The Illusion of Abstraction

• In Event: Schemas, Graphs, Ontologies - II

Abstract

The Illusion of Abstraction
paper abstract by Martha Lampland
for submission to the Schemas, Graphs, Ontologies Open Panel at 4S 2019


Abstraction is an illusion. Recent work on algorithmic culture has drawn our attention to the need to interrogate more carefully the aggregation of data and design of algorithms, lest implicit biases that justify pernicious inequalities be reproduced in their application (Ananny 2016, Dourish 2016, Eubanks 2017, Noble 2018, Radin 2017). This has prompted calls for “enacting algorithms ethnographically” (Seaver 2017 ) or engaging in the hermeneutics of data “at the micro-, meso-, and macro-scales of networked infrastructures” (Acker 2015). In this paper, I will argue that to do so requires we confront the reigning conceit that abstraction in formalizing practices cleanses data of its indexical meanings. By discarding the illusion of abstraction, we may begin to excavate the social and historical conditions of data’s production and trace these tenacles in its reuse. To illustrate, I will analyze a historical example of a mismatch between the meanings conveyed in standardized calculations which have heretofore been overlooked, notably the cultural conception of labor as figured in time and motion studies conducted to set wage rates in the mid-20th century. In doing so, I hope to demonstrate one means of recovering the indexical features of data, and contribute to a wider discussion about how to conduct research that incorporates the historicity of data production into studies of networked infrastructures.

Author

• Martha Lampland, Univ Of Ca-San Diego