| |
An approach to human geography focusing on the spatial arrangement of phenomena and on related flow patterns: its usual methodology is that of spatial science. The philosophy of positivism underpins the approach, which is closely linked to the discipline\'s quantitative revolution.
Work within the locational analysis framework was taken up by several groups of geographers in the US from the late 1950s on, although it had much deeper roots in the work of non-geographer pioneers (Johnston, 1997a): the main initial centres were at the University of Washington, Seattle, the University of Wisconsin, Madison, and the University of Iowa and the ideas were rapidly spread — mainly by graduates from those three centres — to other universities (such as Chicago, Northwestern, and Ohio State): developments at Seattle were significantly influenced by a visiting Swedish pioneer, Torsten Hägerstrand.
The focus of this work was on developing geography as \'the science of locations\'. At Seattle, this was strongly influenced by central place theory and other presentations of regularities in point and line patterns, and in the geography of flows (Garrison, 1959-60); Bunge (1966) generalized on this in his thesis on Theoretical geography, based on the premise that \'nearness [is] a candidate for the central problem in geography\' and defined geography as the \'discovery of predictive patterns\'. Stewart (1956), a physicist who worked at the American Geographical Society, originally a physicist, argued for what he termed social physics, which sought mathematical rules describing human behaviour and its spatial outcomes, and which differed from sociology by its \'avoidance of subjective descriptions\' (cf. macrogeography). Others, such as McCarty at Iowa, were strongly influenced by developments in the field of economics, to which they introduced the spatial variable in searches for correlations between mapped distributions (as in McCarty and Lindberg, 1966): these links led to the close interrelationships between geographers and regional scientists in the 1960s and 1970s (see regional science), and are illustrated by attempts to build economic geography theories of spatial arrangements (as in Smith, 1981).
Locational analysis was introduced to the UK by geographers who had visited the main centres of innovation in the USA (some were graduate students there). One of them, Peter Haggett, codified the paradigm in a textbook — Locational analysis in human geography (first published in 1965: see also Haggett, Cliff and Frey, 1977) — which rapidly attained the status of a classic. He established the pedigree of locational analysis by linking its geometrical focus to the work of early Greek cartographers: others argued that \'of the three classic areas of mathematics, geometry would appear to be the most promising for geography\' (see also Bunge, 1966) although Sack (1972) doubted whether geometry can be used as a language to explain as well as describe form.
Haggett\'s appeal to the relatively neglected geometrical tradition within geography was \'placed squarely on asking questions about the order, locational order, shown by the phenomena studied traditionally as human geography\'. Such a focus needed:
{img src=show_image.php?name=2022.gif }Â to adopt a systems approach which concentrates on the patterns and linkages within an assemblage; {img src=show_image.php?name=2022.gif }Â to employ models as the stimuli for understanding (as with his frequent references to Thompson\'s (1917) On growth and form); and {img src=show_image.php?name=2022.gif }Â to use quantitative procedures to make precise statements (generalizations) about locational order (see also macrogeography).His book, and its successor (Haggett et al., 1977), was in two main parts: the first dealt with models and the second with methods.
Haggett\'s major innovation was his framework for classifying models of locational order. (Interestingly, his schema is paralleled by one developed contemporaneously by Cole and King, 1968.) His initial presentation contained five components: the sixth (diffusion) was added in the second edition. Their separate chapters were ordered in what he saw as a logical sequence for the analysis of regions (see figure). The first (a) was concerned with interactions, or flows across space; the second (b) analysed the networks along which those flows moved; and the third (c) considered the major nodes, or organizational centres, on those networks. In the fourth (d) the hierarchical structure of the nodal system (see urban systems; central place theory) was decomposed, whereas in the fifth (e) the focus was the organization of the space between the nodes — the surfaces. The final component (f) looked at diffusion down the hierarchy, along the networks, and across the surfaces. Haggett decomposed systems, therefore, but — despite his discussion of regions — didn\'t recombine them into wholes.
Regarding methods, there was a major shift of emphasis between the two editions of Haggett\'s book. In the first, he accepted the then conventional view that the methods associated with the general linear model could be adopted for spatial analysis without difficulty but a decade later, influenced by his studies of diffusion and the work of his colleague Andrew Cliff on spatial autocorrelation, he argued that such applications were \'more difficult than might at first appear\' so that, for example, conventional regression procedures were omitted from the book.
{img src=show_image.php?name=bkhumgeofig38a.gif }
{img src=show_image.php?name=bkhumgeofig38b.gif }
{img src=show_image.php?name=bkhumgeofig38c.gif }
{img src=show_image.php?name=bkhumgeofig38d.gif }
{img src=show_image.php?name=bkhumgeofig38e.gif }
{img src=show_image.php?name=bkhumgeofig38f.gif }
Locational analysis Stages in the analysis of nodal regional systems: (a) interaction: (b) networks: (c) nodes: (d) hierarchies: (e) surface: (f) diffusion (Haggett et al., 1977)
From this pioneering conspectus, Haggett explored the component parts of his framework in more detail. A book on network analysis combined human and physical geography (Haggett and Chorley, 1969) but most of his work has focused on aspects of diffusion (Cliff et al., 1975, 1981; see lead-lag models), especially in the context of medical geography.
Haggett\'s major role was as a synthesizer and stimulant, and he promoted both the search for locational order as a major goal of contemporary geography and its adopted positivist philosophy, although he wrote very little himself on the latter. Others followed his lead, providing their own perspectives on the study of what became known as spatial structures (Johnston, 1973). Morrill (1970), for example, was also strongly influenced by the geometrical tradition and his text The spatial organization of society focused on the \'nearness principle\': people, he argued, seek to maximize spatial interaction at minimum cost and so bring related activities into proximity — the result is that \'human society is surprisingly alike from place to place … [because of] the predictable, organized pattern of locations and interrelations\'. Whereas Morrill concentrated on substantive illustration of his principle (which he later revised somewhat, downplaying the geometrical component: Morrill and Dormitzer, 1979), Abler, Adams and Gould (1971) spent the first third of Spatial organization discussing the nature of science and the methods to be used in locational analysis.
Critiques of locational analysis developed during the 1970s, focused initially on the validity of normative theory in geographical studies, on the grounds that this did not reflect the reality of decision-making and so was of little value in predicting locational arrangements. This stimulated growth of the more inductive approach of behavioural geography which, nevertheless, also assumed that generalizations about spatial behaviour and pattern are feasible. This position was countered in further critiques, located within humanistic geography and radical geography, which argued that locational analysis studies imply an absence of human free will whereas societies comprise individuals with the capacity to remember, learn and promote change. As a consequence, while the study of spatial patterns, the use of relevant quantitative methods and the implied search for laws of spatial organization remain part of human geography, they play a smaller role within the discipline than was the case in the 1960s and early 1970s, when they were presented as geography\'s entrée to, and niche within, the social sciences (NAS-NRC, 1965; Taaffe, 1970; Chisholm, 1971).
Some historians of human geography debate whether the development of locational analysis in the 1960s reflected the influence of persuasive leaders such as Haggett, whether it came about because of greater awareness of the existence of relevant models and methods in other disciplines, as Pooler (1977) suggests, or whether it was just a general realization that the traditional form of regional analysis was increasingly less suited to study of the contemporary world and needed to be replaced. Cox (1976) argued that until the nineteenth century societies were predominantly local in their orientation and the local physical environment was the main influence on livelihood — hence traditional regional geography and its society-land focus (what Haggett called vertical links). The twentieth-century integration of societies into a global world-economy (see world-systems analysis; globalization) means that spatial interdependence has become much more important (horizontal links) and \'locally experienced environmental dependencies lost their rationale: men [sic] relate less and less to the land on which they stand and more and more to socially-created geographical patterns over a much wider area\': recognition of this by geographers led to locational analysis ousting regional analysis as the discipline\'s dominant paradigm.
Whatever its origins, locational analysis substantially changed the nature of human geography from the mid-1960s on. It presented geography as a positivist social science, concerned to develop precise, quantitatively stated generalizations about patterns of spatial organization, thereby enriching and being enriched by location theory, and to offer models and procedures which could be used in physical planning. By 1978, Haggett could write that:
the spatial economy is more carefully defined than before, we know a little more about its organization, the way it responds to shocks, and the way some regional sections are tied into others. There now exist theoretical bridges, albeit incomplete and shaky, which span from pure, spaceless economics through to a more spatially disaggregated reality.Twelve years later, he continued to promote the search for \'scientific generalizations\' (Haggett, 1990), while accepting that in the search for spatial order \'the answer largely depends on what we are prepared to look for and what we accept as order\': only a minority of geographers now claim that order is the focus of their quest.
Changes in the nature of modern communications and the diminishing role of distance as an influence on many methods of interaction is making the formal models of locational analysis obsolescent according to some authors (Johnston, 1997b, 1997c). Against this, however, it is argued that making optimal location decisions remains crucial to the profitability of many businesses (see applied geography) and locational analysis models and methods allow an appreciation of earlier societies whose spatial order provides the foundation (such as the urban systems) on which modern societies are built.
Haggett\'s schema made no mention of \'bounded spaces\', divisions of the earth\'s surface into separate territories at a variety of scales, within which much human behaviour is constrained (cf. territoriality). In part this may have been because it proved difficult to develop theories of such spaces, rather than of the point, line and surface patterns which were the focus of locational analysis, but Nystuen\'s (1963) essay on \'fundamental geographical concepts\' identified only three as both necessary and sufficient for the construction of an abstract geography — directional orientation, distance, and connectedness — although he did consider whether boundary should be included as a further \'geographical primitive\'. Space in the locational analysis paradigm was largely continuous in its nature, and boundaries were impediments to its rational use rather than central components of its structuration (see Haggett, 1981).
Locational analysis was very firmly set in the spatial science paradigm of geography that emerged in the 1960s, therefore, but it was not the only approach to understanding spatial structures and the organization of the space-economy. Others, stimulated by the work of Harvey (1973, 1982) and Massey (1984), adopted a political-economy approach, whose goal of accounting for changing geographies of capitalist organization eschewed any search for spatial regularities and instead focused on the contingent influence of place-based factors on the general processes of uneven development (cf. layers of investment). This latter approach now dominates, and the search for \'spatial order\' as set out in Haggett\'s pioneering conspectus is much less important in contemporary geography: locational analysis as spatial science marked an important turning-point in the discipline\'s history — according to some it was crucial to its continued presence in the academy and its acceptance within the institutional structure of the social sciences in North America and the UK (it was adopted much later, and in a much more muted form, in the French- and German-speaking realms), and it continues to be promoted as such by some (Johnston, 1998) — but it now barely rates a mention in contemporary presentations of the discipline\'s contents (e.g. Peet, 1998). (RJJ)
References Abler, R.F., Adams, J.S. and Gould, P.R. 1971: Spatial organization: the geographer\'s view of the world. Englewood Cliffs, NJ: Prentice-Hall. Bunge, W. 1966: Theoretical geography. Lund: C.W.K. Gleerup. Chisholm, M. 1971: Research in human geography. London: Social Science Research Council. Chorley, R.J. and Haggett, P., eds, 1967: Models in geography. London: Methuen; Cliff, A.D. et al. 1975: Elements of spatial structure: a quantitative approach. Cambridge and New York: Cambridge University Press. Cliff, A.D. et al. 1981: Spatial diffusion. Cambridge and New York: Cambridge University Press. Cole, J.P. and King, C.A.M. 1968: Quantitative geography. London: John Wiley. Cox, K.R. 1976: American geography: social science emergent. Social Science Quarterly. 57: 182-207. Garrison, W.L. 1959-60: The spatial structure of the economy I, II and III. Annals of the Association of American Geographers 49: 238-48 and 471-82, and 50: 357-73. Haggett, P. 1965: Locational analysis in human geography. London: Edward Arnold. Haggett, P., Cliff, A.D. and Frey, A.E. 1977: Locational analysis in human geography, 2nd edn. London: Edward Arnold. Haggett, P. 1981: The edges of space. In R.J. Bennett, ed., European progress in spatial analysis. London: Pion. Haggett, P. 1978: The spatial economy. American Behavioral Scientist 22: 151-67. Haggett, P. 1990: The geographer\'s art. Oxford and Cambridge, MA: Basil Blackwell. Haggett, P. and Chorley, R.J. 1969: Network analysis in geography. London: Edward Arnold. Harvey, D. 1969: Explanation in geography. London: Edward Arnold. Harvey, D. 1973: Social justice and the city. London: Edward Arnold. Harvey, D. 1982: The limits to capital. Oxford: Blackwell. Johnston, R.J. 1973: Spatial structures: introducing the study of spatial systems in human geography. London: Methuen. Johnston, R.J. 1997a: Geography and geographers: Anglo-American human geography since 1945, 5th edn. London: Edward Arnold. Johnston, R.J. 1997b: W(h)ither spatial science and spatial analysis. Futures 29: 323-35. Johnston, R.J. 1997c: Geography in a restructuring world. GeoJournal 42: 9-16. Johnston, R.J. 1998: Turning full circle: American geography and the social sciences, 1950-2000. In H.H. McCarty and J.B. Lindberg,A preface to economic geography. Englewood Cliffs, NJ: Prentice-Hall. Massey,D.1984: Spatial divisions of labour. London: Macmillan. Morrill, R.L. 1970: The spatial organization of society. Belmont, CA: Wadsworth. Morrill, R.L. and Dormitzer, J. 1979: The spatial order: an introduction to modern geography. North Scituate, NJ: Duxbury; NAS-NRC 1965: The Science of Geography.Washington: National Academy of Sciences-National Research Council. Nystuen, J.D. 1963: Identification of some fundamental spatial concepts. Proceedings of the Michigan Academy of Science, Arts and Letters 48: 373-84. Peet, R. 1998: Modern geographical thought. Oxford: Blackwell; Pooler, J.A. 1977: The origins of the spatial tradition in geography: an interpretation. Ontario Geography 11: 56-83. Sack, R.D. 1972: Geography, geometry and explanation. Annals of the Association of American Geographers 62: 61-78. Smith, D.M. 1981: Industrial location: an economic geographical analysis, 2nd edn. New York: John Wiley. Stewart, J.Q. 1956: The development of social physics. American Journal of Physics 18: 99-123. Taaffe, E.J., ed., 1970: Geography. Englewood Cliffs, NJ: Prentice-Hall; Thompson, W. d\'Arcy 1917: On growth and form. Cambridge: University of Cambridge Press. |
|
