Tout, absolument tout, peut danser au cinéma. Les images comme les personnes, les objets comme la caméra. La danse est souvent là où on ne la cherche pas. Plus on montre les danseurs et moins le film est dansant. ^[1]

1 Dancing Light, Dancing Forms, and Dancing Objects

Dance has been recorded on film since shortly after the invention of cinema. The first commercial film program screened in the United States (1896) included films depicting dance performances.^[2] Narrative films featuring revuelike dance routines were soon produced in Hollywood. These films are not the type that Man Ray primarily had in mind, however, when he wrote, La danse est un sujet idéal pour le cinéma,^[3] for he was more interested in another development. Dance, once it had been liberated from classical ballet d’action by Loïe Fuller and Isadora Duncan around 1900, became a model for the other arts. Fuller had virtually turned dance into an abstract art form by obscuring her body behind costumes and light effects, and it was that same abstract character that simultaneously allowed music to become a model for other arts such as painting and film. Dance and music were called on as models for the development of cinematic concepts, with the connecting links between dance and music (and ultimately film) being the concepts of ‘movement’ und ‘rhythm’.^[4]

The three time-based art forms of film, music, and dance enter into a kind of circular reasoning in which each is often drawn on to delineate one of the others. Walter Ruttmann’s Lichtspiel Opus I (GER 1921), which sets abstract film images in analogy to music, was described by a contemporary as absolute dance,^[5] for example, whereas dancer Valeska Gert saw Ruttmann’s Opus II (GER 1923) as the paradigm for her definition of dance. Oskar Fischinger’s Studien (GER 1929–1934) transpose popular dance music into a cinematic process usually termed form dance or choreography, given that the forms move across the screen (in rhythm with the music) like dancers in formation.^[6]

Whereas German form dance often was developed through drawing or painting on the animation stand, in French experimental films objects and lights are more likely to enter into dancelike motion. Man Ray repeatedly made lights, forms, and objects perform turns — dancing par excellence^[7] — in his work in order to give form to his cinematographic concept of non-narrative film, the cinépoème.^[8] One of many examples is the collar dance in Emak Bakia (FR 1926). In Fernand Léger and Dudley Murphy’s film Ballet mécanique (FR 1924), the link with dance is already apparent in the title, although this work mainly features filmed objects that are set in rhythmical motion through post-production editing and montage. Whereas the link to music is more implicit than explicit in these cases — established through the orchestration of rhythmic and plastic elements^[9] — in the cinéma pur genre the relationship to music has been theoretically elaborated. In an endeavor to define film as art as early as 1922, art critic Élie Faure wrote the following about his concept of cinéplastique: … la cinéplastique tend et tendra chaque jour davantage de se rapprocher de la musique. De la danse aussi.^[10] Germaine Dulac then repeatedly defines film in her writings on the concept of cinéma pur as musique de l’oeil,^[11] which also possesses qualities inherent to dance. Even though in Thèmes et Variations (FR 1928) she uses parallel montage to unite the movements of light, objects, and plants with those of a dancer, for Dulac cinematic dance is in no way associated with the actual presence of a person dancing, but above all with rhythmical motion: J’évoque une danseuse! Une femme? Non. Une ligne bondissante aux rythmes harmonieux.^[12]

Film experiments with combinations of music and dancing light, objects, and forms were carried out only occasionally after the 1920s, for example in films such as Tarantella (US 1940/1941) by Mary Ellen Bute or Free Radicals (US 1958/1979) by Len Lye.