ABSTRACT
What can we learn from the experience of other disciplines about the geography of cities? The answer to this question comes in two parts. First, there is enough similarity of type of problem to give us some confidence that theoretical geography is now functioning at the frontiers of science in general. Indeed, many of the non-linearities and associated complications (eg. the {W k}, k ~ j, backcloth) of geographical models suggest that some of the geographical problems are harder than those in some other disciplines. Secondly, by investigating methods in other disciplines, it is still possible to obtain new ideas for geographical modelling. We offer two examples of this below, the first the use by some members of the Brussels school of the genetic switching idea; and the second, the application of May's ideas on difference equations to the dynamics of shopping cent res.
