A predictive processing framework for joint action and communication

Area of Science:

• Cognitive Neuroscience
• Social Psychology
• Computational Neuroscience

Background:

• Human joint action and communication are complex abilities requiring individuals to transcend personal perspectives for shared goals.
• Existing frameworks often focus on agent-specific models, which may not fully capture the collaborative nature of human interaction.
• Understanding the computational underpinnings of social cognition is crucial for explaining human cooperation.

Purpose of the Study:

• To propose a novel theoretical framework for explaining human joint action and communication.
• To extend predictive processing architectures from individual cognition to multiagent interactive settings.
• To introduce the concept of agent-neutral models in understanding collaborative behaviors.

Main Methods:

• Development of a theoretical framework based on predictive processing principles.
• Extension of individual predictive processing to multiagent, interactive contexts.
• Formulation of three key claims regarding agent-neutral models in joint action and communication.

Main Results:

• Proposed that co-agents utilize agent-neutral predictive frameworks during joint action.
• Hypothesized that agent-neutral models are updated interactively through communication, shaping predictions.
• Suggested that agent-neutral models facilitate dynamic role allocation in joint actions.

Conclusions:

• The agent-neutral predictive processing framework offers a new perspective on human sociality.
• This framework integrates theories of joint action and communication with computational neuroscience.
• Further experimental research is needed to validate the proposed claims and the framework's efficacy.