What is a model, why people don't trust them, and why they should

It is easier to make one’s way in the world if one has some sort of expectation of the world’s future behaviour. Even when facing a very complex problem, we are rarely in a state of full ignorance: some expectations of system behaviour and the level of risk arising from uncertainty are usually available and it is on the basis of these expectations that most decisions are taken. Humans use models, which are mental or formal representations of reality, to generate these expectations, employing an ability that is shared more or less by all forms of life. Whether it is a tree responding to shortening day length by dropping its leaves and preparing its metabolism for the winter ahead or a naked Pleistocene ape storing food in advance of winter for the same reasons, both are using models. This view leads to two outcomes. The first is that predictions, seen as an expectation of ranges of future behaviours, are not just desirable, but necessary for decision-making. The often-asked question ‘do models provide reliable predictions?’ then shifts to ‘given a certain problem, what type of models provide the most useful and reliable prediction?’ The second outcome is that modelling is no longer a scientist’s activity but is instead a social process. Different types of models can be employed to ensure that all available information is included in model building and that model results are understood, trusted and acted upon.