A margin for error in grasping: hand pre-shaping takes into account task-dependent changes in the probability of...

Area of Science:

• Neuroscience
• Motor Control
• Human-Computer Interaction

Background:

• Grasping movements require balancing success probability with movement costs under uncertainty.
• It is assumed that error probability is managed by adjusting a margin for error in hand opening.
• The brain must estimate uncertainty and compute how movement alterations affect error probability (the 'probability landscape').

Purpose of the Study:

• To determine if grasping movements are sensitive to the probability landscape.
• To investigate if the brain actively controls error probability during grasping tasks.
• To test the margin-for-error account of grasping under varying uncertainty.

Main Methods:

• Subjects performed three grasping tasks with distinct probability landscapes.
• Visual uncertainty was manipulated by blurring vision and monocular occlusion.
• Movements were executed without visual feedback to isolate initial object property estimation uncertainty.

Main Results:

• Changes in hand opening under increased visual uncertainty aligned with predictions from the margin-for-error account.
• Different grasping tasks exhibited qualitatively different responses to uncertainty.
• Grasping movements demonstrated sensitivity to the probability landscape.

Conclusions:

• Grasping movements are sensitive to the probability landscape, supporting the margin-for-error hypothesis.
• The brain actively controls error probability by adjusting movement parameters.
• Findings support the intuitive notion of a true margin-for-error mechanism in grasping.