?Qualculating? a low-carbon future-Assessing the performativity of models in the construction of the French net zero strategy

The aim underlying the notion of a Net zero 2050 future (NZ 2050) is the achievement of a balance between anthropogenic emissions of greenhouse gases by sources, and their removals by carbon sinks, by the mid-point of the current century. This paper delves into the relation between numerical modelling and political decisionmaking in the construction of a NZ future, by exam-ining the recent development of the French national low-carbon strategy (SNBC-2) (2016-2018) and related processes. SNBC-2 was the first attempt to put a NZ2050 France, and the path towards it, into a numerical, calculative form. We show that, aside from yielding a collective capacity to quantify farther into the future, one major lesson of the French attempt at representing NZ2050 was that available models for doing so were of limited relevance. The analysis discusses these shortcomings and their implications with regard to the performativity of numerical models in making NZ2050 a doable horizon. Performativity refers to a circular and dynamic relation be-tween modelling and the realm that they aim to represent (here, NZ2050 future). Modelling a future is thus conceived as an algorithmic and relational process, which contributes to the as-sembly of all the elements that might compose this future, or to bringing about their emergence. Based on the notion, drawn from science and technology studies, of 'qualculation' - which un-derstands calculation as a process that is at once qualitative and quantitative, relational and algorithmic - we underline the crucial, but ambiguous role of non-calculative adjustments used to compensate for the calculative limitations of models. We show that these adjustments can potentially lead to both gain and loss in performativity, depending on whether or not the qual-culation process makes them explicit and allows them to be collectively remembered and criti-cally discussed. We also argue that calculative limitations and the resulting adjustments show that the relationship between models and policymaking is somewhat more reciprocal than what has been argued in the literature: models do encapsulate a form of politics which they enact, but they also can be challenged and reshaped by the emergence of new political objectives.