A special session of the
A special session of the 2017 Annual International Conference of the Royal Geographical Society in London
This year marks the centenary of the first edition of D’Arcy Wentworth Thompson’s book On Growth and Form. Thompson was a mathematically inclined biologist who spent the whole of his academic life in Dundee and St Andrews but who also served for a short while as President of the Royal Scottish Geographical Society (1942–46). So why should we mark the centenary of his work? On Growth and Form was a major influence on the development of the spatial analysis tradition in geography during the 1960s and 1970s, most notably in the work of Haggett and Bunge and in inspiring Tobler to develop cartogram projections. Despite this early work, most subsequent quantitative geography has concentrated on the analysis of the attributes of places, treating their geometry as in some sense fixed. However, aided by advances in geocomputation and greater data availability, recent years have seen a re-discovery of the importance of a geometrical tradition in geography associated with ideas that date back to 1917 to do with the relationship between geometric form and process, similitude and rates of growth, and transformation/projection as an analytical device. This session will examine recent work in the Thompson tradition and along the way celebrate the centenary of a remarkable and stimulating book.
Introduction to On Growth and Form
David Unwin (Birkbeck, University of London, UK)
“Together with Charles Darwin, Thompson is the most culturally influential English-speaking biologist in history” (New Scientist, 11 Feb 2017) but we have only one book, On Growth and Form, first published in 1917, to sustain that assertion. This short talk will introduce the session by tracing the introduction of Thompson’s ideas into the geography of the 1960s and now, fifty years on, into what has been called ‘the new quantitative revolution’.
Shaping geographic enquiry – D’Arcy Thompson’s On Growth and Form
Alan Werritty (University of Dundee, UK)
D’Arcy Thompson’s seminal monograph On Growth and Form (1917) is widely acknowledged by biologists as the first successful biophysical explanation of the size and shape of organisms. In explaining how the shape of plants and animals is determined, Thompson gave mathematical rigour to the concept that forms follows function. His two fundamental ideas – the concept of allometric growth and the theory of transformations – not only radically influenced the development of biology but also stimulated cutting edge advances in fine art, urban design and engineering throughout the 20th century. Thompson’s ideas also played a significant role in the development of spatial analysis in geography during the late 1950s and 1960s, most notably in influential monographs by Haggett and Bunge and Tobler’s liberation of map projections from the constraints of Euclidean geometry. A century after the publication of On Growth and Form Thompson’s ideas continue to inform geographical scholarship in Batty’s studies on the size, shape and scale of cities, Dorling’s World Mapper project and fractal-based analyses of the morphology of river basins. This paper re-assesses Thompson’s contribution to geographic enquiry from the 1950s through to the present time.
Beyond physical form: geocomputational approaches to map projections
Chris Brunsdon (University of Liverpool, UK)
Thompson’s ‘theory of transformations’, articulated in On Growth and Form, has inspired the development of many innovative map projections, of which the cartogram is the best known. A cartogram is a map in which a thematic variable (such as population size, or total land value) is substituted for physical area or distance. The technique can be a powerful tool, in particular when the thematic variable is population-based and used as a proxy for physical area. In this case, other variables can be used as standard thematic variables – but if they are based on characteristics of the population then the choice of a cartogram as a base map ensures that the characteristics of highly populated smaller physical areas remain visible. To those familiar with more usual map
projections of a region, use of other thematic variables to define the cartogram can provide a striking visual contrast, emphasizing the geographical distribution of the variable in use. Thus, cartograms represent an approach to spatial representation that has a geometry, but not the usual geometry of points in a Euclidean plane or on the surface of a sphere. A number of algorithmic approaches to cartograms have been attempted – starting with physical models, and progressing to advanced computer algorithms. This talk will provide an overview of cartograms, including examples of their use and a brief history of approaches to creating them. Finally the `getcartR’ package in the programming language R will be described. This package, written by the speaker, provides a freely available and flexible front end to the Gastner and Newman cartogram algorithm, allowing users to experiment with the cartogram methodology, and providing reproducible, publication-quality spatial data visualizations.
Evolution and Transformation in Cities: Re-evaluating Patrick Geddes and D’Arcy Wentworth
Michael Batty (University College London, UK)
Despite being members of a small community of biologists at University College Dundee in the last decade of the 19th century, Patrick Geddes and D’Arcy Wentworth Thompson did not refer to each others work as far as I can tell. Geddes published the pioneering book Cities in Evolution in 1915 and despite some reservations, preached the message that cities should be observed and planned as evolutionary systems. Wentworth Thompson published his seminal book On Growth and Form two years later in 1917 and to an extent provided the foundations for mathematical biology but also established the notion that physical and natural objects, cities included, evolved to different shapes as they grew. They must have known each other rather well and it is unlikely that they would not have discussed cities and biology at least in passing. I will not solve this riddle in this essay – probably nobody can – and it is possible that the coincidence of their working in the same department for so many years did not make any sense until much later as both Geddes and Wentworth Thompson have come to play such a central role in complexity theory, particularly in the complexity theory of cities. I will attempt to show how early ideas in the theory of evolution resonated in both their works and then illustrate how their work was picked up in the 1960s, in particular by the design theorists such as Christopher Alexander, by those studying geometric form in spatial and city systems such as Waldo Tobler, and by those beginning to fashion a theory of cities from the bottom up such as Jane Jacobs. I will attempt to pull all these threads together especially in contemporary ideas about the morphology of cities such as that which focusses on scaling, self-similarity, and connectivity
(Thursday) 9:00 am - 10:40 am GMT
Royal Geographical Society
Kensington Gore, Kensington, London SW7 2AR
Royal Geographical Society