22 is a live performance for the stage in which projections interact in realtime with the dance and storytelling of Bill T. Jones as he is motion-captured on stage.
22 entails real-time motion capture in front of a live audience as Bill T. Jones improvises a complex narrative dance. At performance begins, Jones establishes a series of 22 named poses, which he then traverses in set order for the remainder of the work, improvising the movements between each pose. He also begins telling two stories, each featuring a child in terrifying circumstances. His narration abruptly intercuts between the two stories, and it is also interrupted as he names a given pose as he reaches it.
For the projections, we created a somewhat analagous system of repetitions, in this case creating seven “stations” (ladder, viewfinder, tightrope, window, door, box, table), with the similar rule that we would also rotate through these stations in sequence. (The quicktime above shows three such stations as established in the initial part of the performance.) Our stations are neither static nor fixed: they unfold in time and they varyin appearance (a box might be pushed in one round, but opened in the next).
The projections are cued either by the stage-manager or by the motion analysis recognition of a given Jones pose. In one sequence, the “door maze,” the progression through the imagery is entirely driven by Jones’ movement across the stage. (see example, which also illustrates a novel way to align video projection with 3D space).
This work let us explore AI agents embodied in representational rather than abstract figures, but much of our research — as well as more than three quarters of our actual imagery — were never incorporated into the finished work.
Artists’ statement
We began with the instinct to avoid duplication. Clearly there was no need to create a virtual double of the live performer, for the physical presence of Bill T. Jones on stage speaks powerfully for itself.
Instead we would evoke a presence just the opposite of his: a silent child, who watches and explores and teeters overhead. When first introduced into the collaboration, the idea of this child guided Bill in his choice of the two stories he tells. And the nature of these stories led us in turn to evoke another figure as well, a man: not the agile dancer we see on stage, but a slower-moving cumbersome figure who remembers, who photographs, and who does hard labor.
The strict structure of Bill’s dance, with its clockwise repetition of successive poses, inspired us to create a parallel (but not identical) form. It too runs through a set order, but with 7 rather than 22 elements. These elements are what we call “stations” rather than poses, from which they differ in two ways: they unfold over time rather than freezing into place; and they repeat only their themes, not their pictures. Thus table first materializes as a desk at which the man sits to write, but returns as a bed on which the boy lays himself down to rest.
Bill T. Jones weaves his way through the rigid choreographic structure he has imposed on himself, finding room to move and to readjust in between the fixed poses he has to hit. The artificial intelligence running our projections must find a way to match that flexibility. And so it peers out at the stage through the infrared cameras trained on the dancer, looking for times and for places in the dance for it to witness, to reframe, or to interpose.
We project scenes that find no exact equivalence in Bill’s stories. He has no tightrope walker, no boy pushing through a hinged labyrinth, no man shouldering a ladder. But he does have boys perched precariously on the edge, and souls lost in fear and perplexity, and men bearing the heaviest of burdens. His protagonists and their situations intersect and interact with ours ambiguously, leading you to draw multiple, uncertain, and unnerving connections between them.
Performances & credits
22 has been performed with live motion-capture only twice: at its premiere at Arizona State University on April 19, 2005 and at the Monaco Dance Forum in Monte Carlo on December 13, 2006.
Additionally, a fixed version of the projections accompanied a dance tour by Jones entitled “As I Was Saying…” in 2005-6, when it was presented at the Kennedy Center in Washington, DC; the Walker Art Center in Minneapolis; the Wexner Center in Columbus, Ohio; and Zellerbach Hall in Berkeley, California.
