Spaces, places and the nature of knowledge

Two decades ago, the study of social geography was confined largely to urban areas, covering topics such as the residential segregation of social class and ethnic groups. Today, class is much reduced in importance to social geographers and numerous other topics have been added to their portfolio.

Introducing Social Geographies defines social geography as the study of the production of social relations, identities and inequalities, of the role of space in that production, and of the resulting spatial variation. Its three parts deal with: "Society, material life and geography", covering work, class and social life, leisure, and communities; "Power, identity and social geography", concerned with race and ethnicity, gender and sexuality, age, generation and life course, and disability; and "Social geography and social problems", with chapters on nature and landscape, housing, crime and inequality, and poverty.

Five themes dominate, according to the authors: individuals are part of societies; space is an important means in the organisation of societies; the local and the everyday are significant; social relations and identities are power relations; and social geography is political, with policy-making roles. Their stance is clearly stated as Marxist and feminist-inspired; positivist approaches are largely eschewed, with humanistic and qualitative methods presented as "most enlightening".

The book is generally much longer on the Marxist and feminist-inspired "theory" and society-wide context than on the spatial arguments. In some chapters, such as that on housing, the latter hardly make an appearance, and elsewhere space is underplayed. Furthermore, some subjects where space is important in social relations and identities - the geographies of health and education, for example - are omitted. Nor is there a coherent approach to the spatial factor - on the neighbourhood effect, for example, on which so much is being written now. Perhaps that is all too positivist?

If one book underplays the spatial factor, however, People and Place brings it to centre-stage. Lewis Holloway and Phil Hubbard are not concerned with sub-disciplinary definitional issues but instead focus coherently on the importance of space and place in the conduct of everyday lives. They cover topics such as "sense of place", "imagining places", "representing place", "place and power" and "struggles for place". And they, too, eschew positivist approaches, not because they are "bad" per se - they throw valuable light on many topics - but because they cannot capture the nature of places and the importance of space in the everyday.

Within their clearly defined agenda, Holloway and Hubbard's is an excellent introductory text. It is well crafted and well written and uses its illustrative material superbly. It shows the great advantage of well-conceived co-authored texts over multi-authored collections. Students are carefully guided through a range of challenging ideas and well-worked examples, and are led to consider the moral underpinnings of much of what social geographers now study. The authors do not feel the need to preach, and have not: they convince by the quality of their arguments, not rhetoric.

Holloway and Hubbard end where Keith Hoggart, Loretta Lees and Anna Davies begin - asking "how do you undertake research on such topics?". Each group is clear that epistemological and ontological questions must be addressed first. Until you are clear on the nature of the knowledge you wish to produce, and of the world it is to be produced about, you cannot address methodological questions, let alone methods.

The first chapter of Researching Human Geography discusses these issues clearly, as does that on "research design" - what decisions have to be made once the general research direction has been determined.

Five chapters follow, dealing with different methodologies and sources: published data, archives and other documents, social surveys, interviews and focus groups, and lived experience as a research process. They are not discussions of the how-to-do-it variety but rather general advice on the advantages of, and problems with each approach. Having read Researching Human Geography , students will not be ready to set out on their chosen subject - details on methods are still to come. But they will have linked the philosophy of human geography with its practice in more detail than has been the case in the past.

Ron Johnston is professor of geography, University of Bristol.