal different learning dynamics

After one-minute breaks, the drop in implicit aiming became smaller as training progressed. The recalibration drop stayed approximately constant. Computational comparisons used contextual inference and cerebellar population coding models to examine these patterns. The authors propose an action-selection and execution framework for motor learning; how these processes interact in natural tasks remains unresolved.[1]

References

  1. News sourcePLOS BiologyMotor learning experiments distinguish two implicit processes↩1↩2↩3↩4