INTRODUCTION: The Data

We now understand the structure of One Flat Thing, reproduced (OFTr) as a form of counterpoint that is created through the interaction of its three systems of organization: movement material, cueing, and alignments. At the beginning of this project those systems had a variety of names, the precise characteristics of which were hard to articulate. It took a collective effort to catalog and interpret the work as a totality. At the center of that effort and understanding is the data of Synchronous Objects.

The process of decoding OFTr was a creative dialog that dilated between insider accounts and outside observation, analytical needs and aesthetic interests. It was a profoundly creative and collective endeavor conducted over three years in close collaboration with William Forsythe and dancers Jill Johnson, Christopher Roman, and Elizabeth Waterhouse.

As we came to fully understand the counterpoint that unfolds in OFTr, we worked to devise methods for quantifying it in the data and expressing it in the objects. We identified what structures were in the dance; gathered the relevant data and found a way to both store and access it; considered multiple ways of understanding the dance; and then standardized terminologies. This effort produced two key sets of data: spatial data taken from our source video of the dance and attribute data gleaned from dancer accounts. You can read more about those sets of data in the column to the right.

Our goal in gathering spatial and attribute data was to discover patterns of organization that we could use to create the objects featured on this site. We weren’t concerned with documenting or reconstructing the dance for the stage, nor were we concerned with purely scientific questions. Instead we worked with the Forsythe Company to unearth the choreographic building blocks of OFTr, quantify them, and repurpose this information visually and qualitatively. As in many forms of inquiry, quantification requires a reductive process that necessarily obscures certain aspects of knowledge (the dancers’ intentions, performance quality, and kinesthetic awareness) in order to reveal others (in this case, choreographic structure). We drew from the methodologies of many disciplines—dance, design, computer graphics, geography, and statistics—and invented new methods when needed.

Spatial Data
Our animators generated location coordinates of the dancers by tracking a single point on each dancer in both the top and front views of the source video of OFTr. By combining the coordinates from both views we were able to generate a three-dimensional data point for each dancer’s location at every moment of the dance. Many objects, including Movement Density, Generative Drawing Tool, and Cue Visualizer, make use of this spatial data to visualize the choreographic structures of OFTr.

Attribute Data
This data set is built from the dancers’ firsthand accounts. We cataloged when dancers said they gave or received a cue, what alignments they were aware of (called sync-ups), when they were improvising, and what themes they were performing in every second of the dance. When dancers’ accounts differed, we made a determination based on other data and our own observations. The attribute data catalogs the three systems of the dance: movement material, cues, and alignments.

Movement Material: In their accounts dancers noted when they performed a theme (set choreographic sequence) and when they improvised. We recorded the theme name and its duration. We then analyzed these in relation to the accounts of our project advisors and other performers in the company. We had many discussions about terminology and decided that for clarity’s sake we would number these themes (T1, T2, T3, etc.) based on when they first appear in our source video of the dance.

Cues: We noted when a dancer is giving a cue for another dancer to move and who is receiving it. We also made note of any performative nuances mentioned by the dancers in their accounts.

Alignments: The entire system of alignments was not quantified in the attribute data as it is based more on observation and Forsythe’s choreographic eye than on dancer accounts. There is a subset of alignments, sync-ups, that the dancers were aware of because they had been instructed to insert a specific alignment at a precise moment. Because they were contained in the dancers’ firsthand accounts, sync-ups are quantified in the attribute data. While not quantified, other alignments that Forsythe uses in OFTr are annotated in the Alignment Annotation object as well as the Full Video Score.

Because we focused on the dance as a choreographic resource—rather than scoring it for the purposes of preservation—we were empowered to take this rigorous process of data collection into new creative spaces. We hope our choices, aesthetic and analytic, generate new possibilities for ongoing creativity and research, both in the studio and in the lab.

—Norah Zuniga Shaw, Columbus, Ohio, March 2009