How to Draw a Dog

What is walking, exactly? The choreographer Trisha Brown explored the movement in 1971 when she staged Man Walking Down the Side of the Wall, a dance piece that consisted of Brown’s then-husband Joseph Schlichter wearing a harness and pulleys to walk down the side of a building in SoHo. Two belayers pitched Schlichter perpendicularly on the seven-story building at 80 Wooster Street, where he balanced his body in a vertical walk down the wall and into the building’s interior courtyard. As Schlichter descended to the soon-to-be-gentrified sidewalk below, his continuous, forward gait felt at once pedestrian and heroic. When had walking ever been so alien and dangerous, and yet so utterly commonplace?

More recently, robotics engineers, GPS specialists, and prosthesis manufacturers, in an effort to authentically replicate the movement, have probed what walking is by breaking it down into its basic components. Like an algorithm, walking is a series of steps. Variables such as the body’s center of gravity, the slope of the ground, or the presence of other objects either diversify the possibilities of walking, or, in the question of robotics and industrial design, are problems to be solved. Replicating walking means replicating proprioception, the awareness of one's own body in space.

In September 2019, Boston Dynamics released its all-terrain robot dog, called Spot, and made it available for leasing to businesses for industrial sensing and remote operations. The company’s video ad campaign highlights the robot’s responsiveness to variables in space and terrain. The jazzy videos, free of spoken dialogue, evoke the succinct pedagogies of William S. Gray’s mid-century grade-school readers:

see Spot walk up a pile of debris
see Spot walk down a flight of stairs
see Spot course-correct after bumping into a cart full of supplies
see Spot rise regally from an inert and crumpled position after a fall
see Spot traverse spaces where humans cannot go

To Boston Dynamics, the 160-pound yellow robot dog is a bank of gestures, coded, selected, and stored; a compartmentalization of repetitive tasks that are no longer considered lucrative or safe for human beings to do. But to the public, Spot’s ontology remains in flux. A small canon of Spot’s test videos has gone viral, showing workers pushing the robot over and Spot clambering back up, dutiful and guileless. The machine’s misfortunes—and the ensuing public outcry—have drawn it into the realm of living flesh.

Continuity between steps is everything, it seems. This was Trisha Brown’s theorem, tested out on New York’s urban brickscape: that the flows we establish by walking on and within things are a frequently occluded but almost ecstatic form of engagement with the world. Brown knew, too, that human beings are banks of gestures. While each creature’s walk is a singular mix of ground, skeleton, and motion, it is also subject to a long and loving history of scientific schematizations—like Eadweard Muybridge’s animal locomotion studies of the 1880s. Perhaps this is why we can’t help but draw Spot back into the world of sentience. The robot breaking his delicate gait supports the fantasy of Spot as an individual, allaying our anxieties about what the robots can be used for, about the large and sinister totalities of which they are a part.

In 1981, Trisha Brown collaborated with Charles Gaines, a professor of art and visual theory at Cal State University, Fresno. Gaines was an African American conceptual artist in a mostly-white milieu, curious about how movement became data and, eventually, types and stereotypes. He drew things—trees, people—by plotting them inside grids, to isolate the tropes we rely on to categorize things. For the collaboration with Brown, Gaines photographed 25 seconds of her dancing in rehearsal and used the photographs to hand-plot the images of Brown’s body onto a grid. His approach invoked the ways that medical and military researchers use data to first condense motion into the smallest possible units, and then recreate the motion using synthetic materials. Robotic data builds a mathematical description of how legs and torso work together; then it’s translated into mechanical and electrical impulses. Mimicking data transcriptions that computer systems generally perform, Gaines showed how the human hand both failed and shifted in the face of systemic approaches. Imitations of the body created their own continuities.

Why does drawing haunt so many conversations about walking? Is it because, as Trisha Brown imagined, drawing and walking are key ways to feel information in space? Is it that, as Gaines showed, the act of transcribing or thinking through something is when we begin to notice its unpredictabilities? Or is it because data is always lying in wait to be mobilized—the way that Spot sits, faithful and leashless, at the construction site?