ABSTRACT

In many contexts, the scientific community produces multiple different, and sometimes conflicting, models of the same target system. What drives the construction of multiple models, and what strategies are there to cope with the resulting abundance? We begin by introducing the problem and by trying to understand why, and how, an abundance of models of the same target emerges. We then turn to Robustness Analysis, a method to try to extract veridical conclusions from model ensembles. Perspectivism offers an alternative approach which sees different models as embodying different perspectives on the same target. In some situations, there is significant latitude in model construction and scientists disagree over the appropriate way of modelling the target. These are situations of severe scientific uncertainty, and there is a question about how such situations can be managed.