If a dancing figure falls in the forest and nobody sees her...
Australian Film, Television and Radio School
This short provocation explores a question I have frequently found myself asking when watching productions of the screendance genre known as ‘dance on camera’. I shall argue that this is one of three distinct genres of screendance, each of which positions the audience slightly differently. The other two genres I will delineate, dancefilm and video dance, each have clear ways of positioning of the audience, but my question in relation to dance on camera is frequently: who or what I am watching? Or, how is the dancing figure being positioned by the film?
In considering this question, I will make use of the distinction drawn by film theorist Tom Gunning (1990) between early cinema, which he characterizes as a ‘cinema of attractions’, and cinema after 1906, which is narrative cinema, to inquire about contemporary dance on camera. The question that this paper puts to makers of dance on camera is: if the construction of the dancing figure’s position in relation to the spectator falls into neither frame – if it is not an ‘attraction’ or exhibition and it is not narration to be observed (as Gunning puts it) voyeuristically – then how does the film construct the dancing figure in relation to the spectator?
My inquiry will begin with a brief discussion of Gunning’s distinction between exhibitionist and voyeuristic cinema, aided by the analysis Teresa Rizzo has carried out in her essay YouTube, The New Cinema of Attractions (2008). Following this, a brief survey of the ways the dancing figure is positioned in fiction films, documentaries, reality TV, dancefilm and video dance will further refine the various ways in which the dancing figure is positioned in relation to the screen audience. Having established the means to specifically pinpoint the positioning of the dancing figure in each of these kinds of screen production, I pose my question about the screendance genre known as ‘dance on camera’ and ask whether it is a return to the cinema of exhibition, or is intended to be understood as a voyeuristic experience of a closed story world, or is a hybrid of attraction and narration that sometimes succeeds in shifting the spectator back and forth between two modes, and sometimes leaves one mystified as to what – and who – we are watching.
Cinema of Attractions v. Narrative Cinema
Film Theorist Tom Gunning distinguishes between two modes of spectatorship when he describes early cinema as the cinema of attractions and later cinema as a cinema of voyeurism. Scholar Teresa Rizzo quotes and interprets Gunning when she writes:
“attractions address the viewer directly, soliciting attention and curiosity through acts of display. … As moments of spectacle, their purpose lies in the attention they draw to themselves,” rather than in the development of narrative devices such as “characterization, causality, narrative suspense, and the creation of a consistent fictional world” (Gunning 1994: 190)
According to Gunning, unlike contemporary narrative cinema, which solicits a voyeuristic spectatorial gaze, early cinema is an exhibitionist cinema where the spectator is overtly acknowledged and invited to look. As Gunning puts it, “It is the direct address of the audience, in which an attraction is offered to the spectator by a cinema showman, that defines this approach to film making” (Gunning 1990: 58) (Rizzo 2008).
This distinction between cinema where the spectator is “invited to look”, and cinema which tells a story that the spectator sits outside of, unacknowledged by the performers, is useful to the discussion of dance on camera in part because films of live dancing (and acrobatics and other forms of physical display) were frequently featured in early cinema. The dancing figure in these pieces of early cinema was clearly and precisely positioned: they were dancers, performing. The performance was a spectacle we were invited to watch, and the performers might acknowledge us with a wave, or wink or nod or even bows at the end of the performance. One of the clearest signals that the performer was aware of herself presenting a performance to an audience is the performer’s own gaze: she often looks directly at the lens, or, in more common parlance, ‘down the barrel’ at the spectator, smiling at us and acknowledging our presence. This mode of positioning is what Gunning calls ‘direct address’ – the dancer addresses her performance directly to the spectator. See, for example, the film of Loie Fuller’s Danse Serpentine:
What Gunning’s theory usefully provides is a clear articulation of two possible relationships constructed between spectator and performer in film. The objective of this paper is to relate these particularly to screen dance, describing how in the different manifestations of screendance the dancing figure is located as a character (unaware of the spectator’s gaze), as person (telling their story to an onscreen listener), as dancer (such as in Reality TV), or a form (an abstraction or distortion of the human figure). My discussion will then consider which of these modes or what hybrids of these modes are being used in the genre of screendance known as dance on camera and how this particular genre does or does not position the dancing figure as either an exhibition or a part of a narration.
Positioning the Dancing Figure
Fictional feature film narratives with dance as a central story element will be considered first, since these generally offer a clear and consistent positioning of the figures on screen. These films position the dancers as characters in the story. When they dance in the narrative the spectator understands it is the character that is dancing, not the dancer who has been employed to play the character, just as it is the character that is involved in the dramatic story, not the actor who plays the character. The dancers never ‘look down the barrel’ directly at the camera lens or acknowledge the cinema audience. They may, in the context of the story perform for an audience within the story, in which case that audience is shown onscreen. Or they may, as characters, dance without an audience. The characters may dance to express their emotions. (See for example the character Fred Atwell [played by Fred Astaire] dance in a bar on The Sky’s the Limit [Griffith, 1943]. The dancing is an expression of the character’s fury about his predicament, not a performance for an onscreen audience.)
There is an interesting hybrid sequence in Billy Elliot [Daldry, 2000] where Billy (played by Jamie Bell), is forced up onto the kitchen table to dance for his father and brother. From there, a montage sequence cuts to Billy dancing in the street, and other places, expressing his frustration with his world and his family. The sequence starts with Billy performing for an onscreen audience (his family) and then shifts to dancing without an audience, for emotional release. However in the sequence, as in all narrative fiction films that feature dance on screen as a central element, whether dancing for an onscreen audience or dancing for emotional release the position of the performer never switches into direct address. The performer stays in character and the cinema audience stays outside of the character’s world, unacknowledged by the character.
For the purposes of this inquiry, we will call this positioning of the dancing figure: character.
Reality TV dance contests present dancing figures as dancers. They are quite explicit about this, frequently reminding us that the dancing is being watched and judged. In these Reality TV dance contests, the dancers will have an opportunity, after their short performances, to turn to the judges or look directly ‘down the barrel’ at home audiences and appeal to us to judge their performance kindly. In this way, Reality TV dance contests hark back to early cinema. They present the dancing as a spectacle and acknowledge that it is designed and executed specifically for the spectator’s gaze, as an attraction or an exhibition.
The onscreen audience and the television audience are specifically acknowledged in the structure of the contests, where audiences have the opportunity to vote. What we vote for is our favorite dancer, so we will call this positioning of the dancing figure: dancer, meaning a dancing figure who is self-consciously executing dance for the spectator’s gaze rather than a character dancing inside the closed world of a story.
Documentaries about dance and dancers also feature explicit positioning of figures on screen: they are people. The dancers are the skilled people who are the subjects of the documentary – their dancing is seen as work being done by people. Sometimes we hear the stories about what they have done and what motivates them as in the 1972 documentary called I am a Dancer (Pierre Jourdan), about Margot Fonteyn, or sometimes we don’t, as in the 2009 documentary La danse – Le ballet de l'Opéra de Paris by Frederick Wisemen. But in either case the dancers are still people – people telling a story about doing their job, or people just doing their jobs.
People in documentaries are usually asked not to break the fourth wall by ‘looking down the barrel’ and addressing the camera lens, and by proxy the audience directly. The implied contract between audience, filmmaker and subject is that a filmmaker is making a film about a person and audiences are watching the film, not interacting directly with the person in the film. The subject interacts with the filmmaker, not the spectator. So, we are not invited to watch their dancing as exhibition, rather, we are invited to watch their dancing as an instance of them at work, or as an illustration of their story about their work, in the same way we might watch a clockmaker build a clock. While building the clock the clockmaker does not acknowledge us watching him, he is positioned as a person being observed at work, just as is the dancing figure in a documentary. So we will call this mode of positioning the dancing figure: person.
This inquiry will now turn to screendance as a discrete form of creative endeavor. Definitions first: screendance in this discussion is a form usually but not always found in short works, generally created with an artistic rather than a commercial motivation, unrestricted, for better or worse, by the requirements of feature film narrative, reality TV, or documentary form. However, rather than define it by what it is not, I will put forward the case that there are three genres of screendance which fulfill all of the above criteria and also have distinct characteristics of their own. These genres, and their many potential sub-genres, will form the basis for framing the discussion of whether the dancing figure on screen is being positioned as a character, a dancer, a person or alternatively a fourth category, not so far discussed, that of dancing figures as ‘form’.
The opening question of the American Dance Festival State of the Art Conference in 2006, put to us by conference convener Doug Rosenberg was: ‘can we propose a theoretical paradigm for understanding screendance?’ This question ultimately led to a Venn diagram of three overlapping disciplines: dance, cinema, and visual art . Unlike the typical result of these models, it was determined, however, that the ‘ideal’ screendance production was not necessarily a mix of all three. Rather, that each approach and each overlap provided a way of comprehending a given work. For the sake of clarity I have identified each of these different genres by a name that distinguishes it from the other two. The names are: dancefilm, video dance and dance on camera. Their distinguishing characteristics, and discussion of the ways in which each positions the dancing figure are detailed below.
A dancefilm, which is working in the overlapping areas of cinema and dance, will prioritise the directorial vision and emphasise the collaborative coordination of all of the elements of cinematic production, making its meanings through script development, mise-en-scene, montage and sound design. Like narrative cinema, dancefilm positions the dancing figure as a character, so their dancing is inside a closed world, something a character is performing either for the character’s emotional expression, or, less typically in the dancefilm genre, for an onscreen audience. These dancing figures do not break the fourth wall, they do not look down the barrel or engage in direct address, rather they inhabit a closed fictional world as characters with, as Gunning puts it, “causality, narrative suspense, and the creation of a consistent fictional world” (Gunning 1994: 190)
Two dancefilms provide examples of the use of cinematic techniques to position figures onscreen as characters rather than people, dancers or forms: No Surrender (Richard James Allen, 2003) and Break (Shona McCullough, 2006).
In No Surrender (Figure 1) a camera awakens and pursues a woman, it is a threatening force that motivates her to escape, and eventually to fight back in a sequence of choreography that entraps the camera. The figure onscreen is a character with motivations and when she dances it is in pursuit of a goal the character is motivated to achieve.
Figure 1: No Surrender, The Physical TV Company in association with the Australian Broadcasting Corporation, chor/dir Richard James Allen, 2001.
In Break, (Figure 2) a family communicates physically, first in naturalistic gestures, then in increasingly abstracted dance language, but still framed as characters whose movements have internal causality. The emotions expressed by this movement belong to the character, not the dancer, and the cinematic and choreographic representations of this emotion stay within the techniques and aspects of storytelling that create a consistent fictional world.
Figure 2: Break, Human Garden Productions, chor/dir Shona McCullagh, 2006
A video dance is a work based in the thinking of a video art maker, a performance art maker or a visual artist. It will have its effect through techniques, theories and premises of those disciplines. It prioritizes aesthetics over story or is structured by patterns of sensory stimuli such as colour, movement, light, shape, line, sound or texture. A video dance in this sense presents a new positioning of the dancing figure. By transforming bodies into abstract patterns of movement, light, shape, colour, etc., it presents the dancing figures as: form.
An example of video dance in this sense would be Gina Czarnecki’s 2005 composition Nascent (Figure 3) made with the dancers of the Australian Dance Theatre. Czarnecki clearly positions these dancing figures as forms. These human shapes are plastered against a black wall and manipulated as one might manipulate paint or water or some other plastic substance to make streaks, flows, layers of shadow and light.
Figure 3 -
Nascent. Directed by Gina Czarnecki. Forma Art and Media UK, co-produced by Australian Dance Theatre. 2005.
Video dance thus constitutes my fourth and final frame for describing the positioning of the dancing figure in relation to the spectator. In summary these are:
Fictional narrative cinema, including the screendance genre known as dancefilm positions the dancing figure as a character, shielded by the fourth wall, watched voyeuristically by an unseen and unacknowledged audience. Its creation is governed by directorial choices about mise en scene, montage, music and sound to cinematically convey a narrative and its meaning.
Reality TV returns us to early cinema and positions the dancing figure as a dancer, directly addressing the audience after their display of dancing and asking for our judgment.
Documentary positions the dancing figure as a person, doing their work, which we are privileged to observe, voyeuristically, unobserved ourselves. It may be that person’s work to create a fictional character, as Margot Fonteyn creates the character of Juliet, for example, but in documentary we watch the real person at work, not the journey of the fictional character.
Video dance or visual art driven screendance, positions the dancing figure as a form, abstracted. Interestingly, this also returns, with a slight variation, to the cinema of attractions. It is not that the performer acknowledges the audience, but that visual art is ‘exhibited’. Video dance is time based visual art and we watch it (sometimes literally, in a gallery), as an exhibition of the art.
So where does this leave dance on camera? Is it narration or exhibition or both? First a quick definition of ‘dance on camera’
Dance on Camera
Dance on camera prioritizes dance as its central discipline rather than visual art or cinema. Dance on camera foregrounds the composition and exhibition of the danced movement.
This definition would seem to qualify it as exhibition. But the performers do not acknowledge the audience, as in exhibitions of early cinema or reality TV, nor are they usually abstracted, as in an art exhibition. We will often see the whole dancing figure, since this will be the best display of the danced movement, but seeing the whole figure does not make this a documentary about a real person. We are seeing the dancing, not seeing the person who, as part of their life does the work of dancing. Then are they characters in a danced narrative? Certainly the conceit of the ‘fourth wall’ is in place, as it would be in a live performance of dance. However, as is also the case in many live dance works, none of the other hallmarks of narrative are in place. To return to Tom Gunning’s definition, there is no “causality, narrative suspense, and the creation of a consistent fictional world” (Gunning 1994: 190).
The mix is, as I say, complex, and dance on camera unfortunately, in my experience, frequently fails to consciously navigate these complexities, resulting in a lack of clarity about the positioning of the figure. I make an example of Strand (Michaela Pegum, Siobahn Murphy, and Dominic Redfern, 2010), a worthy effort, with tremendous potential, unfortunately let down by lack of clarity about the positioning of the dancing figures.
Strand is set in the Australian desert. Two women dance side by side in this desert. They are usually framed for the optimal clarity of display  of their movements and gestures, which is to say that the framing displays the dancing, but does not itself carry significance, except for rare, but stunningly beautiful extreme longs shots when the dancers are made into tiny figures in a vast landscape. In these rare moments, the positioning of the dancers jumps from being ambiguous to being clear: at these moments, they are characters in an implied narrative.
The extreme long shot gives their gesture causality – they dance in defiance of vastness and emptiness. The frame gives their gesture narrative suspense – I want to know, will these dancing figures triumph against this vast empty world or be swallowed by it? Finally the frame gives the dancing figures a fictional world. The long shots present an image of humanity valiantly creating significant form in an empty desert, characters motivated by their very humanity to make art, to dance, against the futility of vastness and lost-ness. The poignancy of the human condition is fleetingly captured.
Then it is lost when the camera repositions. Not only are the beauty and meaning dispersed, but also it is possible to remember that this was an image created by a camera crew, not really two women in an endless ocean of air. Making me conscious of the machinations of image-making ruins the effect of the image completely for me, I am jolted into remembering that humans are not tiny figures alone in a vast landscape, rather that the artifice of making such an image involves cameras and camera crews – they are anything but alone. Perhaps this was intentional, and the filmmakers can, of course, pursue whatever intent they have. But in this case, as in so many works of dance on camera where the display of the dancing is prioritized over the meaning created by the mise en scene I question their choices in positioning the dancing figure. Why make them a character and then make them a display in such rapid alteration? Which frame is more important to my understanding of the filmmaker’s meaning and intent: the narrative frame or the exhibitive?
When it fails to position the dancing figures explicitly as characters, dancers, people or forms, and fails to navigate the complexities to make a virtue of the ambiguities, dance on camera sits uneasily between what Gunning calls cinema of attractions and narration, losing the strengths of either form. I believe it is possible to hybridise narration and exhibition with dance on camera, but it relies on a crystalline comprehension of when each mode is being employed, and why, if it is to be fully convincing.
Altman, R., (1989) The American Film Musical BFI, London
Gunning, T. (1990) “The Cinema of Attractions: Early Film, Its Spectator and the Avant-Garde”, in Early Cinema: Space Frame Narrative, ed. Thomas Elsaesser, London: British Film Institute,(pp56-62)
Gunning, T. (1994) “The Whole Town’s Gawking: Early Cinema and the Visual Experience of Modernity”, The Yale Journal of Criticism, 7.2: Fall, 189-201.
Pearlman, K. (2006) “A Dance of definitions” RealTime, http://realtimearts.net/article/issue74/8164 Rizzo, T. (2008) “YouTube, The New Cinema of Attractions”, http://scan.net.au/scan/journal/display.php?journal_id=109, accessed 12 November, 2010.
 The original diagram had many contributors – almost everyone at the conference had something to say on the subject, but in particular, Katrina McPherson, Douglas Rosenberg, Ellen Bromberg, AllaKovgan, Richard James Allen and Bob Lockyer engaged in the discussion, making significant contributions. The discussion and diagram were first written up in RealTime, Australia’s National Arts Bi-Monthly, and the article is still available here: http://realtimearts.net/article/issue74/8164
 Displaying a dance via a camera is not the same as mise en scene, which is the art of staging and framing movement, space and design to convey an idea or a particular significance.
Dr Karen Pearlman is Head of Screen Studies at the Australian Film, Television and Radio School and co-artistic director of The Physical TV Company.