Anthropomorphism and service humanoid robots: an ambiguous relationship

Purpose – So‐called humanoid robots, among a large class of service robots, are designed to work in close harmony with humans. Their anthropomorphism and its consequences have, however, been little studied. The purpose of this paper is to tackle this question by differentiating the psychological meaning of anthropomorphism from its technical meaning, understood as a human‐like device. The author shows that the former generates salient projections which can be interpreted with respect to Mori's uncanny valley. The role of the task is highlighted with a theoretical attempt to integrate the robot as a social player into a Heider balance‐theory inspired model. This psychological anthropomorphism, however, must be compared with technical anthropomorphism, which leads to underlining present‐day difficulties in designing highly human‐like functional machines with, as a consequence, running the risk of giving them the delusion of a human behaviour that they are not able to realize. Design/methodology/approach – This is a theoretical paper aimed to highlight a double meaning of anthropomorphism for humanoid robots and its consequences. Findings – Task‐based interpretation of the Mori's uncanny valley and link between psychological anthropomorphism and technical anthropomorphism. Originality/value – The originality of the approach consists in applying to the humanoid robot a double approach of anthropomorphism. The first one corresponds to the classical psychological meaning producing peculiar anthropomorphic projections on a non‐human being, while the second corresponds to the technical realization of a human‐like machine dedicated to be integrated into a human environment.