The quantitative revolution plus 55 years: relevant, testable and reproducible modelling?

The quantitative revolution in geography in the 1960s and 1970s was an exciting time to be a post-graduate student. Like many social scientists, geographers at that time wanted to make the ‘intellectual music’ of science, the anthem of geography, from an orchestra of common methods and techniques. It is therefore with great appreciation that we thank Peter Haggett for providing retrospective comments on this volume as one of the conductors of the revolution. The term ‘maestro’ aptly applies to Peter and his many contributions, since the seminal work by Chorley and Haggett (1967) in Models in Geography was a catalyst and confirmation that models should be an integral part in the evolution of geographical knowledge. One of the features of the revolution was a massive borrowing of ideas from other disciplines under the umbrella of General Systems Theory. This enabled information on the properties of one real system to be used to present the properties of another little known system (Von Bertalanffy 1962). By the end of the 1970s, this reasoning by analogy had reached its zenith. Yet, the theoretical evolution of geographical models during the revolution was a haphazard but creative affair. One only has to look at the evolution of consumer spatial behaviour (Fig. 1) to appreciate the complex cross-fertilisation of different ideas that underpins this concept. This process was intellectually stimulating and good to be part of. However, in many ways, although the models were creative, they were naive in their appreciation of