ABSTRACT
Modelling is central to science and should be central to pupils’ understanding of what scientists do, so there is every reason to clarify just what we think it involves. To do so we might begin with physical models, which are a great support for scientists and teachers, and in the classroom there is a reassuring definiteness about having something tangible on the bench-an orrery as we talk about the solar system, a plastic ‘replica’ of the eye if we are discussing vision, or even just a length of Visking tubing when we discuss how intestines work. It helps to have something to handle and look at-so much better than ‘mere talk’.1 At another level, however, models are ‘mere talk’. None of the useful toys I have mentioned could have materialized on the bench at all without their designers being guided by a mental process in which the objects helped them to explain what they had in mind. First and foremost, it is mental models which matter, and my purpose is to emphasize the role of language at two important stages in their generation and use:
• Stage 1 — when someone is developing a new model (the ‘re-describing’ stage). • Stage 2 — when someone else is trying to appreciate it (the ‘persuading’ stage).
