This book presents a comprehensive methodology for segmenting, transcribing, and analysing feature films, especially classical Hollywood films. It offers an in-depth study of three scenes from Roman Holiday (1953), each showcasing different combinations of verbal and non-verbal performance. Through detailed study, the book demonstrates how these scenes can be broken down into Events, Facts, Tasks, and Actions, while also exploring the semiotic modes involved in acting, art direction, cinematography, and editing. It reveals how acting constructs characters, relationships, situations, and the story's generic structure, and examines how art direction, cinematography, and editing contribute to both performance and genre development. Building on these investigations, the book hypothesises performance features. These hypothesised features are subsequently tested on the Chinese film Raise the Red Lantern (1991), showcasing their flexibility and potential for broader application in film performance. The book makes a significant contribution to existing research by bridging film performance theories with social semiotics, synthesising them into a cohesive framework that is applicable to both Hollywood and non-Hollywood films. This book is especially valuable for scholars, practitioners, and students of film studies and social semiotics.