Ron Johnston on Emrys Jones's A Social Geography of Belfast .

While reading for my undergraduate dissertation on a railway town I came across Howard Bracey's Social Provision in Rural Wiltshire, which stimulated my MA research. It seemed I was going to be a rural geographer. But my first academic post was a teaching fellowship at Monash University, and I became an urban geographer instead.

Geographers had done very little work on the Melbourne area by then and it was suggested that I study it for a PhD. Urban geography was a new subfield, and I explored the urban sociology and ecology literature for ideas. Geography was also beginning its quantitative and theoretical revolution, and I was seduced by social area analysis and factorial ecology. Small-area census data were available for Melbourne, and I learned to use computers.

The data analysis on its own was insufficient to my current conception of urban social geography, however. I wanted to understand the place as well, and to appreciate the links between social processes and spatial form. A fascinating seminal volume provided the needed stimulus, and underpinned the structure of my thesis.

ÁñÁ«ÊÓƵ

Emrys Jones's A Social Geography of Belfast was a path-breaking book in a whole variety of ways. Published in 1960, it predated most geographers' forays into statistics, with a number of technical innovations. It introduced the classic Burgess and Hoyt models of urban social morphology, subjecting them to a telling critique. Very importantly for me, it postulated a direct link between social regions and urban landscape regions, which was graphically illustrated by the developments linked to country residences on Belfast's Malone Ridge.

Revisiting the volume after more than a decade, I have immediately recaptured the excitement of a seminal work, which lay outside the received wisdom of geographical analysis. Many later developments (in the study of intra-urban migration, for example) were foreshadowed. Most importantly, alongside the focus on general processes operating in a large modern city there was an appreciation of the particular features of a very individual city - and insights into what was to happen there a decade later.

ÁñÁ«ÊÓƵ

The history of the discipline of human geography (and no doubt many others) identifies many pioneering scholars the impact of whose work is less than their quality deserves. Emrys Jones is one of the them. Most of the first generation of urban geographers (like myself) working in the 1960s and 1970s were increasingly attached to data analysis and theory development based on statistical generalisation than to the nuanced appreciation of the complex mosaic that makes up a city's social geography - and when we abandoned statistics we turned to versions of Marxism instead, with their own generalising frameworks. And so I left behind studies of high-status residential areas, street patterns and building types; perhaps if I had reread Emrys's book more frequently, things might have been different - but then perhaps I would not have had the pleasure of rereading it now.

Ron Johnston is former vice chancellor of the University of Essex.

BOOKS