Models and Formats of Representation

Models are generally used by scientists to obtain predictions and to provide explanations about phenomena. Their predictive and explanatory power is usually thought of as depending on their representative power: a model is a device standing for a target system in virtue of some relationship between its features and features of the system. This representational relationship enables scientists to draw inferences and to obtain information concerning the target system by reasoning with and in some cases literally manipulating the model. It is still not clear, though, in virtue of which features models allow such inferences to be drawn. Some philosophers have argued that appealing to isomorphism between models and target systems is not su cient to explain how models are used by cognitive agents, in practice, to represent the systems they study. The fact that the structure of theories seems to adequately capture the "structure" of the world is a genuine philosophical problem roughly, the problem of the use of mathematics in science but it does not solve the problem of how scientists use representing devices to draw inferences about the systems they study. Examining this problem does not commit anyone to any particular conception of the structure of scienti c theories. When it comes to representation, one has to pay attention to the kind of cognitive operations and sometimes to the concrete manipulations scientists perform when they learn, develop and