Abstract
This study examines the possibilities and challenges of a sign language–script–based model for producing health information videos for deaf audiences. The case analyzed is the production process of a Nihon Shuwa (Japanese Sign Language: JSL) diabetes education video developed by the National Cancer Center Japan. Drawing on participatory ethnography, the study analyzes meeting recordings, draft scripts, medical review documents, and field notes collected between December 2021 and March 2025.
The findings indicate that positioning sign language as the primary language of meaning design enables the systematic use of modality-specific principles, including early presentation of the overall conceptual framework, spatial visualization of causal relationships, and the integration of non-manual markers into multilayered discourse structures. In contrast to translation-centered models that begin with written Japanese scripts, the sign language–script–first approach restructures information design according to visual-spatial processing logic.At the same time, the study reveals structural tensions in the medical review and reverse-translation process. Institutional validation relied on written Japanese documents, making reverse translation necessary and exposing asymmetrical linguistic power relations embedded in existing knowledge-production systems. Furthermore, filming and editing were not merely technical post-production stages but integral components of linguistic realization, requiring simultaneous language and visual design.
These findings suggest that the sign language–script–first model is not simply an alternative production technique but an epistemological reconfiguration that repositions linguistic authority in health information design. However, its sustainable implementation requires institutional support, including direct sign language–based review systems and the development of deaf-led professional infrastructures.
