Michael Batty

Environment and Planning B: Urban Analytics and City Science Volume 50, Issue 7

A basic canon of the systems approach applicable to any field is the notion that a system is separable and distinct from its wider environment. In short, to formally study such a system, it must have a well-defined boundary beyond which it has no substantial impact on its wider context, while its wider context is usually composed of similar systems which have minimal impact on the system in question. The implication is that the environment defined by its boundary ‘excludes’ any significant actions or interactions essential for the functioning of the system itself. This is, in some respects, equivalent to the notion that we are defining a closed system which we can study in isolation from any extraneous or exogenous factors that might affect its operation. It is the definition used by Karl Popper (1959) to justify the use of the classical scientific method as fashioned in experimental science where the laboratory must be closed from the outside environment for robust theories to be tested and validated. In the case of cities, historically or at least from the middle of the last century, such boundaries are typically defined to minimise the overall interactions between the system and its environment. The implication is that insofar as there are many distinct systems, to minimise the interactions between one another, they are often arranged as a hierarchy. To minimise the exchange of energies between the system and all the systems within its environment, a good working definition of a system is that it contains the most significant interactions within the system itself (Simon, 1969). This question of course turns on what is regarded as ‘significant’.

Read the full article at: journals.sagepub.com

• Papers