Abstract

Raúl Ruiz’s Mammame (1986), a cinematographic adaptation of the ballet play of the same name created by Jean-Claude Gallotta, is one of his most understudied films. In this article we propose that this film is a privileged site to explore the complex relationship that Ruiz established between sound and image during his whole career. At first, Mammame appears to be a film that documents something like the rehearsals of the Emile Dubois dance company. A close study of the film, however, makes apparent that Ruiz uses the movements of the bodies on screen, as well as the soundtrack (with music composed by Henry Torgue and Serge Houppin), in order to further explore the creative and experimental displacement of sound and image, language and bodies. This displacement points to a fundamental and more general problematization of cinematic dualism. Mammame, we argue, is a film that manages to deconstruct the experience of cinematographic synchronicity by breaking up its components: it makes soundtrack (music, sounds, silences, and noises) and image (bodies, shadows, and locations) work separately throughout the film. The dissolution of voice into exhalations, the blurring of bodily shapes into shadows, and the disjunction of music and moving image – via different strategies to establish cinematic continuity – are all elements that contribute to the experimental display of a general disarticulation of sound and image. More specifically, Mammame establishes a double relation between voice and mouth: on the one hand, voice appears as an overflowing noise (guttural, harsh, rough noises coming out of the mouth); on the other, mouth appears as a body part able to dance (emphasizing dislocated facial gestures). By subverting the notion of cinema as a systematic display of an accordance between what is seen and what is heard, this film puts into question the dualism between body and discourse. Nevertheless, image and sound constantly refer to each other, thus contaminating one another and invading their respective and (quasi) autonomous regimes of representation.