Abstract
"Digital representations are ubiquitous in the workplace. Screen displays, forecasts, simulations, indicators, multi-dimensional models, figures, and images are increasingly central to work of all kinds. Representations are simultaneously transparent and opaque. They contain and reveal information about the organization. At the same time, they conceal the computational work used to convert data about the physical world into abstract depictions. Computational opacity is consequential when representations become misaligned with the physical world they depict. We examine how computational opacity can be breached, allowing non-programmers to repair misalignments between representations and the physical world. Drawing on an ethnography of a machine-shop floor, we show how operators develop practical computational literacy skills—the capacity to visualize and talk about physical objects and processes independent of them; to translate this noncomputational thinking and talking into computational symbols, syntax, structure, and assumptions; and to create computational solutions. We show how operators develop this skill vicariously, observing programmers as they solve problems. We contribute to understanding how, in an increasingly digitized workplace, those without programming capacities may decrease their dependence on programmers and increase their capacity to create and alter representations of the physical world."