Epilogue

Abstract

Considerable work has been published in recent years by geographers, spatial economists, regional scientists, spatial statisticians, and scholars from other allied disciplines, on the topic of space-time transformations. A main emphasis of this work is description of the spatio-temporal trajectories followed by geographic systems. Good examples of affiliated dynamic spatial models are provided in Griffith and MacKinnon (1981), while interesting examples of some initial attempts to formulate evolutionary geographical models are furnished in Griffith and Lea (1983). Unfortunately many of these models tend to be linear in form, tend to treat time as reversible, and tend to view locations in space as independent entities. But spatial autocorrelation mechanisms, the irreversibility of space-time processes, and the multiplicative or exponential nature of descriptors of change highlight the inappropriateness of most dynamic as well as ‘evolutionary’ spatial models. As soon as non-linear formulations embracing interaction effects are coupled with bifurcation possibilities, the modelling game becomes dramatically changed.