Models refer to the conceptual or mathematical simplifi cation of reality; they are constructed in order to understand some aspect or process. For some aspects of physical geography such as climatology and understanding the relationships involved in chaotic and complex systems (Chapter 8), modelling is an essential tool to understanding; it could be argued that modelling is the only means available to facilitate our understanding of reality in these areas. Models are not reality itself and should not be viewed as such. Researchers who use models do not necessarily believe that their models correspond directly to reality. Nevertheless, on occasion some researchers may place too much confi - dence in models, and, likewise, some decision-makers may behave as if the models are reality. If this were not so then the shock associated with the failure of a model would not be such a surprise. Harvey (2010), in relation to the ongoing economic crisis, suggests that economic theorists (citing Samuelson in the Washington Post as his source) ‘became too interested in sophisticated forms of mathematical model-building to bother with the messiness of history and that this messiness had caught them out’ (p. 235). Add ‘space’ and ‘place’ to history and you have the nub of the problem. Models are simplifi cations, potentially complicated simplifi cations, but still simplifi cations of the messiness of reality. Belief in models as reality is a very correspondence-based view of reality (Chapter 2) and by implication the practice of science.