GaWC Research Bulletin 69

This Research Bulletin has been published in R Johnston and M Williams (eds)(2003) A Century of British Geography Oxford: Oxford University Press, pp. 347-68 (British Academy centenary monographs).

Please refer to the published version when quoting the paper.

(Z)

Global, National and Local (in a Century of British Geography)

P.J. Taylor

This chapter describes how geographical scale has been handled in British geography over the last century. With geography meaning, literally, ‘writing about the Earth’ it would seem that the matter of scale should be largely settled: geographers are the ‘global scientists’. In practice this has not been the case and, in fact, the global has been a relatively neglected scale of geographical analysis. Most geographers have spent most of their time ‘writing about their country’ and those from Britain have been no exception. In this chapter I explore how and why this privileging of the national scale has operated in a century of British geography.

As a topic in its own right, scale has attracted a minute amount of the total effort of British geographers but, and it is a ‘big but’, choice of scale is implicit in everything they teach and research. Geographical scale is something you cannot ‘get out of’, all studies take place at a given scale. Thus, far from having too little to review, the potential literature for reference in this chapter is massive, literally all of British geography. In fact, although remaining broad in scope, I do place limitations on the subject-matter considered here. I am concerned with scale in just human geography and focus largely on selected works through the twentieth century that have had a disproportionate effect on British geography. The result is a story of economic, political and urban geographies weaving their way through global, national and local scales.

I take a traditional chronological approach to my story telling. Put simply, British geographers began with a focus upon global and national scales, recovered a local scale which led on to a focus on the national and the local, and currently the emphasis is on the global and the local. In order to understand these alternate combinations of scale preferences it is necessary to begin with a brief discussion of geographical scale as a social construct.

THE NATURES OF GEOGRAPHICAL SCALE

At first glance, geographical scale is not a particularly complex concept. I was able to use it in the preamble above without defining it. This implies a straightforward concept, one that readers will easily understand: different geographical scales mean different sizes in the areal extent of study. But, of course, there are dangers in this simplicity. Certainly I do not want to imply that there is anything ‘natural’ about the geographical scales that geographers employ in their researches. Two questions are raised here as a way of problematising the concept. First, why are there three scales specified by the title of this chapter – why not 4 or more? Second and in any case, have these scales kept their meaning over the last century. For instance, is the idea of ‘global’ in 2000 the same in it was in 1900?

Social Construction: State as Pivot

Geographical scale is a social construction. It may seem to ‘just appear’ in our studies as a ‘given’ but behind every choice of scale there is a bundle of social processes that have created it. In our modern world scale seems so simple because its construction is, to a large degree, the outcome of a single institution, the state. The rise of the nation-state and its dominance over our collective geographical imaginations is directly reflected in the three scales of our title. The state defines the pivotal scale (‘national’), with a scale above (previously called ‘international’, today ‘global’ is preferred) and a scale below (‘intra-national’, commonly referred to today as ‘local’). The social sciences have been built upon this largely unexamined spatial order of scale: the core concerns of political science (national government), sociology (national society) and economics (national economy) are at the pivot scale (the use of the adjective ‘national’ is not necessary, replace it simply with ‘the’ and the scale meaning remains in all three cases) with international (political) relations and local politics, international comparative sociology and community sociology, and international economics and regional economics, all poor relations within their respective disciplines. Elsewhere I have referred to this specific ordering of knowledge as the result of ‘embedded statism’ (Taylor, 1996).

Geographical scale is, therefore, a largely hidden dimension within contemporary social knowledge, a testimony of the power of the state to define our modern world. Human geography is a social science, of course, but we would expect geographical scale to be rather less taken for granted in this discipline. And this is indeed the case. The state has been an important player in geographical discourses but it has not always been treated as the pivotal scale because of geographers’ traditional use of the regional concept to frame their studies. Regions as areal units of study are immensely flexible. The study region may indeed be defined by a state’s boundaries but this is not necessarily so. Regions can be large (continental scale) or small (neighbourhood scale) and can cover all areal extents in between. Other social sciences use regions, of course, but here they are less flexible and are employed to indicate studies above and below the state (e.g regional power blocs (of states) in international (political) relations and regional divisions of states in regional economics). Defining regions, with or without the use of states, has been a major preoccupation of human geographers and this has led to a rich tradition of researches at a variety of scales. However, over the last century, human geographers have come to conform more and more to social science norms resulting in a severe erosion of this distinctive tradition.

Social Meaning: Inconstancy and Nexuses

Today it is the global scale that dominates most social thinking. Globalization is a buzzword of our times and all the social sciences are adapting their concepts in efforts to come to terms with this ‘new’ scale of operations. Human geography is no exception. The irony is, however, that geographers have a strong global tradition that has had little or no impact on current thinking. For instance, a century ago Mackinder (1904, 22) was describing processes with a ‘worldwide scope’ in which ‘[e]very explosion of social forces … will be sharply re-echoed from the far side of the globe’. In effect, human geography has studied ‘two globals’, first as imperial/political geography then, and second as corporate/economic geography now. The key point is that geographical scales do not ‘stand still,’ as it were, while political, social and economic processes continue to operate unabated. This follows from the identification of geographical scale as a social construct, an outcome of those very processes that are forever changing. In other words the meaning of geographical scales is not a constant.

These changes in meaning are perhaps most obvious for the global scale but it is equally the case for other scales. When the local was imported into British geography from the French regional school, it was very much a rural idea of place, today localities are usually defined in urban terms. As for the ‘national’ scale, states change their boundaries, states divide, states coalesce, new states form, the nature of states change – clearly the meaning of a scale based on this institution has perforce been dynamic. And all this, like the global, is reflected in changing human geographies and the scales they study.

Finally, it is important not to see different geographical scales as somehow separate from one another. Scales may be treated separately for analytical purposes but in reality scales exist together – when we are local we are simultaneously national and global. There may be processes we study that have more salience at one scale, say the global, but it is impossible for ‘global processes’ to operate only globally, they will be ‘grounded’ at particular local places and be subject to many specific mixes of local cultures and national jurisdictions. The processes that make up distinctive geographical scales operate together. This is where the simplicity of scales gives way to complexities. Below I use the idea of a causal nexus to indicate inter-relations between different scales whose processes are so intertwined that analytical separation is quite counter-productive. Focusing on three scales produces three possible nexuses – global-national, national-local and global-local – and I will argue that each nexus has had its period when it has dominated British geography in the last century.

A GLOBAL-NATIONAL NEXUS: IMPERIAL GEOGRAPHIES

For most of the modern era geography has been closely intertwined with exploration. Geography was largely the reporting of the practices of explorers. Since the exploration was by Europeans of non-European lands, Stoddart (1982) has dubbed geography ‘the European science’. But this exploration leading to ‘discovering’ new lands is not a simple encounter whereby Europeans brought ‘new worlds’ into contact with Europe (Taylor, 1993). There were very clear power relations in these encounters that led to European domination of the world, consolidated in the late nineteenth century by the ‘new imperialism’.

After 1900 exploration and geography separated. The former had always been a very practical pursuit but with no new ‘exploitable’ lands to discover, exploration was reduced to its ‘heroic’ dimension: new ‘adventures’ to the poles and to the tops of high mountains. In contrast geography became consolidated as an academic discipline taught in schools and universities. Basically, geography took over the practical dimension from its exploration forebear and this meant, rather than being a European science, it was to be, more specifically, an imperial science. Thus this first geography discipline was perforce global in its scale of enquiry. The global and the practical came together explicitly in Herbertson’s (1905, 1910) work on ‘natural regions’: he devised environmental land divisions of the world as regions which he subsequently suggested as areal units for imperial planning by European states.

The new geography of the universities included two particular ‘subdisciplines’ that encapsulated the global-national nexus at the heart of the discipline: commercial geography (Chisholm, 1922) and political geography (Mackinder, 1904). Both were concerned with the power and wealth of states within their territories and with how worldwide relations between states, both economic and political, affected levels of power and wealth. I will focus briefly on the work of two leading British geographers in the early twentieth century, Halford Mackinder and C. B. Fawcett.

Halford Mackinder (1904, 1919) is well known for his political geography and in particular for his ‘heartland theory’. Basically this provided a framework for British foreign policy as it moved from imperial rivalries outside Europe to concern for those rivalries being expressed within Europe. Mackinder warned of the dangers on Britain’s doorstep of continental land alliances that could overthrow her world leadership based as it was on long-standing naval supremacy. This was a global-national nexus expressed directly as statecraft. But Mackinder was more than a political geographer. Much less well known are his writings on commercial geography where he again took a global view. In this work his concern is for the ‘great trade routes’ that define Britain’s economic position within the international economy (Mackinder, 1900). Of course, Mackinder did not see these as two distinct global-national nexuses. Trade policy (tariff reform) was central to the politics of his era and, indeed, his ideas on this led to him changing political parties. These two topics are brought together when he explicitly highlights the national scale in his text Britain and the British Seas (Mackinder, 1907). Most contemporary geographers would find the reference to the ‘British seas’ a little odd but for Mackinder this was crucial to his argument: the seas were the conduit of trade and the basis of defence. After a physical and regional description of Britain the book concludes with chapters on ‘strategic geography’, ‘economic geography’ and ‘imperial Britain’ – this is a regional geography that does not neglect relations with the rest of the world.

C. B. Fawcett has been less celebrated in geography but we find a very similar global-national nexus in his work. His Provinces of England (Fawcett, 1960) has the most longevity among his writings and in this he offered a geographer’s regional division of the country as a new basis for government administration – part of the ‘home rule all round’ reaction to Irish nationalism. This ‘domestic’ statecraft is complemented by his magisterial geography of the British Empire (Fawcett, 1933). As well as physical and economic surveys of the constituent territorial components of the empire, he emphasises the worldwide connections with chapters on ‘Seaways and Airways’ and ‘World Relations of the British Empire’. However, it is in the organization of the text that the global-national nexus is most clearly shown: there are two chapters each on the major territories (Britain, Canada, Australia, South Africa, India), one considering ‘internal’ matters, the other ‘external’. Topics such as Britain’s position in the ‘Ocean Gate of Europe’ and Canada’s domination by the USA take this text way beyond a territorial ‘survey’ of the empire; it is a truly global geography text for a British audience.

RECOVERING THE LOCAL: A NATIONAL-LOCAL NEXUS

Fieldwork is one of the few vestiges of exploration to be found within the discipline of geography. Merging with a parallel influence promoting ‘social surveys’ from Patrick Geddes’ (1915) civics and planning movement, fieldwork and field courses have long since been raised to the level of a fetish in geographical teaching compared to other social studies in schools and universities (see Wooldridge and East, 1951, chapter 9). Fieldwork is inevitably local in scale and therefore the potential for local study to become an important element in the development of British geography was there from the beginning. Initially overshadowed by the global, as imperial-national concerns gradually gave way to national and local planning concerns in the middle of the twentieth century, a new scale nexus emerged linking the national with the local.

This was expressed in three different developments in British geography: regional geographies of the country, contributions to national and local planning, and the development of the new sub-discipline of urban geography.

Regional Texts on Britain: Geographical Contributions to the Theory and Practice of Britain

The mid-twentieth century was the heyday of regional geography, now most definitely defined at the national and sub-national scales. The key text for this new focus is Ogilvie’s (1928) Great Britain: Essays in Regional Geography which was published for the International Geographical Union’s conference in Britain in 1928. Going through many editions, this set the style for a particular synthetic ideal of regionalism in which the social was described as the product of the physical. Chapters are devoted to a total of 23 regions wherein each begins with a region’s physical geography before describing its human geography. The physical basis of this exercise in regionalisation is illustrated by the identification of regions such as ‘The London Basin’ and ‘The Fenlands’. This formula for geographical study was applied to undergraduate dissertations for several decades as students were obliged to ‘pick their region’ and then proceed ‘from geology to industry via agriculture’ in a hopeless search for ‘regional synthesis’. The strength of this form of regional geography is shown by the imitation of Ogilvie’s book some 34 years later in Great Britain: Geographical Essays (Mitchell, 1962). This text has chapters devoted to 27 regions all organised with the physical geography first and although ‘London Basin’ has now become ‘London Region’, the Fenlands keeps its status as a region with its own chapter.

The idea of a synthesis between land and people is, of course, at the heart of all nationalist projects. Nations need to be integrated with their homelands. Thus, the scale implications of this regional geography are plain to see: Britain consists of a collection of physical units that together constitute the land of Britain. The nexus between the local units and the nation-state is to be found in the presentation of the sovereign territory as a coterie of countrysides. Geographical scale is being used to promote a British national image that privileges the rural over the urban in a new theory of Britishness. This is the discipline of geography’s contribution to the broader intellectual and political twentieth century project that was eliminating Britain’s industrial heritage from ‘real Britain’ and, in particular, from ‘real England’ (Colls and Dodds, 1986; Taylor, 2000).

Synthetic regional geography did not go unchallenged in this period. By the time the International Geographical Union returned to Britain in 1964 the text composed for the event overtly proclaimed a different type of geography: The British Isles: A Systematic Geography (Watson and Sissons, 1964). This followed an alternative ‘one-scale approach’ to the geography of Britain, pioneered by Stamp and Beaver (1932), in which different themes were studied across the whole country. This approach proved to be of much more practical utility than the synthetic studies.

Geography for National and Local Planning

The 1930s depression followed by World War II stimulated programmes of national reconstruction in which British geographers were much to the fore. Planning was conceived at three nested scales: national economic (demand) planning (Keynesianism), regional economic (supply) planning, and local government (physical) planning. While the former was the domain of economists in the Treasury, geographers made important contributions to the latter two levels of planning and, in particular, showed how the levels fitted into the larger whole.

The Land Utilization Survey initiated by Dudley Stamp in 1930 remains the largest national exercise in applied geography in Britain. Harnessing a huge team of schoolchildren, the land use of every field across the whole of Britain was surveyed in the early 1930s. The idea was to produce a basic database for British agriculture at the time of its nadir. From the scale of the individual field, information was aggregated to three higher scales and published as: local map sheets (1 inch to a mile) covering all of England and Wales, and Scotland except for the less populated Highlands; 92 county reports; and a national volume The Land of Britain: its Use and Misuse (Stamp, 1950). This research became very important for wartime food planning during the German blockade and especially for post-war planning. In 1941 Stamp was made Vice-Chairman of the Committee on Land Utilization in Rural Areas whose report in 1942 contributed to the Town and Country Planning Act (1947), the Agriculture Act (1947) and the National Parks Act (1949). Note that this most urbanised of countries did not organise city or urban planning, rather it is ‘town and country’ that is to planned. In reality this was in essence what Peter Hall and his colleagues (1973) have termed, in the title of their classic critique, The Containment of Urban England. Town and country planning was a scale-strategy to save all of the British countryside from urban people through the employment of ‘green belts’ around cities and towns. In this way the power of rural Britain continued to impact directly on urban dwellers resulting in large numbers of them being herded into high-rise flats in the 1960s. This is a particularly British nation-local nexus creating a great British planning disaster.

Geographers, of course, did not ignore industrial and urban Britain. In 1938 the Royal Geographical Society provided evidence to the Royal Commission on the Geographical Location of the Industrial Population in the form of numerous maps prepared by E. G. R. Taylor (1938). The idea of industry being concentrated in an ‘axial belt’ from Lancashire through the Midlands to London was included in the Commission’s report and excited much controversy in British geography. Following Mackinder, Baker and Gilbert (1944) suggested there was no single ‘belt’ rather there were two separate regions, Metropolitan England and Industrial England. Clearly the latter critique is more about thinking in terms of processes rather than simply reading patterns off maps but it is also a matter of scale. Any spatial scale can be divided into component parts, informed choice requires understanding the salience of different scales for a given purpose. Methods to facilitate such decision making involve much more than traditional mapping and did not become available in geographical research until after the quantitative revolution as described later below.

Urban Geography: New Contents, New Scales

In the mid-century classical statement of the nature of geography by two leading British geographers of the time (Wooldridge and East, 1951), there are chapters on physical geography, historical geography, economic geography, political geography and regional geography but not on urban geography. In fact this sub-discipline not only fails to have a chapter devoted to it, it does not even feature in the book’s index. As we have seen, what geographical interest there is in the local was largely rural, the coterie of landscapes that required protection from the people. For this land-obsessed discipline, any democratic expression of urban interests is, as one geographer put it, ‘happily … quite unthinkable’ (Freeman, 1958, 56).

Despite this infertile disciplinary landscape, to use a perhaps apt metaphor, an urban geography did emerge as a strong tradition of British geographical research which had its roots in the mid-century. Not surprisingly, the stimulus to this development came from the discussion around the industrial population distribution map which was soon realised to actually be a map of towns and cities. Thus to interpret the distribution map required understanding cities and how they relate one to another. The person who appreciated this was Arthur Smailes who produced two classic papers at the time describing the urban hierarchy of England and Wales which led him to describe the ‘axial belt’ as simply an ‘urban mesh’ (Smailes, 1944. 1946). With Dickenson (1947) who brought German studies of urban regions to the notice of British geographers, Smailes set the framework of a new ‘local’ form of study involving two scales: internal (urban land use regions) and eternal (urban service fields). This organisation carried through into quantitative geography in which a vibrant urban geography was able to find models from other social science to fit the internal/external division of urban knowledge.

THE NEW QUANTITATIVE GEOGRAPHY AND ITS SCALE PROBLEM

Regional geography severely declined in the decades after mid-century. Quite simply, treating the description of regions as an end in itself gradually lost its intellectual appeal as the core of the discipline. Led by economic geographers who had long used regions as a tool of explanation rather than as the goal of their researches, and abetted by urban geographers whose emphasis on functional regions had a similar motive, regional geography as a synthesis of multiple factors in homogenous areas was attacked as an intellectually sterile pursuit. Synthetic regional geography largely disappeared in the 1960s to be replaced by a ‘new geography’ based upon an altogether different paradigm (Haggett and Chorley, 1967). Thus this was not just a victory of a more systematic geography, additionally geography portrayed itself as self-consciously ‘scientific’ in contrast to the ‘art’ of regional geography. Generally going under the name of the ‘quantitative revolution’, this new geography can be interpreted as a ‘modernization’ of the discipline to bring it into line with the other social sciences. For many years the odd one out in social studies, human geography was transformed from an old fashioned synthesizing discipline into a specialised social science using the techniques and methods of the latter to inform spatial patterns and mechanisms. In the process, a new rigour was brought to bear on the question of geographical scale.

This transformation of geography was largely accomplished by geographers based in the USA and came to the notice of most British human geographers only after the ‘revolution’ was deemed to have occurred (Burton, 1963). Nevertheless British geographers were instrumental in consolidating the new geography in the mid-1960s with six influential texts: Statistical Methods for Geographers (Gregory, 1963), Frontiers in Geographical Teaching (Chorley and Haggett, 1965), Locational Analysis in Human Geography (Haggett, 1965), Models in Geography (Chorley and Haggett, 1967), Quantitative Geography (Cole and King, 1968) and Explanation in Geography (Harvey, 1969). Not only do these books mark a watershed in British geography, they also show the beginnings of distinctive British offerings moving from reviews of largely American studies to fresh research contributions. Perhaps the clearest case of the latter is the way Peter Haggett brought questions of geographical scale to the heart of the new geography. With Haggett (1965a, 1965b, Haggett and Chorley, 1967b) geographical scale became an explicit concern of quantitative researchers as ‘scale problems’ (Haggett, 1965a).

The Statistics of Geographical Scale

The basic problem of using statistical techniques in spatial analyses is that there is no ‘geographical individual’ to act as the basic object of study (Chapman, 1977). Measurements are collected for a given set of areal units, analysis is executed, and results obtained, but interpretation cannot simply end there. This is because the results of all such spatial analyses are scale-dependent. For instance, correlations between two variables for small census units will be different from the correlation for the same variables calculated using city regions as areal units. In other social sciences this is the ‘aggregation problem’ which is somewhat less severe because there are individuals – decision-making social agents - that constitute the key level of analysis, whereas in spatial analysis the key level of analysis – geographical scale – varies with the purpose of the research. Thus transferring rigorous quantitative analyses from mainstream social sciences to human geography provided interesting intellectual puzzles for the new geographers.

Haggett (1965a) identified three scale problems. First, the problem of information available at different scales required a means of standardization as solution. Haggett and his colleagues suggested a new measure of areal extent, the G-scale, based upon logarithmic units of the Earth’s surface (Haggett et al., 1965). This seemingly ‘natural’ measure allowed direct comparison between the scales of any sets of geographical information. Second, the areal coverage problem was about how geographers could adequately cover their domain, the land surface of the Earth. The solution required sampling which led directly to the third problem, the scale linkage problem. The latter was the crucial issue: how to relate results at one scale of analysis to other scales. For instance, interpreting local fieldwork as a sample of a larger spatial unit (a region or country) is problematic if we expect different results to obtain at different scales. The solution involved one of the most innovative elements of quantitative geography whereby scale was incorporated into the analysis as the crucial variable. In such ‘scale component analysis’ the operation of different processes can be identified at different geographical scales (Haggett, 1965b, 265-9).

Notice that the very empirical nature of quantitative geography tended to reinforce the local scale of analysis. The emphasis on data collection strengthened the fieldwork tradition in geography (Board, 1965; Collins, 1965). Further, quantitative methods teaching was inserted into the geography curriculum in the ‘practical’ slot thus supplementing or even superseding map work. However, the whole matter of the scales of study in quantitative geography reveals more about the new geography than first meets the eye.

Scale Revealing the Critical Limitations of the New Geography

The analytic treatment of scale in the new geography had severe critical limitations. In becoming a social science, human geographers took on more than simply a new bundle of techniques and models, they engrossed a privileged scale of study – the national. The social sciences as creations and creatures of the state divided up all human activities into three policy spheres - economic, social and political - and therefore human geography was reorganised into the necessary trilogy: economic geography, social geography and political geography. However, along with the new models available from economics, sociology and political science, there came the unexamined geographical assumption in the form of the embedded statism of the social sciences (Taylor, 1996). As we have seen, human geography was certainly very familiar with operating at this scale as in statecraft and national planning but this was different: regions and their scale flexibility was replaced by an inflexible and critically unquestioned national scale. This may have produced a better social science but it impoverished geography as a distinctive discipline.

In Table 1 all the human geography information contained in tables and diagrams are classified by scale for four of the key quantitative geography texts mentioned above. (Gregory (1963) and Harvey (1969) are omitted from the table because of the paucity of such data in these texts that are essentially about statistics and methodologies.) The results are as expected. The new need for data creates two outcomes. First, the local scale is well represented as a result of data collection in the field. Second, where such collection is not undertaken, there is reliance on the state to provide for data needs so that the national and sub-national scales are well represented. The neglected scale in this process is the international scale and, in particular, the global. In fact the latter rarely appears in these texts, nearly all the evidence at this scale concerns international comparisons at far less than a global scale.

The irony in all of this is that for all the claims to revolutionary credentials and the enhanced concern for scale, the new quantitative geography focussed upon local and national scales as did the geography it was replacing. However the treatment of these two scales differed somewhat. Whereas we have described a national-local nexus for the previous geography this would be an inappropriate characterisation for the new geography. The analytic bent of the latter led to a separation of scales as in the scale component analysis that measures distinguishing scale effects. The key point is that the linkage between scales being sought in the new geography is not a connection but rather a division of variance.

Reinforcing the National Scale: Development Without a National-Global Nexus

The transplant of embedded statism into human geography as social science is best exemplified in studies of economic development. Since this development is supposed be about lessening material inequalities at a global scale we might expect that here we could find evidence for new post-imperial studies at this scale. In fact this is not the case. Importing the social science approach to development led to a development geography that, far from re-introducing the global, actually reinforced the national scale of analysis in geography.

In the social sciences the process of development, like its social twin modernization, was conceptualised as a national project. Each country was considered to be a separate laboratory for development policies that will project the country forward. The forward march was assumed to be the same for all countries who were thus all marching along parallel paths to development (Taylor, 1989). The only difference between countries is that they are located at different points along the same journey. Therefore any differences in the material wealth is merely a matter of timing, eventually, given the correct policies, all will reach the goal of, as the most famous model described it, ‘high mass consumption’. There are many critical points that can be made about this model but what is noteworthy here is the elimination of the global scale. This just does not exist for development social science, there are only countries to develop.

Keeble (1967) in Models in Geography provided the main introduction of this social science into British geography. His argument is set up by the usual dismissal of previous geographical studies of development in favour of model-building (pp. 243-6). Given the new geography interest in scale, Keeble does introduce geographical scale into his consideration of economic growth models. Initially he provides a table with three scales which he describes thus: National (areas are state units); Sub-national (areas are regions of state units) and Supra-national (areas are continents, or even whole world). The use of ‘even’ implies the rarity of use of this scale; in fact in his table Keeble provides examples of models for the two lesser scales but not for the supra-national, whether global or not. After discussion of Rostow’s (1960) famous ‘non-spatial’ stages of growth model with its promise of high mass consumption for all countries, Keeble organises his text using the three scales above but has by far the least to say about the supra-national using just one main example, the pattern of economic differences in western Europe. Symbolically the map of Europe is portrayed as 18 separate country maps (Keeble, 1967, 272). In the same volume Hamilton (1967, 399-400) draws upon the most famous ‘spatialisation’ of Rostow’s (1960) model, Taaffe et al.’s (1963) transport development model, and converts it into a national industrial development model. Whether general development or just industrialisation, the point is that geographers are studying one country at a time without recourse to anything resembling a global scale. Thus instead of a new development geography heralding a new global-national nexus, we have only more national analyses.

CHALLENGING THE NATIONAL: NEW LOCAL AND NEW GLOBAL SCALES

In the aftermath of the quantitative revolution, if there were to be a challenge to the dominance of the national scale of study in human geography the most likely challengers would seem to be urban geographers. This sub-discipline strongly consolidated its position within geography in the 1960s and 1970s and clearly offered the possibility of enhancing the local scale of analysis. This did not happen: urban geography research increasingly took on a national agenda. This trend had its origins in late 1960s US geography where ‘urban problems’ were seen as indicative of ills of society as a whole. Whether studying national urban systems for national planning or intra-urban patterns and processes for national social policies (such as in housing), rather than this marking a return to a national-local nexus there was a generalised absorption of one scale (the local) into another (the national), clearly marked by the national scale being commonly designated ‘an urban society’. This was exacerbated with the radical challenges to public policy heralded by David Harvey’s (1973) highly influential Social Justice and the City in which he turned from searching for policies to make cities more just places to raising questions about the nature of society that prevented cities becoming just places. From this point the thrust of researches increasingly turned to the internal side of urban geography to study social processes in a capitalist state. The overall outcome is reflected in a comprehensive review of urban geography at the end of this period in which the external models so popular during the quantitative period (e.g. central place theory) are conspicuous by their total absence (Bassett and Short, 1989). Cities had become objects of study as places where society’s problems were accentuated and in this sense urban geography was no longer local.

But the local did re-emerge as a powerful theme within British geography. The new ‘rise of the local’ had a very specific origin in a new neo-Marxist regionalism. At about the same time there was a new ‘rise of the global’ which had a much more multifarious origin. However both re-engagements with non-national scales emerged from British geographers trying to make sense of the worldwide economic downturn that was particularly severe in its effects on Britain in the 1970s and 1980s.

The Localities Project

The downturn in the world economy had different impacts on different parts of Britain. Thus it rekindled the long-running ‘north-south debate’ in national economic policy. This crude political characterisation of the geography of recession was interpreted in a more geographically sensitive manner in a new neo-Marxist theory of regional development. In the classic statement of this approach, Doreen Massey’s (1984) Spatial Divisions of Labour, there is a return to the idea, banished by the quantitative revolution, that all places are unique (p. 117). However, unlike the physically unique sub-regions of traditional regional geography, Massey defined regional uniqueness in terms of the social inheritance of places. Every location has had a different experience of past periods of influential social change so that it brings to the contemporary situation its own particular historical mix of social and economic attributes and relations. In the famous geology metaphor of this theory, every region has different layers of investment reflecting its past successes and failures within the national and world economy. Thus. the national economy is made up of a coterie of local regional economies each differentially affected by the world economic downturn.

In 1984 the Economic and Social Research Council set up a research programme based upon this new regional theory. Entitled the Changing Urban and Regional System (CURS), the research focused upon seven study areas, 3 from the south and 4 from the north and midlands. The aim was to understand how national and global economic restructuring was impacting in different ways in these different places. The results are reported in Cooke’s (1989) Localities in which the policy issues are discussed, particular the social and political contexts, both local and national. The basic question asked was in what circumstances and how could local economic initiatives be helpful? With the local firmly back on the geographer’s research agenda this new regionalism has become a standard approach for relating different scales in British human geography. Integrating the localities research with other similar work on places but which focus more on cultural and political processes, Johnston (1991) has been able to encompass a wide range of topics in this highly scale-sensitive manner.

Putting the ‘Geo’ Back into Geography

The economic downturn in the world economy put the global back on the agenda in British geography. Self-evidently not a local or national problem, to understand social change clearly required a return to consideration of supra-national processes. Hence there were soon calls for putting the ‘geo’ back into geography in which British geographers were prominent (Johnston, 1985; Thrift, 1985, Taylor, 1989). In practice there were two main avenues through which the global made its reappearance: economic geography and political geography.

As we have seen, the economic restructuring in the localities research included an important global dimension but there were more explicit treatments of the global in 1980s economic geography. In particular three books stand out. First, Taylor and Thrift (1982) brought together a series of essays on multinational corporations. These large firms were re-writing the rulebook on industrial location theory by investing in manufacturing beyond their home countries, and, more importantly, beyond the ‘developed world’ itself. This creation of a ‘new international division of labour’ was obviously just as significant for policy in the ‘developed’ countries as in the ‘developing world’. Second, Grigg (1985) produced an extensive overview of the world food situation, its production and maldistribution. Third, the first edition of Peter Dicken's (1986) Global Shift appeared which was to become the standard reference for the economic geography of the world economy going through new editions to the present day. Quite simply, by the end of the 1980s it was not possible for students to study economic geography without incorporation of a major global dimension.

In contrast, and as we have seen above, political geography had a strong global tradition going back to the imperial influence on geography. Although, this interest waned somewhat after World War II due to association with German geopolitics, Mackinder and his heartland theory remained a staple of the fare provided to political geography students. Thus it features in Muir’s (1975) textbook which began the tradition of organising political geography into the three scales, local, national and international. Explicitly followed by Short (1982), political geography became the most scale conscious sub-discipline of geography. The weakness of this work was to be found in the deficit of linkages between the three scales. A solution was provided by bringing a world-systems approach to bear on a materialist political geography (Taylor, 1982, 1985). Political geography proved to be the main vehicle in which this most global of social theories was transferred in geography. By the end of the 1980s the global dimension was once again integrated into political geography studies in both teaching and research.

It was not just in economic and political geography that the global scale became important again. One set of essays published at this time (Johnston and Taylor, 1986) brought the global scale to human geography in general with a title that proclaimed A World in Crisis. This included essays on demographic and cultural issues as well as economic and political subjects. But the key point was that, in hindsight, we can see that the roots had been established for British human geographers to make important contributions to the study of globalization which came to dominate much social science research in the 1990s.

TODAY'S GLOBAL-LOCAL NEXUS

Globalization is a remarkable concept. Although the processes it encompasses have been variously traced to decades before 1990, the idea only really took off as a way of seeing the world in the 1990s. By the end of the decade social scientists were publishing on globalization at the rate of 33 items per week (Taylor, et al., 2002)! Why this massive interest compared with what was still a minority social science concern in the 1980s? Two changes in world geography seem pertinent. First, the end of the Cold War in 1989-91 meant the division into First, Second, and Third Worlds called out for rethinking and therefore revision. Second, the rise of Japan and consequent economic success of Pacific Asia meant that old assumption that the core of the world economy was a western European/northern American monopoly also needed rethinking and therefore revision. Both rethinkings could and did lead to revisions suggesting a more integrated ‘one world’. Globalization is the term that expresses both political and economic ‘new worlds’ and became the heavily favoured term to describe the bundle of many processes that promote trans-national outcomes. The term is, of course, not neutral in the way it is employed and its contentiousness is part of its fascination. I am not concerned with the politics of the concept here but rather I deal with how it has been researched in British human geography relative to other geographical scales.

It seems to me that the exploration of linkages between geographical scales is one of the prime contributions that geographers can make to globalization debates. This is why I have used the phase ‘global-local nexus’ to describe this section. As a nexus this implies that geographers are doing more than recognising processes at different scales, they are investigating the causal mechanisms through which processes at different scales reinforce each other. The fact that it is the local that is commonly paired with the global in such thinking indicates the way globalization implies an erosion of the power of the state: key processes are moving to above and below the national scale. It is, of course, the only scale nexus identified in this story that does not include the national. In practice national-level processes are not as neglected as this formulation might suggest – they are sometimes included in the ‘local’ category in opposition to the global - as we will see below.

Glocalization: Connections Across Scales

Since geographers were not ‘late’ discoverers of the global, it follows that they had transcended elementary contemplation/description of this new focus in social science by the time globalization had become such a popular topic. Relations between the global and other scales were widely discussed in the literature but mostly the detailed research itself did not go beyond a chosen scale. This was largely the case with the localities project but in a follow-up study Towards Global Localization, Cooke et al. (1991) do address a series of key questions at different scales. Analysis of the communication and computing industries in different countries shows how innovation in firms create enabling technologies that impact on corporate strategies producing industrial shifts. It is this single bundle of processes that is simultaneously producing new local and global scales of activity. Subsequently, one of Cooke’s co-authors, Swyngedouw (1992) coined the term ‘glocalization’ to describe a rescaling of economic activity as part of corporate strategies of flexible accumulation. This term emphasises that the contemporary politics of scale is not a singular scale effect, as globalization implies, but operates precisely through constructing different scales of activity.

This multi-scalar approach to understanding globalization is best illustrated in the popular Open University texts Geographical Worlds (Allen and Massey, 1995) and A Shrinking World? (Allen and Hamnett, 1995). In the former the chapter on ‘local worlds’ is explicitly about how various places are linked together by ‘global connections’ that are continuingly reconstituting local worlds (Meegan, 1995). More generally the books describe processes producing a very uneven globalization, one made up of many different local regions continually being remade by processes that are simultaneously local (e.g. inheritance), national (e.g. regulation) and global (e.g. corporate). The very same inter-scalar arguments can be found in another popular text that appeared at the same time, Geographies of Global Change (Johnston et al. 1995). In this collection of geographical essays it is the range of processes in which the global scale is implicated that is emphasised: geoeconomic, geopolitical, geosocial, geocultural and geoenvironmental changes are all charted. Global change may be the focus but in every case it is multi-scalar processes that are identified.

The most famous attempt to provide a framework for linking geographical scales is Amin and Thrift’s (1992) conceptualisation of ‘nodes’ of economic activity operating through ‘global networks’ of flows. They were referring generally to concentrations of firms within ‘neo-Marshallian’ districts. Although such districts are traditionally ‘industrial’ in nature one of the most explicit examples of concentrations of firms operating through global networks is to be found in the service sector. The activities of advanced producer service firms located within world cities constitute the classic example of a global-local nexus within contemporary globalization.

World Cities as a Local-Global Nexus

In contemporary research world cities are usually defined as major urban concentrations of economic activities that have worldwide connections. In the words of Peter Hall (1966, 7) in his pioneering book on the subject, world cities are where ‘a disproportionate share of the world’s most important business is conducted’. From this starting point, a world city literature has grown which is both very international and very interdisciplinary. Nevertheless, British geographers have continued to make prominent contributions, not least Hall himself (e.g. Hall, 2000). With London commonly identified as a premier world city, British geographers, notably Nigel Thrift (1987, 94), have had an important role in developing ideas about how world cities work.

The key point made in research on world cities is that no city can be studied in isolation. This reverses the ‘internalist turn’ in urban geography reported above. Cities operate within networks of cities and therefore understanding any city is impossible without knowledge of its geographical connections. World cities, themselves, are a special case in which those links are intensely global. But this renewed interest in the external relations of cities has not meant a consequent neglect of relations internal to the city. World cities are global-local nexuses in which the local intensive concentration of economic activity creates global economic connections and the global reach provides for more local economic opportunities in a causal spiral of mutually reinforcing scale effects. Again, we can use Open University texts to illustrate how contemporary British geography is developing these new ideas on geographical scale. In Massey et al.’s (1999) City Worlds the two key essays are on ‘worlds within cities’ (Allen, 1999) and ‘cities in the world’ (Massey, 1999) which between them provide us with an image of networks within cities within a network of cities. This argument is comprehensively backed up by the essays in a companion volume, Unsettling Cities (Allen et al., 1999), which cover a wide range of economic, political, social and cultural processes that link the local with the global. In a somewhat narrower vein, Taylor (2001) has specified a world city network as an interlocking network where global service firms, through their worldwide office locations, ‘interlock’ cities in the servicing of global capital. This work is part of a global internet project, the Globalization and World Cities (GaWC) Study Group and Network, an international and interdisciplinary centre for world city research but very much a product of British geography and thus with a direct interest in globalization as rescaling.

Globalization with its reliance on worldwide instantaneous communication has sometimes been interpreted as heralding the ‘end of geography’. The diminishing importance of distance in some social and economic activities does not foreclose geography, it is an integral part of the creation of new geographies. And at the heart of these new geographies there comes an enhanced importance for geographical scale. For much of the last century questions of scale have been hidden by the embedded statism of the social sciences. However, from a position where the question had been settled at the national scale, with globalization, the concept of scale has been freed to imbue the whole of the social sciences. Whereas it was mainly human geographers who incorporated scale relations in their social analyses, today relations between the global, national and local are crucial issues in all social sciences. Within contemporary British geography ‘global, national and local’ compose a critical agenda for making sense of the twenty first century.

REFERENCES

ALLEN, J. 1999: Worlds within cities. In Massey, D., Allen, J and Pile, S. (eds) City Worlds (London, Routledge), 53-98.

ALLEN, J. and MASSEY, D. (eds.) 1995: Geographical Worlds (Oxford, Oxford University Press).

ALLEN, J. and HAMNETT, C. (eds.) 1995: A Shrinking World? (Oxford, Oxford University Press).

ALLEN, J., MASSEY, D. and PRYKE, M. 1999: Unsettling Cities (London, Routledge).

AMIN, A. and THRIFT, N. 1992: Neo-Marshallian nodes in global networks. International Journal of Urban and Regional Research16, 571-87.

BAKER, J. N. L. and GILBERT, E. W. 1944: The doctrine of an axial belt of industry in England. Geographical Journal 103-4, 49-73.

BASSETT, K. and SHORT, J. 1989: Development and diversity in urban geography. In Gregory, D. and Walford, R. (eds.) Horizons in Human Geography (London, Macmillan), 175-93.

BOARD, C. 1965: Fieldwork in geography, with particular emphasis on the role of land-use survey. In Chorley, R. J. and Haggett, P. (eds) Frontiers in Geographical Teaching (London, Methuen), 186-214.

BURTON, I. 1963: The quantitative revolution and theoretical geography. Canadian Geographer 7, 151-62.

CHAPMAN, G. P. 1977: Human and Environmental SystemsL a Geographers’s Appraisal (London, Academic Press).

CHISHOLM, G. G. 1922 [1889]: Handbook of Commercial Geography. (London, Longman, Green and Co,).

CHORLEY, R. J. and HAGGETT, P. (eds.) 1965: Frontiers in Geographical Teaching (London, Methuen).

CHORLEY, R. J. and HAGGETT, P. (eds.) 1967: Models in Geography (London, Methuen).

COLE, J. P. and KING, C. A. M. 1968: Quantitative Geography (London, Wiley).

COLLINS, M. P. 1965: Field work in urban areas. In Chorley, R. J. and Haggett, P. (eds.) Frontiers in Geographical Teaching (London, Methuen), 215-238.

COLLS, R. and DODDS, P. (eds.) 1986: Englishness: Politics and Culture 1880-1920 (London, Croom Helm).

COOKE, P. (ed.) 1989: Localities (London, Unwin Hyman)

COOKE, P., MOULAERT, F., SWYNGEDOUW, E., WEINSTEIN, O. and WELLS, P. 1991: Towards Global Localization (London, UCL Press).

DICKEN, P. 1986: Global Shift: Industrial Change in a Turbulent World. (London, Paul Chapman).

DICKENSON, R. E. 1947: City, Region and Regionalism (London, Routledge & Kegan Paul).

FAWCETT, C. B. 1933: A Political Geography of the British Empire (London, University of London Press).

FAWCETT, C. B. 1960 [1919]: The Provinces of England (London, Hutchinson).

FREEMAN, T. W. 1958: Geography and Planning (Londo, Hutchinson).

GEDDES, P. 1915: Cities in Evolution (London, Williams and Norgate).

GREGORY, S. 1963: Statistical Methods and the Geographer (London, Longmans)

GRIGG, D. 1985: The World Food Problem (Oxford, Blackwell).

HAGGETT, P. 1965a: Scale components in geographical problems. In Chorley, R. J. and Haggett, P. (eds.) Frontiers in Geographical Teaching (London, Methuen), 164-85

HAGGETT, P. 1965b: Locational Analysis in Human Geography (London, Arnold).

HAGGETT, P. and CHORLEY, R. J. 1967: Models, paradigms and the new geography. In Chorley, R. J. and Haggett, P. (eds) Models in Geography (London, Methuen), 19-41.

HAGGETT, P., CHORLEY, R. G. and STODDART, D. R. 1965: Scale standards in geographical research: a new measure of areal magnitude. Nature 205, 844-7.

HALL, P 1966: The World Cities (London, Heinemann).

HALL, P., GRACEY, H., DREWETT, R. and THOMAS, R. 1973: The Containment of Urban England (London, Allen and Unwin).

HALL, P. 2000: Global city-regions in the twenty first century. In Scott, A. J. (ed.) Global City-Regions (Oxford, Oxford University Press).

HAMILTON, F. E. I. 1967: Models of industrial location. In Chorley, R. J. and Haggett, P. (eds) Models in Geography (London, Methuen), 362-424.

HARVEY, D. 1969: Explanation in Geography (London, Arnold).

HARVEY, D. 1973: Social Justice and the City (London, Arnold).

HERBERTSON, A. J. 1905: The major natural regions of the world. Geographical Journal25, 300-12.

HERBERTSON, A. J. 1910: Geography and some of its present trends. Geographical Journal 36, 468-79.

JOHNSTON, R. J. 1985: To the ends of the Earth. In Johnston, R. J. (ed.) The Future of Geography (London, Methuen), 326-38.

JOHNSTON, R. J. 1991: A Question of Place (Oxford, Blackwell).

JOHNSTON, R. J. and TAYLOR, P. J. (eds.) 1986: A World in Crisis? Geographical Perspectives (Oxford, Blackwell).

JOHNSTON, R. J., TAYLOR, P. J. and WATTS, M. J. (eds.) Geographies of Global Change (Oxford, Blackwell).

KEEBLE, D. E. 1967: Models of economic development. In Chorley, R. J. and Haggett, P. (eds) Models in Geography (London, Methuen), 243-302

MACKINDER, H. J. 1900: The great trade routes. (The connection with the organization of industry, commerce and finance.). Journal of the Institute of Bankers 21, 1-6, 137-55, 266-73.

MACKINDER, H. J. 1904: The geographical pivot of history. Geographical Journal 23, 421-42.

MACKINDER, H. J. 1907 [1902]: Britain and the British Seas (Oxford, Clarendon).

MACKINDER, H. J. 1919: Democratic Ideals and Reality (London, Constable).

MASSEY, D.1984: Spatial Divisions of Labour. Social Structures and the Geography of Production (London, Macmillan).

MASSEY, D. 1999: Cities in the world. In Massey, D., Allen, J and Pile, S. (eds) City Worlds (London, Routledge), 99-156.

MASSEY, D., ALLEN, J and PILE, S. (eds) City Worlds (London, Routledge).

MEEGAN, R. 1995: Local worlds. In Allen, J. and Massey, D. (eds.) Geographical Worlds (Oxford, Oxford University Press), 53-104

MITCHELL, J. 1962: Great Britain: Geographical Essays (Cambridge, Cambridge University Press).

MUIR, R. 1975: Modern Political Geography (London, Macmillan).

OGILVIE, A. G. 1928: Great Britain: Essays in Regional Geography (Cambridge, Cambridge University Press).

ROSTOW, W. W. 1960: Stages of Economic Growth (Cambridge, Cambridge University Press).

SHORT, R. J. 1982: An Introduction to Political Geography (London, Routledge & Kegan Paul).

SMAILES, A. E. 1944: The urban hierarchy of England and Wales. Geography 29, 41-51

SMAILES, A. E. 1946: The urban mesh of England and Wales. Transactions of the Institute of British Geographers 11, 85-101.

SMAILES, A. E. 1953: The Geography of Towns. (London, Hutchinson).

STAMP. L. D. 1950: The Land of Britain: its Use and Misuse (London, Longman).

STAMP, L. D. 1960: Applied Geography (London, Penguin).

STAMP, L. D. and BEAVER, S. H. 1933: The British Isles : A Geographic and Economic Survey (London, Longman).

STODDART, D. R. 1982: Geography – a European science. Geography 67, 289-96.

SWYNGEDOUW, E. 1992: The Mammon quest: ‘glocalisation, interspatial competition and the monetary order: the construction of new scales’. In Dunford, M. and Kafkalas, K. (eds.) Cities and Regions in the New Europe (London, Belhaven Press), 39-68.

TAAFFE, E. J., MORRILL. R. and GOULD, P. R. 1963: Transport expansion in underdeveloped countries: a comparative analysis. Geographical Review 53, 503-29.

TAYLOR, E. C. R. 1938: Discussion on the geographical distribution of industry. Geographical Journal 92, 22-39.

TAYLOR, M. J. and THRIFT, N. (eds.) 1982: The Geography of Multinationals (London, Croom Helm).

TAYLOR, P. J. 1982: A materialist framework for political geography. Transactions of the Institute of British Geographers NS7, 15-34.

TAYLOR, P. J. 1985: Political Geography: World-Economy, Nation-State, Locality (London, Longman).

TAYLOR, P. J. 1989: The error of developmentalism in human geography. In Gregory, D. and Walford, R. (eds.) Horizons in Human Geography (London, Macmillan) 303-19.

TAYLOR, P. J. 1993: Full circle, or new meaning for the global? In Johnston, R. J. (ed.) The Challenge for Geography. A Changing World: a Changing Discipline. (Oxford, Blackwell), 181-97.

TAYLOR, P. J. 1996: Embedded statism and the social sciences: opening up to new spaces. Environment and Planning A 28, 1917-28.

TAYLOR, P. J. 2000: Which Britain? Which England? Which North? In Morley, D. and Robins, K. (eds) British Cultural Studies (Oxford, Oxford University Press), 127-44.

TAYLOR, P. J. 2001: Specification of the world city network. Geographical Analysis 33, 181-94.

TAYLOR, P. J., WATTS, M. J. and JOHNSTON, R. J. (2002) Geography/globalization. In Johnston, R. J., Taylor, P. J. and Watts, M. J. (eds.) Geographies of Global Change (Oxford, Blackwell, 2^nd ed.), 1-**.

THRIFT, N. 1985: Taking the rest of the world seriously? The state of British urban and regional research in a time of economic crisis. Environment and Planning A 17, 7-24.

THRIFT, N. 1987: The fixers? The urban geography of international commercial capital. In Henderson, J. and Castells, M. (eds.) Global Restructuring and Territorial Development (Beverly Hills, Sage), 203-233.

THRIFT, N. 1994: On the social and cultural determinants of international financial centres: the case of the City of London. In Corbridge, S., Martin, R. and Thrift, N. (eds.) Money, Power and Space (Oxford, Blackwell), 327-355.

WATSON, W. W. and SISSONS, J. B. 1964: The British Isles: a Systematic Geography (London, Arnold).

Table 1: Geographical Scales in British Quantitative Geography

Text	%International	%National	%Sub-national	%Local
Chorley/Haggett 1965 (n = 24)	8.3	41.7	12.5	37.5
Haggett 1965 (n = 147)	7.5	29.9	39.5	23.1
Chorley/Haggett 1967 (n = 43)	0.4	46.5	23.3	30.2
Cole/King 1968 (n = 76)	7.9	51.3	27.6	13.2

n indicates the number of items of evidence in tables and diagrams for each book. Only human geography evidence is counted.

Edited and posted on the web on 3^rd December 2001

Note: This Research Bulletin has been published in R Johnston and M Williams (eds)(2003) A Century of British Geography Oxford: Oxford University Press, pp. 347-68 (British Academy centenary monographs)