to star items.

Accepted Paper

How Granular Data Becomes "Precise": Trained Judgments and Comparisons in Establishing Precision in Digital Agriculture   
Mylène Tanferri (IT University Copenhagen)

Send message to Author

Paper short abstract

What is granular data, and how does it become "precise"? Drawing on ethnographic research in digital agriculture, this paper examines granularity as a practitioner-made achievement, showing how precision depends on trained judgment and comparison with other ways of knowing.

Paper long abstract

Technologies of precision promise to replace uncertainty with accuracy. As in other domains, precision in digital agriculture is now tied to granular data. This data would be fine-grained enough to enable precise knowledge of phenomena, both external to plants and internal, such as temperature or chemical states. Yet, what is granular data, and how does it become precise? To answer these questions, I draw from interviews and ethnographic observations with digital agriculture actors. Inspired by material-semiotic insights from STS (Law, 2008; Mol, 1999; Lien & Law, 2011), I consider granularity as a practitioner-made achievement rather than a technical property alone. In doing so, I examine how granular data is brought into existence through heterogeneous practices, showing how plant lives are rendered "knowable, predictable, and governable through big data analysis". This examination speaks directly to the study of precision(s) through two main observations: first, data practices in producing and interpreting granular data depend on experimental infrastructures, interdisciplinary collaboration, and ongoing human interpretation and expertise, similar to trained judgment (Daston and Galison, 2010), in making sense of granular data. Yet these elements often disappear from claims about granularity and precision. Second, the value of granularity is premised on comparisons with other modes of knowing, such as traditional agronomic knowledge, rather than by empirical demonstrations. While granular data is often framed as an ideal of mechanical objectivity, this framing tends to obscure both the trained judgments on which it depends and the comparative work that establishes its advantage over other ways of knowing.

Traditional Open Panel P194
Technologies of precision: Exploring the meanings, practices, and politics of precisioning tools across healthcare, agriculture, and warfare.
  Session 1