Subtle predictive movements reveal actions regardless of social context

Area of Science:

• Cognitive Science
• Neuroscience
• Social Psychology

Background:

• Human action prediction is crucial for social interaction.
• Social context influences how we perceive and predict others' intentions.

Purpose of the Study:

• To identify information enabling action prediction.
• To investigate how social context (cooperation vs. competition) modulates this prediction.

Main Methods:

• Interactive reaching game with human participants.
• Psychophysics experiment with observers predicting movement direction from video clips.
• Machine learning classifier used for comparative analysis.
• Information mapping technique to pinpoint relevant visual cues.

Main Results:

• Both humans and a machine learning classifier predicted movement direction before finger lift-off.
• Predictive information was distributed across the initiator's body in both cooperation and competition.
• Intentions were revealed even when social motivation suggested concealment.

Conclusions:

• Action intentions are discernible from body movements early in the action sequence.
• This predictive information is available regardless of whether the social context is cooperative or competitive.
• Humans readily reveal intentions, even in competitive scenarios.