Analysis of causal argumentation in social-ecological systems research

Abstract

Social-ecological systems (SES) research is a field in which interdisciplinary collaboration is necessary to understand the causal complexity of phenomena such as biodiversity loss and climate change. However, interdisciplinarity poses challenges for evaluation of evidence and causal claims, which is a crucial for generating and integrating knowledge, prediction and informing policy. A single hierarchy of evidence is not possible when research approaches confront different practical and theoretical challenges. We argue that a way forward to evaluate claims is an analysis of causal argumentation, which is an understudied aspect of causation in SES research. Analysis of causal argumentation traces the reasoning linking data and causal claims, discloses assumptions, and makes explicit the support of robustness and generality that qualify claims. We demonstrate this approach by surveying a selection of the field’s most cited papers from the last ten years to analyse the justificatory strategies used to make causal claims. We found that claims in this small corpus are diverse and arguments rely on combinations of reasoning patterns, like inference to the best explanation, mechanism-based thinking, and Mill’s method of difference. There is no causal claim that has a general scope, high strength, and captures causal complexity. Arguments that better captured causal complexity supplemented the argument with causal models that are visually represented. Triangulation of multiple evidence produced claims about more aspects of causal configurations, and achieved more specificity, and stronger claims. Studies that dealt with smaller and simpler causal configurations were able to triangulate evidence more easily. Most of the studies could benefit from considering alternative explanations. We conclude that analysing causal argumentation helps to recognise the limitations of our strategies, make our arguments sharper, and decide what information is useful for decision making.