Abstract

Sign languages are multimodal, integrating manual articulators such as handshape, movement, and location with non-manual elements including facial expression, gaze, and body posture to construct meaning across parallel channels. Although Kress and van Leeuwen’s Visual Grammar (VG) has been extensively applied to static visual texts, its application to dynamic, time-based signed discourse remains underexplored. This study proposes an exploratory socio-semiotic framework that maps VG’s three metafunctions―representational, interactive, and compositional meanings―onto coordinated manual, non-manual, and spatial-temporal resources in Korean Sign Language (KSL). The analysis draws on selected discourse excerpts from the National Institute of Korean Language KSL Nuri Dictionary, transcribed and time-aligned using ELAN, focusing on six focal lexical items: CAR, OIL, ENGINE, FOR LONG, CHANGE, and DANGER. Qualitative findings reveal recurring tendencies: manual parameters foreground representational meaning, non-manual features foreground interactive meaning, and spatial-temporal organization contributes to compositional meaning, while allowing functional overlap across semiotic channels. By operationalizing VG for embodied and temporally unfolding discourse, this study provides an analytic basis for extending the framework to sign languages and motivates future cross-linguistic research using larger datasets.