Abstract Focuses on the levels of complexity underlying the construction of environmental issues and on the ways in which scientific techniques tend to simplify reality in an effort to control them. To demonstrate how such simplifications take place, it explores the cultural politics involved in constructing scientific knowledge in this field. It presents two case studies: firstly, a political–ecological description of an analytical study of soil erosion in the Oaxaca region of Mexico; secondly, of a computer‐based model of ecological and social arrangements in nomadic pastoral regions in sub‐Saharan Africa. These examples illustrate the process of simplifying abstraction at work and how investigative practice can itself affect policy‐relevant outcomes. The chapter concludes with recommendations for a more explicit approach to the assumptions and choices that simplifying abstractions can involve and shows how such an approach could enrich the understanding of socio‐natural phenomena under investigation.