One preferred way of coping with the openness or indeterminacy of future is to elaborate ‘fictional expectations’ that enable action by defining possible outcomes. In this article, I propose to focus on the career of the impact foresight model to analyse how agro-economists combine imaginaries, narratives, data and calculative technologies addressing the long-term future of agriculture. Impact is a partial equilibrium model, which has become increasingly comprehensive. Its modular structure now enables it to interweave scenarios produced by other legitimate institutions, and to run simulations for a number of configurations of climate change and socio-economic evolutions. In this article, foresight models are taken to be material discursive devices. My argument is that their evolutions as technologies and the framing of the future they operate should not be analysed separately. Transforming radical uncertainty into controlled variability – magnitude of change, they explicitly endeavour to ‘bound’ uncertainty. But it is ‘bounded’ in a way that is highly dependent on the knowledge infrastructure upon which the models rely. Quantified modelling also makes it possible for economists to compare rival models and create alignments or negotiate zones of consensus, that is, a certain form of knowledge on the future. In the case under scrutiny, technological choices and data processing work contribute to reinforce a certain point of view – market, production and technology-oriented – on food security. Studying infrastructure and model design therefore allows a better understanding of path dependency and cognitive lock-in effects regarding the way the future is envisaged and narrated.

中文翻译：

---

有限的未来：在粮食安全研究中发展边界先知基础设施

应对未来的开放性或不确定性的一种首选方法是阐述“虚构的期望”，通过定义可能的结果来采取行动。在本文中，我建议关注影响预见模型的职业，以分析农业经济学家如何结合假想，叙述，数据和计算技术来解决农业的长期前景。影响是部分均衡模型，已经变得越来越全面。现在，它的模块化结构使其能够交织其他合法机构提供的方案，并针对气候变化和社会经济演变的多种配置运行模拟。在本文中，预见模型被认为是物质话语设备。我的论点是，不应单独分析它们作为技术的发展以及它们运行的​​未来框架。他们将根本性不确定性转化为受控的可变性-变化的幅度，明确致力于“约束”不确定性。但是，它以高度依赖于模型所依赖的知识基础结构的方式“受限制”。量化建模还使经济学家可以比较竞争对手的模型，并建立一致或协商一致的领域，即对未来的某种形式的知识。在仔细审查的情况下，技术选择和数据处理工作有助于加强对粮食安全的某种观点-市场，生产和技术导向。