Evolution of behavioral attractors with learning: nonequilibrium phase transitions

Area of Science:

• Motor control
• Dynamical systems theory
• Human motor learning

Background:

• Bimanual coordination is crucial for many daily activities.
• Understanding how the brain learns new coordination patterns is key to motor rehabilitation and skill acquisition.
• Coordination dynamics can be modeled as a dynamical system with attractors.

Purpose of the Study:

• To investigate the dynamical process of learning a bimanual coordination task.
• To examine changes in coordination dynamics over 5 days of practice.
• To understand how intrinsic dynamics interact with learned patterns.

Main Methods:

• Participants practiced a bimanual synchronization task with a specific phasing pattern.
• Coordination dynamics were assessed using relative phase as a collective variable.
• Attractor layouts (phase diagrams) were probed before, during, and after practice.

Main Results:

• Learning induced qualitative changes in the phase diagram, reflecting altered coordination dynamics.
• Evidence of phase transitions and loss of stability was observed during learning.
• The interplay between intrinsic dynamics and the learned pattern influenced the learning process (e.g., abrupt vs. gradual change).

Conclusions:

• Learning a new bimanual coordination pattern involves a shift in the underlying dynamical system.
• The relationship between intrinsic and behavioral dynamics dictates the nature and stability of the learned coordination.
• Learned coordination patterns can be robust and long-lasting.