Thinking algorithmically: The making of hegemonic knowledge in climate governance

Abstract

Algorithms – instructions for acting on data, executed by code – are increasingly being enrolled into climate policy governance via the prediction of policy outcomes, the evaluation of climate mitigation and adaptation strategies, and the design of practitioner actions. Yet the political implications of these technological changes in environmental governance are only just beginning to be theorised. In this paper, we examine one particular facet of this emerging politics: the relationship between thinking algorithmically and hegemonic power. Drawing from Laclau and Mouffe’s theorisation of hegemony we argue that algorithmic forms of reasoning lend themselves towards producing hegemonising knowledge regimes, with important implications for a democratic politics of climate change. Recognising that algorithms stand for wider socio‐technical assemblages that structure and create knowledge, we call for greater attention to the reliance on algorithms within climate governance – less for the algorithms themselves than for their particular epistemic commitments that create algorithmic ways of thinking, with associated claims to power. Through a critical review of scholarship at the intersection of critical digital studies and environmental governance, we first identify three key epistemic commitments involved in thinking algorithmically: induction, abstraction, and optimisation. We then examine the correspondence between these key features of algorithmic thinking and the conditions that Laclau and Mouffe propose form the grounds for hegemony: objectivity, universality, and necessity. Better understanding what “thinking algorithmically” entails, and the forms of knowing and acting that it affords and excludes, is vital, we argue, to begin naming the political implications and transformative potential of new forms of climate governance.

Geographies of science and technology II: In the critical zone

“…Butler, 2019), while more recently Andreas Malm (2018) has dismissed ‘constructionism’ as one of the theoretical edifices which keeps critical thinkers from acknowledging the severity of climate change, and from fully participating in the fight against it. This is not the place to revisit the ‘science wars’, but Castree (2021) offers a pithy defense of work which queries how (rather than whether ) climate change is constructed, and which interrogates the effects of dominant constructions (like carbon budgets, for instance) on collective efforts to govern the atmospheric commons (see also Machen and Nost, 2021). But the aforementioned studies of science and technology in ‘the critical zone’ all, in their own way, raise questions about the place and role of critiques of technoscience in a moment of apparently accelerating environmental crises.…”

Non‐ and sub‐state climate action after Paris: From a facilitative regime to a contested governance landscape

“…They highlight societal concerns such as injustices, gender inequality, and neo‐colonialism attached to the climate crisis (Dehm, 2017). Disruptive confronters promote alternative modes of knowing, like indigenous thinking, against “hegemonic knowledge” (Machen & Nost, 2021) and the dominance of modern science. These new perspectives potentially disrupt established ways of evaluating climate change and its effect on society, people, and nature.…”

Graph embedding and transfer learning can help predict species interaction networks despite data limitations

“…Indeed, the use of ecological data is not an apolitical act (Nost & Goldstein, 2021), as data infrastructures tend to be designed to answer questions within national boundaries, and their use often draws upon and reinforces territorial statecraft. As per Machen & Nost (2021), this is particularly true when the output of "algorithmic thinking" (e.g. relying on machine learning to generate knowledge) can be reused for governance (e.g.…”