Humanlikeness as design, anthropomorphism as inference: a conceptual framework for human-robot interaction

Area of Science:

• Human-Robot Interaction (HRI)
• Robotics Design
• Human-Computer Interaction (HCI)
• Cognitive Science

Background:

• Humanlike robots are increasingly common in various settings.
• Research and design are challenged by conflating robot humanlikeness with human anthropomorphism.
• Current definitions often overemphasize appearance, limiting precise understanding.

Purpose of the Study:

• To analytically distinguish robot humanlikeness from human anthropomorphism.
• To propose an integrated framework for understanding these and related constructs.
• To provide resources for principled study design and evaluation in HRI.

Main Methods:

• Conceptual analysis and theoretical distinction of humanlikeness and anthropomorphism.
• Development of an integrated framework encompassing related concepts (humanness, machinelikeness, etc.).
• Curating measurement resources for empirical study.

Main Results:

• Humanlikeness is defined as a multidimensional design property of an agent mimicking human characteristics.
• Anthropomorphism is defined as a psychological attribution process by humans to non-human agents.
• Treating these as distinct yet interacting enables clearer theorizing and design guidance.

Conclusions:

• Analytically separating humanlikeness and anthropomorphism is crucial for advancing HRI research.
• The proposed framework and resources support cumulative, interdisciplinary work.
• This distinction facilitates more precise operationalization and actionable design strategies for humanlike robots.