Modeling habits as self-sustaining patterns of sensorimotor behavior.

Area of Science:

• Cognitive Neuroscience
• Computational Psychology
• Robotics

Background:

• Current psychological and cognitive neuroscience models often reduce habits to stimulus-triggered response probabilities.
• This perspective overlooks the complex, self-organizing nature of habitual behaviors within an ecological context.

Purpose of the Study:

• To propose an alternative theoretical view of habits as self-maintaining behavioral patterns.
• To introduce a computational model that captures the organismic and self-organizing aspects of habit formation.
• To explore the role of sensorimotor loops, embodiment, and situatedness in habit organization.

Main Methods:

• Development of a novel computational model: the iterant deformable sensorimotor medium (IDSM).
• Coupling the IDSM to the sensors and motors of a simulated robot.
• Testing the model in various simulated environments to observe habit formation.

Main Results:

• The IDSM demonstrated the formation of self-maintaining patterns of activity across the model, robot body, and environment under specific conditions.
• The model successfully generated distinct habits in response to different environmental setups.
• The model provides a sensorimotor "meso-scale" abstraction of habits.

Conclusions:

• Habits can be conceptualized as self-organizing, biological-like processes rather than mere automatisms.
• The proposed model offers a framework for understanding behavioral self-organization, historicity, and autonomy.
• This approach moves beyond traditional atomistic and statistical conceptions of habit.