Blech, Christine https://orcid.org/0000-0002-3188-6065
Lembcke, Hanna https://orcid.org/0000-0003-3088-7928
Bouquet, Cédric A. https://orcid.org/0000-0003-4228-3291
Liepelt, Roman https://orcid.org/0000-0002-4656-6267
Funding for this research was provided by:
FernUniversität in Hagen
Article History
Received: 8 July 2025
Accepted: 11 September 2025
First Online: 19 November 2025
Declarations
:
: The success of human-robot-interaction depends on how well people can anticipate a robot’s behavior, which also implies adopting the robot’s point of view. This research starts with the assumption that people can adopt the perspective of a (social) human being more easily than that of a (not so social) robot. Finding full-fledged perspective taking for both, human and robot avatars in a well-established experimental task with static images of avatars, we conclude that submentalizing represents a valid explanation for visual perspective taking. We critically discuss the method and its potential confounding factors (e.g., working memory load) and conclude that more realistic settings are desirable for future research on visual perspective taking.
: The authors declare no competing interests.