This work continues an investigation into what happens when a technical system attempts to name that which it cannot recognize. Drawing on footage of the Stations of the Cross, a computer vision system locates bodies, reconstructs their joints, and infers actions from a previously learned repertoire of categories.
These figures are embedded within a specific territory: roads, vegetation, sugarcane fields, and a landscape shaped by sugar production. Yet the model attempts to read this context through a different archive of relations. In this displacement, bodies and gestures exceed the technical repertoire of “global” labeling systems, and each failed reading exposes a distance, an excess of meaning that their categories cannot absorb. Thus, as Donna Haraway reminds us, their gaze—despite presenting itself as “universal”—is always situated and foreign to the worlds shown here.
In dialogue with Gayatri Spivak, I am interested in what translation displaces, reduces, or renders inaudible when it attempts to make the other intelligible through categories that precede it. Following Éric Sadin, this operation acquires a political dimension when technical systems no longer merely record the world, but produce statements about it: identifying, evaluating, and inferring what comes to appear as true.