Published in Dance Theatre Journal, Vol. 17: No. 4, January 2002
The Software for Dancers project set out with a clear goal in mind: to explore the idea of developing computer software that could be useful to choreographers as a rehearsal tool. It was not looking at how software might be used in stage performance, in archiving, or as a performance medium itself ('digital dance') - although it was certainly recognised that the outcome of the project may have uses or repercussions in these areas.
The project, which took place over 2 weeks in October with funding from the Arts Council of England and the support of Sadler's Wells and Random Dance Company, had been formed as the convergence of two previous proposals. The first came from Random's choreographer, Wayne McGregor. Accustomed to using software in trying out ideas for his own work, he was interested in how other choreographers might creatively respond to ideas emerging from computer processes. He had specifically wanted to involve a group of experienced choreographers who, though new to using computers in their creative work, were nevertheless skilled in thinking abstractly about processes and composition, and who may therefore find creative stimulus in computers.
The second proposal came from Amsterdam-based Scott deLahunta of Writing Research Associates, which specialises in performing arts research. Aware of many of the current developments in software for dancers and choreographers, he felt there was space for a software tool specifically designed for use in rehearsal rather than for staging or performance.
Those two different but overlapping motivations reflect the dual role of 'rehearsal' in dance. More than in any other performing art, dance works are not just produced or interpreted in rehearsal, they are created in, made through and emerge from the rehearsal process. Even within the intentionally restricted area of exploring the idea of a software tool specifically for rehearsal, this tension between practical and creative uses of software was an ongoing undercurrent throughout the project.
Beyond these parameters, the project set out with no preconceptions about how to move towards its aim, or indeed what the goal might look like when it was reached. That, it was hoped, would emerge from the participants, all invited for their experience and knowledge within their fields. The project was co-ordinated and managed by Scott deLahunta. The four participating choreographers were Wayne McGregor, Siobhan Davies, Shobana Jeyasingh and Ashley Page. Though they had already been on 'orientation sessions', they had (with the exception of Wayne McGregor) little or no experience of using computers as part of their artistic or rehearsal practice. On the digital side, the software developers were all chosen for their experience in working with artists or as artists themselves: Christian Ziegler, the designer of the Improvisation Technologies CD-Rom for Frankfurt Ballet; digital visual and performance artist Guy Hilton; digital sound and video artist Joseph Hyde; software artist and programmer Adrian Ward; and digital media artist Bruno Martelli of Igloo, who contributed to the preparation discussions held earlier in June.
The medium is not the method
The first sessions were mainly open-ended discussions, with no specific goal except to try to find a common ground between the two participant groups. This was no straightforward task: there were many possible avenues to explore, complicated by the participants' wide variety of interests and working methods, and not least by problems of communication produced by terminology, jargon and preconceptions. Still, at the end of these initial sessions, several basic concepts were clarified, goals narrowed, and criteria set.
First of all, what do computers do? Fundamentally: modelling, manipulation and 'memorising'. That is, they represent information, they analyse and process information, and they store information. That information can be text, audio, image or video, and it can be either sampled (recorded from video, camera, scanner or microphone), or synthesised, that is, generated by the computer itself.
An example of sampled information is motion capture technology. A performer is rigged up with lights at key positions on his/her body; a camera then records their movement, and the computer reconstructs the movement as animation. It is occasionally used in feature film, such as the latest Star Wars; it was also used, for example, in Merce Cunningham's Biped, though here there was no attempt to render a lifelike figure.
An example of synthesised information can be found in Lifeforms software; its most well-known user is also Cunningham. This animation package has a virtual figure with articulated limbs that can be moved into different positions. In essence, the animator creates positions of the body and limbs at different points in virtual space, and the programme 'tweens' the motion between them - that is, it constructs a pathway for the figure to move along between one position and the next.
With the rapid advances in 3D animation technologies - whether based on sampled or synthesised data - it was not surprising that the first ideas discussed revolved around the notion of a more or less lifelike animated model of the moving human body - in other words, a simulation of the choreographic medium. However, this avenue was fairly quickly abandoned because of a number fundamental problems.
First, any modelling system has its own ways of handling, filtering and representing information - and these parameters limit the ways that they can be used. Lifeforms, for example, works best with shape, design and composition - it has a very visual, external sense of the human body. But it is not very effective with, for example, gravity and acceleration, internal body impulse, or flows of energy. Choreographers less concerned with the design/shape aspect of the body may not find it interesting or useful.
Furthermore, 3D animation programmes aiming for a lifelike simulation of body movement take up a lot of time, processing power, and disc space. They are require expensive equipment and skilled users to operate them effectively, and are potentially very time-consuming. So: fine for the cutting edge of digital technology, but certainly not accessible or even practical for the choreographer.
Instead of looking towards simulating the choreographer's medium, the research group then began to look at their methods - a line of enquiry which proved much more fruitful. Of course choreographers' working methods are wildly diverse in the way they create, the role of the dancers, the relation of idea to movement, the choice of subject and inspiration. There is simply no adequate model for that; if there were, choreographers may as well throw in the towel.
However, at a more general level, there were some practical methods common to all the choreographers, albeit used for different ends. First among these was using video, now an almost ubiquitous rehearsal tool, whether as a record of material made, a way of teaching and learning material, or a way for dancers and choreographers to reflect on what they are making together. Then, all the choreographers used some form of note-making, such as written text or drawn diagrams and sketches, which could include instructions to dancers, reminders of ideas for themselves, references to sources of inspiration. And those sources of inspiration could come in any medium - music, sound, text, an image, a TV programme, a book, a building.
Although the content of these notes and recordings, as well as the uses to which they are put, varies widely from one choreographer to the next, nevertheless all the choreographers do go through the process of taking notes and making recordings. One idea for an at least potentially useful rehearsal tool was becoming clearer, in the shape of some kind of multi-media notebook.
There were two different proposals for the multi-media notebook, the first based on video input, the second on a stage plan combined with a timeline. Both would aim to use the computer at what it does best - representing, storing and processing information - while at the same time keeping in mind the all-important criterion of usability.
The primary data input for the video-based tool would be studio video recordings - already commonly made as part of rehearsal practice. The software envisaged was based around a simple and intuitive video interface - such as iMovie, with its straightforward but effective tools for basic editing of clips. But in order to function as a notebook, the choreographer also has to be able annotate the recordings - something that's hard to do with videotape, but much easier with digital movie clips. These notes could take several forms. One might be to draw directly onto the screen, to indicate, for example, directional lines of energy or flow at a particular point in the recording. Other annotations could take the form of text notes or other diagrams - or indeed any other type of media input (sound or voice for example) - that could be linked to certain sequences or clips.
Of course, once the video is in digital form, it is easy not only to annotate studio footage, but to try out certain kinds of ideas. For example, a clip could be played back and tried out with different soundtracks; or could be sped up, or sequenced differently - all activities that take a lot of time in the studio, but could be easily tested out with a tool such as this.
Like the iMovie annotation tool, the stage plan idea - which emerged mainly from discussions between Ashley Page and Christian Ziegler - is also for a multimedia notebook, but here based on a visual map of the stage space coupled with a timeline. The map can include entrances and exits, props, scenery and so on; onto this can be placed sound files, video, text or drawings, again linked to a time sequence. Pathways can also be drawn onto the plan - and video clips, for example, could move across the stage plan over time. The effect then is of a multi-media, but highly schematic representation of the progression of a piece, a kind of staged storyboard which can include sound, video, image, text and so on.
Two important factors for the usability of these proposals are their schematic rather than life-like modelling of the rehearsal process, and their flexibility and adaptability. They enable recording, annotating and representing of information from rehearsals, not as a simulation of rehearsal but as an adjunct to it. Furthermore, they enable a range of different types of input (sound, image, etc), but at the same time allow them to be used in different ways (instruction, direction, mnemonic, cross-reference, etc) according to the needs or interests of the choreographer. The notion is, too, that they will be adaptable to different uses through enabling plug-ins - modules of extra code that expand the software's capacity in specific ways.
Both these ideas focus on practical working processes; but as mentioned above, dance rehearsals are not only a means to staging but also a means of creating. Running through the project was also the thought of developing a very different type of idea, one that could harness the computer's processing power to stimulate choreographic ideas. Again, this started with an idea of simulation. Choreographers take ideas from any source and transform them - through experiment, invention, interaction, chance and play - into movement. Given that transformation of data is something that computers do well, the question arose: could computers model any part of that process?
The answer again, was: not directly. The ways, ends and means of different choreographers are too diverse, unpredictable and ever changing. Instead of simulation, however, computers could be harnessed as a source of stimulation - providing choreographers with source of ideas that they may find interesting not as a model of their artistic process, but as a stimulus for it.
The research project explored some ways in which computer processing could transform media input (sound, image, text, etc), for example by randomisation, geometric transformation or organic transformation. This last was potentially the most interesting, a set of rule-based transformations that have closely modelled 'natural' life-like forms (plant growth for example, or the movement of birds within flocks). By the end of the project, however, this avenue of exploration was still at the embryonic stage.
To put this research into a context of current developments, the results of the Software for Dancers projects were shared in a two-day seminar with other groups also developing software in conjunction with dance making and performing: Igloo (London), Troika Ranch (New York) and Barriedale Operahouse (Vienna). Barriedale has also been developing an idea similar to the stage plan notebook concept. And finally, these groups and the Software for Dancers group presented their work at an open forum at the Clore Studio in the Royal Opera House.
Meanwhile, the participants had had their own discussion of the findings. Some of the potential advantages of the tools were clear: for some parts of the rehearsal process, it could save studio time and space; it may also be a neat and portable way to share information, both between choreographers and dancers, and between choreographers and collaborators such as composers and designers. But there are potential disadvantages too. For example, sketching on the computer may simply take up more time, not less - anyone familiar with writing on a word-processor knows how easy it is to fiddle with the editing and cut and paste features, trying our ideas without committing to them. So expanding the range of choices and the ease with which those choices can be changed does not necessarily lead to more effective use or results.
So for the development of the software tool to lead to effective use, a number of criteria need to be kept in mind: is it usable by choreographers? Is it also useful to choreographers? Does software already exist that could be used or adapted? And finally, regardless of usability, can choreographers simply work better without it?
This was the first stage of a proposed development project, and the answers to those questions will doubtless be developed in the next phase, still in the planning stage, of Software for Dancers.
Sanjoy Roy is a writer, editor and designer.