What exactly do robophilosophers refer to when they criticize certain representations of gender? How can they critique representations without reproducing stereotypes themselves? What is referred to in the descriptions, if what is to be described is socially constructed? In order to address these questions, this poster aims to answer the overall question: How can gender as a social construct be described in social robotics without reproducing problematic stereotypes? We conducted a multi-layered analysis of the article “Gendering Humanoid Robots: Robo-Sexism in Japan”, which confirms that the social practice of academic writing in the field of social robotics bears a risk of reproducing gender stereotypes. Our overall research question on how gender as a social construct can be described in social robotics without reproducing problematic stereotypes can be answered as follows: gender in social robotics can be described by approaching robots as a sociotechnical system, which involves their interdependency with society and social structures, and should not be reduced to possible physical attributes of social robots.