Decoding the Neural Stages from Action and Object Recognition to Mentalizing

Area of Science:

• Neuroscience
• Cognitive Science
• Psychology

Background:

• Inferring intentions from actions requires integrating conceptual action and object information with semantic knowledge.
• The neural mechanisms underlying the interaction between action/object recognition and mentalizing remain unclear.

Purpose of the Study:

• To elucidate the processing stages from basic action/object recognition to mentalizing.
• To investigate how neural systems for action, object, and mental state representation interact.

Main Methods:

• fMRI-based crossmodal multiple regression representational similarity analysis.
• Examined human female and male participants' brain activity during action interpretation tasks.

Main Results:

• A modality-general network of frontoparietal and temporal regions supports intention inference.
• Distinct but overlapping networks exist for action, object, and mental state representation, challenging a strict hierarchical model.
• A left-lateralized core network (ventrolateral prefrontal, inferior parietal, anterior lateral occipitotemporal cortex) shows representational overlap for action, object, and mental state information.

Conclusions:

• The findings argue against a sequential bottom-up hierarchy from recognition to mentalizing.
• A core left-hemispheric semantic hub integrates conceptual action/object information with mental state representation.
• This core network is crucial for semantic interpretation and bridges recognition and mentalizing systems.