On the limit: Working memory capacity constrains learning of complex visuomotor mappings

Area of Science:

• Cognitive Neuroscience
• Motor Control
• Human Learning

Background:

• Human sensorimotor adaptation involves mapping sensory input to motor output.
• Explicit cognitive strategies, utilizing working memory, play a significant role alongside implicit learning.
• The scalability of these explicit strategies for complex sensorimotor tasks is not well understood.

Purpose of the Study:

• To investigate the capacity of explicit cognitive strategies to support learning of complex sensorimotor mappings.
• To determine the limits of working memory in sensorimotor adaptation.
• To explore how task complexity affects the effectiveness of explicit learning strategies.

Main Methods:

• Parametrically tested explicit strategy capacity in a visuomotor rotation task.
• Varied the number of target-rotation pairings participants had to learn (4 vs. 8 mappings).
• Utilized a parameter-free working memory model to predict performance based on stored associations and random guessing.

Main Results:

• With 4 mappings (within working memory capacity), participants achieved near-perfect compensation via stored associations.
• With 8 mappings (exceeding capacity), participants failed to fully compensate, reaching ~50% asymptotic performance.
• Performance limits were accurately predicted by a working memory model, indicating a capacity bottleneck.

Conclusions:

• Explicit cognitive strategies for sensorimotor adaptation are constrained by working memory capacity.
• A cognitive bandwidth limit exists, beyond which adaptation effectiveness plateaus.
• Task complexity exceeding working memory capacity fundamentally limits the utility of higher-order cognition in sensorimotor learning.