Growth waves, sliding average and technical calculations of plant lives: A glimpse on time as organized by digital agriculture tools

This presentation is based on initial data collected for a post-doctoral project on digital agriculture (DA) and its accompanying sustainability promises. DA rests on numerous presuppositions and possible redistributions of agency and responsibility, which this project explores ethnographically in the everyday practices of developers and users of a specific innovation. Here, we focus on the “sliding, or moving average”, a statistical tool used to analyze the electrophysiological signals of tomato plants captured by a phytosensor in an agronomic greenhouse. I look at the links between these calculations and the temporalities of plants’ lives, growth, and needs, and I ask what this kind of statistical tool can tell us about how futures emerge from present practices (Bryant and Knight, 2019; Pink et al., 2022). As one participant explained: “A sliding average is not really a prediction where you would have a probability calculation, but a comparison tool. It compiles past events and organizes them into an average - where what happens in the present is compared with what would be expected given what has happened in the past – and as data is treated on an ongoing basis, this comparison is going on all the time”. The sliding average, thus, captures plant growth over time and produces an expected, yet moving, « normality ». When applied to plant signals, specific deviations, in the past or the present, can then be investigated, explained, and possibly controlled – by humans: using collective situated knowledge, memories, and logs of past events, external factors (a window left opened, a heatwave affecting the region) can be discriminated from potential internal issues in plants and lead to potential (non)- interventions.
While growing practices were always concerned with time, and in addition to being related to “farming-by-numbers” (Krzywoszynska, 2024), this specific DA application can be seen as a way of reorganizing the temporality of (un)controlling edible plants’ growth, just as the role of human memory and local knowledge about events in time are re-specified. Finally, I consider how such calculi can be of interest to studying food system temporalities, by providing a glimpse into how futures are being calculated (Doganova, 2024), from the ground up, in the details of exploratory digital growing practices.
Bibliography
Krzywoszynska, A. (2024) ‘“You can’t manage what you can’t measure”: Regenerative agriculture, farming by numbers, and calculability in soil microbiopolitics’, Environment and Planning E: Nature and Space, 7(4), pp. 1691–1710.
Bryant, R. and Knight, D.M. (2019) The Anthropology of the Future. Cambridge University Press.
Doganova, L. (2024) Discounting the Future: The Ascendancy of a Political Technology. Brooklyn, New York: Zone Books.
Pink, S. et al. (eds) (2022) Everyday Automation: Experiencing and Anticipating Emerging Technologies. London: Routledge.