Historical Geography and Historical Sociology: Our Honest Toil and the Spatial Turn

Susan W Friedman. Handbook of Historical Sociology. Editor: Gerard Delanty & Engin F Isin. Sage Publication. 2003.

With the growing interest in space within sociology and a willingness among some in geography to begin to think theoretically one might expect mutual interest and convergence between the fields of historical geography and historical sociology. Historical sociologists who have experienced the ‘historical turn’ and the ‘cultural turn’ may wonder just what form the ‘spatial turn’ could take. Despite some convergence, the interaction between these fields has been limited. What historical geography and historical sociology have shared is an interest in history, but what they have taken from history has differed, reflecting their divergent disciplinary histories and aims. Here I will attempt to introduce historical geography to those in historical sociology who may be intrigued with at least their conception of what such a field could include, and then turn to questions of convergence and overlap. To limit this unwieldy topic, I will concentrate on North American and British authors and, for historical geography, on those directly associated with the label. As a historical geographer, my reflections on historical sociology will be both suggestive and tentative.

Parallel and Divergent Disciplinary Histories

At first glance, the histories of historical geography and historical sociology appear rather similar. Apparently subfields within the social sciences, they have defined themselves in part by their expressed interest in history. Following some currency and attempts at delimitation in the 1920s and 1930s (see, for example, Barnes, 1921; Becker, 1934; East 1933; Fairgrieve et al., 1921; Gilbert, 1932; Morris et al., 1932), both experienced renewed attention in the 1970s and 1980s, encouraging a flurry of conferences, the founding of flagship journals and other attempts at institutionalization. Both have strong Anglo-American biases and have felt the impact of a series of popular methods and approaches, including quantitative methods, world-system theory and feminism. However, once one begins to explore the particular relationships of historical geography and historical sociology with the field of history as well as their early attempts at self-definition, important differences emerge. Here I will concentrate on historical geography followed by some brief comparisons with the history of historical sociology.

In Britain, the early practitioners of historical geography had often been trained as historians (Darby, 1983b). They saw their objects as two-fold, both subsidiary to history: first, beginning in the 1840s, the study of changing administrative and political boundaries (for example, E.A. Freeman, 1881); and, second, emerging at the end of the century, the study of geography’s influence on history (for example, Mackinder, 1904; G.A. Smith, 1894). In both cases, the history to which geography was tied was a political history, and in fact, the field was often labelled historical and political geography (Darby, 1987: 117-18). As the prominent historical geographer Henry Clifford Darby later testified: I recall hours spent in disentangling the various connotations of the name Burgundy—the kingdoms, the duchies, the county, the imperial Kreis or ‘circle’ of later times; little wonder that Freeman wrote: ‘no name in geography has so often shifted its place and meaning,’ (1983b: 421).

Clearly there were ties between this early historical and political geography and the imperial projects of the times. Deeply involved in both establishing geography as a discipline in Britain and British politics, Halford Mackinder promoted a form of historical geography closely tied to political history, made no attempt to hide the connection between the two. In a position paper prepared for the Royal Geographical Society, he stated: ‘I believe that on lines such as I have sketched a geography may be worked out which shall satisfy at once the practical requirements of the statesman and the merchant, the theoretical requirements of the historian and the scientist, and the intellectual requirements of the teacher’ (1887: 159). This paper was used by the society, led at the time by colonial administrators and men of affairs (Stoddart, 1986: 66), to persuade Oxford University to appoint a Reader in Geography, a position that was first held by Mackinder (Kearns, 1985: 72). However, as Felix Driver (1992) has argued, the ties between the discipline of geography and imperialism were not deterministic but instead were constituted in complex and often contested ways. Although much remains to be done to uncover this complexity, Robin Butlin’s study (1995) of the historical geographies of the British Empire from 1887 to 1925 has demonstrated that only some of these works gave a laudatory and a militaristic image of empire.

By the advent of the First World War, imperialism was increasingly seen as problematic, and some began to question the wisdom of linking history and geography. For example, in a 1914 review of Mackinder’s series ‘Elementary Studies of Geography and History, P.M. Roxby, a geographer at Liverpool trained in history at Oxford, argued that the ‘combined presentation of history and geography’ could lead to ‘too determinist a view of social development.’ To illustrate the dangers of such an approach, he wrote, ‘[L]et us join furious battle with Feuerbach, who said,’ perhaps in grim anticipation of his country’s recent action, that ‘history is nothing but the operation of geographical laws…’ (1914: 407).

By the late 1920s and the early 1930s, historical geographers were eager, in Darby’s words, ‘to forge a method and to create an academic discipline’ at a time when geography still lacked a secure institutional base (1987: 131). The great postwar expansions of both secondary and university education, which enabled geography to become more established and independent (TW Freeman, 1987: 9), provided the needed opportunity. Darby, who became the first Ph.D. in geography at Cambridge, later described this period: Whereas, hitherto, some historians had believed in the relevance of geography to history, now some of the increasing numbers of geographers in Britain began to reverse the thinking, and to consider the relevance of history to ‘geography’ (1983b: 422).

At the ‘First International Congress of Historical Geography’ held in Brussels in 1930, Darby was impressed by the archivist Charles Pergameni (1931: 212), who pleaded for a redefinition of the field as the ‘human geography of the past,’ drawing on the earlier discussions of ‘past geographies’ by the German geographer Alfred Hettner. This view was in opposition to the ‘older view’ of geography’s relationship to history held by Belgian historians who organized the conference. Two years later at a joint meeting of the Geographical and Historical Associations on the question What is historical geography?, the geographers present argued that the theme of the influence of geography on history was one of geographical history, mistakenly labelled historical geography by historians, and that instead their field should be redefined as ‘the reconstruction of the geographical conditions of past times’ (Morris et al., 1932: 43)—an approach that soon became the predominant theme of British historical geography (see also East, 1933; Gilbert, 1932).

Although there were some notable exceptions, many of the academic geographers in the United States were trained in geology rather than history and often held posts of physical geography in geology departments (Dunbar, 1981; James and Martin, 1978). Among the early American publications associated with historical geography, the theme of the influence of geography on history was particularly strong, often drawing on the study of physiography. In 1903, Albert Perry Brigham, an early student of the physical geographer William Morris Davis, published Geographic Influences in American History. In the same year, Ellen Churchill Semple, who studied history at Vassar and geography with Friederich Ratzel in Germany, published American History and Its Geographic Conditions. Both Brigham and Semple used features of physical geography to help explain the course of a primarily political American history, stressing themes of natural defences, territorial expansion, the location of routes, the suitability of areas for settlement and the progression of events during battles in the French and Indian War and the Civil War.

Attacks on these works came from historians including Frederick Jackson Turner, whose own frontier thesis was not entirely free from environmental determinism (the explanation of social phenomena by the physical environment). Turner (1905), while finding much to praise in Semple’s and Brigham’s books, criticized them for relying on questionable secondary sources. Others were even more sceptical. In 1907 at a conference of the American Historical Association on ‘the relations of geography and history,’ George Burr from Cornell criticized Semple’s work, arguing that ‘geography, though a factor in history, is only a factor,’ and that man ‘too is a factor, and oftener the active than the passive,’ and George Adams argued, ‘Most of the matters which they [the geographers] rightly call upon us to include in history are conditions, not causes’ (American Historical Association, 1908)

Despite these criticisms, other North American geographers took up the theme of geographic influence, and some would go to even greater excesses, lapsing into rather blatant forms of environmental determinism often mixed with evolutionary theory. Two of the more extreme examples were Ellsworth Huntington, also a student of Davis who later studied and taught at Yale, and T Griffith Taylor, who taught successively in Australia, at Chicago, and finally became the first chair of the new Geography Department at Toronto in 1935. Both were trained as geologists and became known for advocating climatic explanations for racial and regional character types, often invoking neo-Lamarckian themes (Livingstone, 1992: 225-31). Excesses such as these soon gave North American geography a bad name at a time when its institutional basis was still not solid, often being tied to geology departments. Some of the harshest criticism came from the anthropologist Franz Boas. Initially trained in physics and geography, Boas came to reject the deterministic explanations he had been taught in favour of historical and empirical modes of research (Livingstone, 1992: 291-4).

Against this background, several prominent historical geographers emerged who are now seen as the founders of the discipline, and in all cases, they attempted to sever the links between historical geography and both political history and environmental determinism. In Britain, the dominant figure became Darby. Trained in the earlier versions of historical geography, he became the leading spokesperson in Britain for the redefinition of historical geography as the reconstruction of past geographies. In 1936, while he held both a lectureship in geography at Cambridge and a research fellowship in history at King’s College, his edited An Historical Geography of England before 1800 was published, in part due to the assistance of the economic historian J.H. Clapham, also at King’s, who helped him receive backing from Cambridge University Press. Through a ‘sequence of cross-sections taken at successive periods,’ Darby (1936) intended to demonstrate historical geography’s potential. As he later explained, ‘Our aim was not to produce some broad general views, but geographical descriptions based as far as possible on primary sources; and our hope was to match the scholarship of contemporary historians’ (1987: 124-5). The book was, in fact, favourably reviewed by historians despite its challenge to their conception of historical geography. In the English Historical Review, after expressing surprise at the lack of attention to either ‘the fixing of boundaries, whether civil or ecclesiastical’ or ‘the course of military operations,’ Sir George Clark reported ‘[historical geography, still a little uncertain of its place among the sciences, has great promise for the future. Mr. Darby and his collaborators have shown that it has already won an important place in the study of British history’ (1937: 140).

The book was, as Darby (1960) later admitted, a methodological mix. It included not only cross-sections but also discussions of developments through time, including his own work on the draining of the Fens bordering the Wash, drawn from his dissertation. Nevertheless, in it the seeds of Darby’s form of historical geography were clear: the interest in cross-sections and ways to combine them with narrative descriptions of change; the themes of active transformations of landscapes; and, most importantly, the determination to use primary sources—to meet the historians’ standard of scholarship. Darby’s later substantive publications, which exemplified this approach, included his edited A New Historical Geography of England (1973), which alternated ‘horizontal’ cross-sectional chapters with ‘vertical’ chapters on ‘geographical changes through time,’ Domesday England (1977), and numerous papers on the changing English landscape discussing woodland clearances, drainage and moorland reclamation as well as the study of place names.

In his promotion of the field, Darby made a series of methodological statements in which he stressed problems of presentation and organization, addressing, for example, ‘the problem of geographic description’ (1962). Paying particular attention to the relations between geography and history, he suggested four possible approaches: the geography behind history (corresponding to the geographical explanation of history), past geographies (cross-sections), the history behind geography (the vertical approach) and the use of history to explain features in the present landscape (1953). Conceptual issues were of less interest, making him leery of overly close relations with ‘social studies’ (Andrews, 1980: 205; Darby, 1983b: 424; Lawton and Butlin, 1989: 16; Williams, 2002). He preferred to cultivate relations with economic history promoting a careful examination of documentary sources, and ended his commentary on his writings shortly before his death with the following quotation from Clapham taken from his introduction to Darby’s fenland books in 1940: ‘He is a very imperfect economic historian who is not also a tolerable geographer; and I cannot picture to myself a useful historical geographer who has not a fair working knowledge of economic history’ (quoted in Darby, 1989: 8).

Contemporary with Darby in the United States were two American geographers, Ralph Brown and Carl Sauer. Closest in interests and approach to Darby, Brown came to geography from a background in agriculture and economics. Like Darby, he identified historical geography as ‘the geography of the past’ (1948: iii), was particularly interested in the use of primary documents, and appeared to view historical scholarship as a model (McManis, 1978). Because of his early death in 1948 and his avoidance of methodological discussion, Brown never became an active spokesperson for North American historical geography, but is still seen as ‘a major figure’ in the field (Butlin, 1993: 36). In his substantive writings, he carried his determination to rely on primary documentation to an extreme. In his first book, Mirror for Americans: Likeness of the Eastern Seaboard, 1810 (1943), he attempted to create past geographies, stressing landscape transformation and environmental perception, using only the data sources available in 1810, and in his second book, Historical Geography of the United States (1948: iii), he gave sketchy and partial coverage given his insistence on sufficient primary documentation.

Although both men reacted against the character of much of American geography, the approaches taken by Carl Sauer and Ralph Brown were very different. In common with Brown and even Darby, Sauer stressed the importance of primary research, bemoaning the tendency of those promoting geography as a synthetic field to base their work on ‘borrowed materials’ (letter to L.S. Wilson, 6 April 1848, quoted in Kenzer, 1988: 335). However, following the physiographer Rollin Salisbury, with whom he studied at Chicago, Sauer saw such primary research as closely tied to ‘physiographic studies’ and careful fieldwork (Kenzer, 1988: 335). Recoiling from the ‘environmentalist tenet’ taught by Semple, Harlan H. Barrows and others at Chicago, he, like the Pergameni paper Darby admired, turned to the writing of German geographers such as Hettner.

Once at Berkeley, where he headed the new geography department, which had recently separated from geology, Sauer published a paper on the distinctiveness of a geographical approach. Citing the anthropologist Kroeber (a former student of Boas), ‘The Morphology of Landscape’ attacked environmental determinism, depicting it as a misguided attempt to make geography part of biophyisics. His alternative vision of geography was based on the morphological study of landscape, a study which he insisted would include the ‘modification of the area by man and its appropriation to his uses.’ Here he defined historical geography ‘as the series of changes which the cultural landscapes have undergone and therefore involves the reconstruction of past cultural landscapes’ (1963a [1925]: 344), a view reinforced in 1927 when he described historical geography as the study of ‘landscapes of the historic past’ (1927: 200). Later Sauer would repudiate parts of this vision, stressing man’s agency in changing the landscape, rather than the reconstruction of a series of landscapes. (for example, Sauer, 1963b [1956], 1963c [1941]). In his own substantive work, he would come to focus on agricultural origins and dispersals in Latin America and the American Southwest and the exploration and exploitation of the area by Europeans (for example, Sauer, 1952, 1966, 1968, 1971).

In addition to reacting against the ‘environmentalist tenet,’ Sauer found himself out of step with much of the geography that replaced it in the United States. In 1940 in a presidential address to the Association of American Geographers (AAG), he protested against ‘the neglect of historical geography’ and charged that those who limited their work to the contemporary period were ‘held by a peculiar obsession’ (1963b [1956]: 366). Terming the period from Barrows’s ‘Geography as Human Ecology’ (1923) to Richard Hartshorne’s recently published methodological treatise The Nature of Geography (1939) as the ‘Great Retreat,’ he lamented the attempt both to separate human from physical geography and to limit the former to ‘a nongenetic description of the human content of areas, sometimes called chorography’ (1963c [1941]: 353). In his attempt to distinguish geography from other subjects and give it academic respectability, Hartshorne had given a very narrow definition to historical geography, limiting it to a study of cross-sections and suggesting that some of Darby’s and Sauer’s writings were really works of history or anthropology, not geography (1939: 178, 184-8).

Although professing an interest in ‘comparative regional geography,’ Sauer limited such comparisons to a particular cultural area and those adjacent to it.

Such work obviously cannot be done by sample studies ranging widely but may require a lifetime given to learning one major context of nature and culture. One may thus extend one’s learning outward to the limits of a culture area and explore the contrast on the other side of the boundary line. Or one may undertake excursions to areas characterized by important kindred qualities. But always there must be the base, the area for which the observer is making himself expert. (1963c [1941]: 362)

Accordingly, he argued, ‘there are no general laws of society, but only cultural assents.’ Instead of turning to sociology or economics, he suggested cultivating relationships with anthropology, ‘the most advanced of the social sciences’ (Sauer, 1963b [1956]: 357, 378), whose ‘culture area’ concepts he found particularly relevant. With time, he would argue that even anthropology was becoming too much of a social science, interested in theorizing and universalizing, rather than a field of culture history as he envisioned (Williams, 1987: 220).

In 1948, after Brown’s death, Sauer was approached to write the chapter on historical geography for a fiftieth-anniversary volume for the AAG, a chapter that Brown had been expected to write. Despite repeated requests, he refused, being disillusioned with American geography, and suggested a student of his, Andrew Clark, who was working, in Sauer’s words, on ‘the overseas Anglo-Saxon culture world’ (Williams, 1983: 18). Clark soon became the spokesperson for North American historical geography, fulfilling the role undertaken by Darby in Britain, in a way not done by either Sauer or Brown. Later, Clark would testify that his heart was never in his methodological pronouncements, which had been something thrust upon him due, among other things, to Brown’s death (Meinig, 1978: 24).

Clark had come to geography from mathematics and economic history. In 1930, he received a BA in mathematics and physics from McMaster, and, after working as a statistician, completed a Master’s degree in geology, economic history and geography at Toronto in 1938. There, he worked with both the economic historian Harold Innis and the geographer Griffith Taylor. Innis, who was closely associated with the geography department, was known for his ‘staple theory,’ which explored the importance of commodity production (particularly cod and furs) for the economic history of Canadian regions. He is still seen as a founder of Canadian historical geography. From both Innis and Taylor, Clark gained an appreciation for the importance of the physical environment, even though he rejected Taylor’s determinism (Ward and Solot, 1992: 14-15), and from Innis, training in the use of archival sources (Meinig, 1978: 11).

At Innis’s suggestion, Clark had transferred to Berkeley to work on his doctorate with Sauer, whose emphasis on careful fieldwork fitted well with Clark’s training under Taylor. Clark’s thesis was on the impact of European settlement on the New Zealand landscape, focusing on the introduction of plants and animals. Although a topic well suited to Sauer’s interests, Clark’s framing of it was distinctive in his emphasis on the region, a theme that he would develop in his later works. For his next monograph, Three Centuries and the Island (1959), he examined Prince Edward Island, combining cross-sections with vertical studies of change, mapping selected phenomena, ratios between them at particular times, and the changes over time. In his last monograph, Acadia: The Geography of Early Nova Scotia in 1760 (1968), Clark organized his discussion by eras and topics rather than by cross-sections, but still give a detailed picture of areal patterns. Throughout his substantive work, Clark strove to debunk easy explanations, whether they were those of environmental determinism or of cultural transfer. Instead he came to view historical geography as the investigation of the complexity of particular regions (for example, Clark, 1948; see also Meinig, 1978: 16-17).

In the AAG commemorative volume, Clark charted a course between Sauer, his former adviser, and Hartshorne, now his colleague at Wisconsin. He refused to remove time, and Sauer, from the field, and yet, like Hartshorne, emphasized a cross-sectional approach to areal differentiations. Defining historical geography as the study of ‘geographical change through time,’ he claimed that such an approach focused on processes because such cross-sections could be ‘understood as momentary states in continuing and complex processes of change.’ In contrast to history, which studied changes to human society, geography paid attention to ‘the surface of the earth, in whole or in part, and to areal associations and differentiations thereon’ (1954: 71-3, 85). Despite his characterizations, Clark has been criticized for examining the results of geographical changes rather than the processes, and so really studying changing geographies rather than geographical change (Meinig, 1978: 21-2).

In 1972, Clark contributed a chapter on ‘Historical Geography in North America’ to Progress in Historical Geography. Having been subjected to attacks by those who thought Clark represented a very restricted view of their field (Koelsch, 1970; Prince, 1971), he was now less willing to provide a clear distinction between history and geography, speaking instead of the ‘history-geography borderlands’ and arguing that ‘there are, indeed, no peculiarly “geographical” or “historical” “facts” or “factors”.’ Nevertheless, he suggested that the more recent work of Sauer and his students was not historical geography, given their ‘cultural-environmental interests’ and ‘often strong anthropological overtones,’ but instead a ‘stream of geographical culture history, parallel to but often quite distinct from the work of the more historiographically focused historical geographers.’ Although admitting that their particular approach had much to offer, he noted that ‘difficulties have arisen through lack of documentary skills or of contextual information’ when they tackled historical topics (1972: 130, 138-9).

Clark’s work and his statements of intent were a study in contrasts. Despite his pronouncements on the nature of historical geography, he is perhaps best known within the field for his insistence on high standards of scholarship, a value that is still seen as central to the discipline. Although describing his research programme as a study of British rural settlement and land use overseas (Ward and Solot, 1992: 19), his insistence on the exhaustive use of both archival and field sources, or ‘hyperempiricism,’ (Ward and Solot, 1992: 13) prevented him from making many comparisons or generalizations (for example, Clark, 1968: 371) and he came to focus on relatively small and isolated regions. He would also come to champion a humanistic vision of historical geography, and yet that vision was often obscured in his own painstaking empirical work. Admired for his scholarship, he has been attacked for his excessive ‘scholarly prudence,’ a phrase that Clark used to justify his hesitancy to make generalizations (Meinig, 1978: 16).

None of these four historical geographers established close relationships with sociology, and all were hesitant to theorize, even viewing generalizations with scepticism. Instead their primary interdisciplinary links were to history, especially economic history, in the cases of Darby, Clark and Brown, and to anthropology for Sauer. Although both Clark and Sauer expressed some interest in comparative studies, their insistence on prudent scholarship strongly limited the kinds of comparative work they found acceptable.

Unlike historical geographers with their active attempts to ‘create an academic discipline,’ historical sociologists have been hesitant to proclaim a distinctive domain, and many leading proponents continue to claim that their aim is to promote an approach from which all branches of sociology could benefit, not to establish a subdiscipline (Skocpol, 1984; Tilly, 1988). Efforts to study earlier practitioners are often directed to social thinkers of all sorts whether or not they would have adopted a label of historical sociology (for example, Skocpol, 1984; D. Smith, 1991), and those who explicitly adopted the label are often ignored.

The term ‘historical sociology’ did have some currency in the United States by the 1930s, when it was identified by the American Sociological Society as synonymous with general sociology. F.N. House identified its approach as one studying the ‘theory of social evolution and progress’ (1936: 297), an approach that was soon seen as outdated and tainted by evolutionism (Schwartz, 1987: 2). There were, however, two American sociologists who tried to redefine the term and remove its association with abandoned evolutionary theories: Harry Elmer Barnes and Howard P. Becker. Barnes (1921: 17) suggested the earlier flaws could be overcome by turning to the ‘historico analytical method, introduced by Boas and his disciples,’ and Becker (1934: 26) suggested the same end could be best achieved by turning to ‘Max Weber’s ideal-typical method.’ Nevertheless, both men were outside the mainstream of American sociology, and their efforts to revive the term were unsuccessful.

By this time, American sociology had become very distinct from history, and the Chicago school of sociology was becoming very influential. Albion Small, who started the sociology department at Chicago and had a background in history, later recalled a joint 1903 meeting of the American Economic Association and the American Historical Association. At that meeting, Franklin Giddings, who was soon to publish his Descriptive and Historical Sociology (1906), presented a paper entitled, A Theory of Social ‘Causation’ (1904), which was harshly criticized by the historians, including George Burr, who labelled it as a form of philosophy of history. In Small’s words, ‘In brief, as the sociologists understood it, the historians virtually declared that they had no use for the conception of social science, and especially not for the conception of history as science.’ Even more appalling to Small was a comment attributed to Professor Emerton of Harvard: ‘It is not even essential that what the historian writes down shall be true, provided it lends itself to dramatically interesting treatment.’ According to Small, that remark ‘marked the decisive parting of the ways between American historians and sociologists,’ and at the time of his writing their relations had still not recovered (1923: 45).

In Britain, sociology was very slow to gain an institutional foothold, and for many years, there was only one professor of sociology and so little room to promote a particular kind of sociology. This chair at the London School of Economics (LSE) was held first by Leonard Hobhouse from 1907 until his death in 1929, and then by his student Morris Ginsberg. Hobhouse was interested in theories of social development, without associating development with progress, and led a study that assembled a large collection of reports on 643 pre-literate peoples, using records from anthropologists, travellers and missionaries. Ginsberg worked more directly with historical materials and saw the sociologists’ role as ‘the discovery of general laws’ drawing on sources from history, anthropology and elsewhere (Banks, 1989). Although neither identified himself as a historical sociologist, both Barnes and Becker would point to them as precursors. Following the Second World War, there was an influx of students into sociology at the LSE and some interest in the use of historical materials but also a focus on contemporary issues. Little real growth in historical sociology came in Britain until the 1960s, when the sociologists at Leicester, many of whom had come from history, started to have an impact (Banks, 1989: 527, 532-3).

In contrast to the founding historical geographers, those first associated with historical sociology saw history as a source of materials but had little serious interest in the historical method. Their interest was in the big picture, and their goal to systematize history starting with unilinear evolutionary approaches and later with modifications of them to create more acceptable theories of social development. It was, in the words of Piotr Sztompka, ‘sociology above history’ (1986: 335), as compared with the ‘sociology without history’ that had come to dominate in the United States.

Renewals during the 1960s, 1970s, and 1980s

When both Clark and Sauer died in 1975, their many students were well established, as were those of Darby, who continued an active presence until his death in 1992. It has been the students of Clark and Darby who have identified themselves with historical geography and so have dominated it, whereas Sauer’s students have often chosen the label of cultural geography Clark’s success as a promoter of historical geography made him an easy target during the 1960s and early 1970s, when historical geography came under attack from those both within and outside the field. In a review of Clark’s Acadia in the journal Economic Geography, Willliam Koelsch spoke of an ‘intellectual crisis’ in the field due to its neglect of ‘those conceptual frameworks, models, and techniques which have distinguished geography in the 1960s’ (1970: 203), and in the same journal, Martyn Bowden, noting the ‘rift’ between geography and historical geography charged that historical geography had become ‘the stronghold of a group of regional geographers whose prime objective is seemingly to make historical geography respectable in the eyes of historians’ (1970: 203).

Much of geography had indeed changed during the preceding decade as more and more geographers became intrigued with the use of quantitative techniques and more analytical approaches, including those of environmental perception. In part, Livingstone has argued (1992: 324), this quantification served to reframe geography as an objective science following the turbulent McCarthy era. The initial impetus for the quantitative changes began in economic geography during the 1950s but by the 1960s had become more widespread. At first, it was simply a matter of applying statistical techniques developed in other fields, but by the late 1960s, geographers were beginning to rework many of these techniques to take into account the particular problems encountered when applying them to geographical distributions, and some began to talk of reformulating the field of geography into a spatial science.

Clark, with his background in mathematics, did not dismiss quantitative approaches outright, but was worried that the ‘current fad’ would restrict information to ‘very limited categories,’ leading to an inadequate understanding of the regional context (1972: 137). More critically, Cole Harris one of Clark’s many former graduate students, questioned the logical foundations of a ‘theoretical geography of spatial relations’ given that all phenomena exist in both space and time. Even more broadly, he challenged the concept of theoretical geography, which suggested that ‘all explanation must be either deductive and nomological or probabilistic.’ As an alternative approach, he proposed that historical geography concentrate on geographical synthesis built around concepts such as region, landscape and place. Such a synthesis would pay more attention to both ‘the actions of ordinary men’ and the physical land than would typically be the case in historical synthesis and would not abandon the field’s tradition of scholarship (Harris, 1971: 157-8, 160).

Some, however, pushed for great change. Hugh Prince (1968: 110) advocated ‘fresh approaches,’ including the study of past perceptions and the use of models, and three years later complained, ‘Knowledge and respect for material sources of evidence has, for a few zealous practitioners, turned into a narrow devotion to archives as the final repositories of truth’ (1971: 22-3). Showing a similar disdain for the ‘orthodox doctrines,’ Alan Baker argued that ‘methodologically the main advances can be expected from an increased awareness of developments in other disciplines, from a greater use of statistical methods, from the development, application and testing of theory, and from exploitation of behavioural approaches and sources’ (1972: 13). The ‘other disciplines’ from which he hoped to gain so much included economic and social history and anthropology, but he appeared to expect less from sociology.

Despite a remark by Baker that historical geography should become less of a distinctive subfield and more of an approach within the ‘branches of systematic geography’ (1972: 28), this period was a time of institutionalization for historical geography marked by the founding of separate journals and the development of conferences. What was to become the Historical Geography Research Group (HGRG) associated with the Institute of British Geographers (IBG) started in 1967 and by 1975 claimed status as one of the largest research groups of the IBG with over 200 members. At about the same time, a loosely organized group known as the Eastern Historical Geography Association (EHGA) began meeting as well, drawing together historical geographers from both Canada and the United States. Following two joint meetings between the HGRG and a group of Canadian historical geographers (1975 and 1977), this Canadian and British group was expanded to include geographers from the United States, Australia and New Zealand, and given the acronym CUKANZUS, a group which would meet in Los Angeles (1979), Toronto (1981) and Oxford (1983). In 1986 at Baton Rouge, the conference was renamed the International Conference of Historical Geographers (ICHG) to encourage even wider participation. This rapid growth in organizations and meetings caused Hugh Prince to wonder, following the Toronto CUKANZUS in 1981, ‘Do historical geographers spend too much time and effort organizing and attending conferences?’ (1982: 64).

Journals specifically devoted to historical geography were also started. In 1971, the Historical Geography Group associated with the AAG launched the Historical Geography Newsletter, co-edited by Martyn Bowden, whose critical review helped to start the debates within historical geography. Initially circulated free of charge, early issues were dominated by reports on meetings, thesis abstracts, research objectives and short book commentaries, but the hope was to include ‘extended bibliographic contributions, inventory essays, critical appraisals of primary sources, and items similar to those now characteristic to the Historical Methods Newsletter’ (Bowden and Vicero, 1972a: 2). Articles guidelines suggested a focus on specific research materials, methodology, the teaching of historical geography and literature reviews on a particular theme or region (Bowden and Vicero, 1972b). The circulation increased rapidly. More articles were included, and soon a subscription charge was imposed.

The success of the Newsletter soon led to plans for a more ambitious journal, the Journal of Historical Geography. Announced as early as 1973 in the pages of the Historical Geography Newsletter (‘Journal of…,’ 1973: 47), publication began in 1975 with an English base and both an English editor, John Patten, and a North American one, Andrew Clark. The journal aimed to aid ‘recruitment’ to historical geography through an ‘increase of courses, seminars and research activities in historical geography within universities and colleges all over the world’ (Clark and Patten, 1975). Despite this goal, British and North American historical geographers dominated the journal, which included substantial articles and a large book review section along with notes on conferences and other announcements. In 1978, the Historical Geography Newsletter, which had been growing as well, was reorganized to encourage lengthier submissions and retitled Historical Geography. Predominantly a North American journal, one of its mandates is to present the work of ‘emerging scholars.’ Both journals aimed to encourage interdisciplinary dialogue, the Journal of Historical Geography welcoming contributions from ‘allied subjects’ and Historical Geography promoting the field’s contributions to the ‘geographical, historical and social scientific disciplines.’

The institutionalization of historical sociology occurred later and was less explicitly oriented toward the definition of a subdiscipline. Accordingly, the growth of the field has often been charted by the publication of particular books and articles (for example, D. Smith, 1982), rather than by the organization of specific societies or journals. This lack of enthusiasm for disciplinary definition is evident in Charles Tilly’s 1988 statement:

Fearfully, I predict the institutionalization of historical sociology: fixing of a labeled specialty in sections of learned societies, journals, courses, a share of the job market. I fear these likely outcomes for two reasons: first, because the ‘field’ lacks intellectual unity and, by its very nature, will forever lack it; second, because institutionalization may well impede the spread of historical thinking to other parts of sociology. The other parts need that thinking badly. (1988: 709)

Another indication of this ambivalence has been the unwillingness of many to separate sociology from history, so, rather than arguments explaining the difference between geographical history and historical geography, one finds statements such as Abram’s hope ‘to integrate history and sociology as a single unified programme of analysis’ (1982: xviii). Two other defining books in historical sociology, Theda Skocpol’s edited collection (1984) and Dennis Smith’s The Rise of Historical Sociology (1991), focus on key figures from the past and include some such as E.P Thompson and Marc Bloch who probably would have been uncomfortable with the label. What tends not to be included in these disciplinary accounts is discussion of the earlier evolutionists who had also used the label. That history is forgotten as sociologists discover history with untainted eyes, marking a ‘new departure’ rather than reviving the past (Tilly, 1988: 706).

Despite this uneasiness with disciplinary definition, some of the feared institutionalization did occur. In the mid-1970s, there was an ‘explosion of historical sociology Ph.D.s’ in the United States (Abbott, 1991: 219), and in Britain, the Leicester sociology department, with its historical slant, was overtaking London as the major source of sociology teachers in British universities (Banks, 1989: 533). Norbert Elias and Ilya Neustadt, both European refugees from the start of the Second World War, led the department at Leicester and promoted the comparative study of societies, stressing their variety and changes over time and drawing on anthropology and history. A number of historians were recruited, and, according to Banks, this may have contributed to the more detailed historical studies of particular areas over short periods of time that came from the department as well. Nevertheless, few of the sociology students at Leicester from the late 1960s on engaged in historical research, preferring instead the techniques of participant observation and interviewing.

In 1976, both the British Journal of Sociology and Social Forces, published by the University of North Carolina Press, produced special issues devoted to historical sociology. That year also marked the start of three social history journals that were to become influential: Social Science History, History Workshop and Social History. The Sorokin prize was awarded to sociologists with a historical orientation in both 1976 and 1977, an event that Victoria Bonnell suggested marked its achievement of ‘full status’ within the field of sociology (1980: 157). For Theda Skocpol (1984: ix-xiii), a crucial event appears to be the Conference on Methods of Historical Social Analysis held at Harvard in 1979. By the early 1980s, according to Dennis Smith (1991: 3), articles with a ‘historical dimension’ accounted for nearly a quarter of those published in the main sociology journals.

There was not a major journal specifically devoted to the field until the Journal of Historical Sociology began in 1988. Historical sociologists during the interim period often published their works either in the older Comparative Studies in Society and History, begun in 1958, Theory and Society, begun in 1974, or one of the new social history journals. Not unlike the Journal of Historical Geography, the Journal of Historical Sociology sought to promote ‘openness, exploration, and diversity’ and not to define and therefore confine the field (Corrigan and Sayer, 1988: 3). Explicitly interdisciplinary in orientation, the editors pointed not only to history but also geography and anthropology as fields with shared interests.

When Daniel Chirot introduced the special issue of Social Forces on ‘The Uses of History in Sociological Inquiry’ in 1976, he argued that one could not yet identity particular schools of historical sociology or even ‘recognizable clusters of positions’ (1976: 232). By 1991 Abbott (1991) identified groupings associated with participation in the Section on Comparative Historical Sociology, one of the largest sections of the American Sociological Association by the early 1980s (D. Smith, 1991: 3), and the Social Science History Association (SSHA). These organizations, he argued, have served two very different groups of historical sociologists. The first was started by the Weberians’ but after 1983 shifted focus to the macro-political sociology of the nation-state, with an emphasis on theory and a sceptical attitude towards quantification. The SSHA, by contrast, was strongly associated with quantitative history often at a micro-scale and attracted those sociologists interested in working more directly with the historians often on past social groups such as families and occupations.

The relationships that the very different strands of historical sociology developed with history have coloured their relationships with historical geography. For the macro-sociologists, the primary interest has been in the political history of the nation-state, although Abbott has argued that, at least in the 1980s, the degree of ‘self-referentiality’ in this group has worked to limit even that connection (1991: 221). For the sociologists attracted to the SSHA, the initial links were made with quantitative history. These were forms of history that many historical geographers treated with caution. The legacy of geographical influence continued to hamper their consideration of political history. In addition concerns over the presentism of many of the ‘quantifiers’ in geography caused many to be sceptical of quantitative history, although some urban and historical geographers did make links to quantitative history, in part through the SSHA.

Contrasts, Encounters, and Intermediaries

Given the reluctance of historical sociologists and even historical geographers to define their domains, any discussion of their relationships becomes problematic. One way to compare these fields is to explore their reactions to approaches associated with the Annales and world-systems theory. Spokespersons for both fields have testified to their debts to the Annales ‘school’ of social history and named Marc Bloch as an important precursor. The interpretations of Bloch’s message have, however, differed considerably. For historical geographers, Bloch has come to represent synthesis, a greater use of concepts, and a more social object of study, but historical sociologists have tended to stress his comparative approach, even going so far as to argue that he sought causal generalization and theoretical conclusions (Friedman, 1996: 174-5). Practitioners of both fields have also admired Fernand Braudel, but historical sociologists have tended to be less critical of his work. Abrams, for example, ended his influential book on historical sociology with the words, ‘[I]t is works such as Braudel’s Mediterranean that point the way’ (1982: 335). By contrast, historical geographers have often been uncomfortable with his sweeping approach and have criticized Braudel’s ‘geographical’ interpretations as being far too deterministic, coming too close to the earlier traditions of geographical influence and geographical history (for example, Clout, 1988: 71; Prince, 1975: 104-5). World-systems theory, so central to much of historical sociology, has also been treated with some scepticism within historical geography and portrayed as too simplistic and divorced from place. In the mid-1980s, when political geographers attempted to introduce Wallerstein’s work on the world-system to geography, Kearns complained of his ‘universalist rhetoric,’ which paid little attention ‘to the history of market institutions, to the evolution of rationality,’ and so forth (1988: 282-3), and Peter Hugill (1988: 111-27), while proclaiming an interest in ‘macro-historical geography,’ criticized Wallerstein for being historically inaccurate (1988: 111-27).

At a broader level, historical geographers have had mixed feelings about two of the defining characteristics of historical sociology, the use of theory and of the comparative method, as well as the quantitative methodology adopted by some in the field. Nevertheless, with time, some use of theory has become more acceptable. In 1975, Richard Dennis asked for a ‘more sophisticated integration of diverse aspects of social, economic and demographic theory’ (1975: 405), but in 1991 wrote of the dangers of an uncritical use of theory divorced from archival or fieldwork: A more recent danger is that of flirting with a trendy social theory which may provide the flimsiest of wrapping; almost anything can be dressed up as ‘structure’ and ‘agency’; ‘narrative’ is sometimes an excuse for the absence of a conclusion (1991: 281). Even Cole Harris, who had defended of historical geography in 1971 as a field of synthesis, has argued for a closer examination of theoretical context but again without abandoning crucial links to empirical work (1988: 329; 1997: xiii). Whereas historical sociologists have been reluctant to use the writings of Foucault and Giddens as models, the former being too cultural and the latter linked to a sociology seen as insufficiently historical, their writings have inspired a number of studies in historical geography. One of the theoretical works from historical sociology that has had the most positive reception in historical geography is Michael Mann’s The Sources of Social Power (1986), combining as it does an interest in the territorial arrangement of power with, in Harris’s words (1991: 675), a ‘command of the historical record.’

Uneasiness over comparative and quantitative methods continues. In 1986 at the ICHG held at Baton Rouge, some, such as Cole Harris, advocated more generalized comparative studies, but others, such as Sam Hilliard, saw ‘little need to move away from traditional concerns or methods’ (Baker, 1987b: 195). Many historical geographers feared that quantitative methods would reduce history to data for testing (Dennis, 1991: 267; 2001: 18; Mitchell, 1987). For example, in a critical 1987 review of a work by William Sewell, Michael Heffernan charged that he had replaced the historical voices with ‘the formal language of quantitative sociology’ and treated nineteenth-century Marseilles as a laboratory ‘more concerned with data and method than people and places’ (1987: 211-12). Despite these doubts, Dennis has called for a ‘re-integration of quantitative and qualitative perspectives’ and, like many others, including Heffernan, has championed ‘eclecticism’ (2001: 19-20).

Recently there has been somewhat greater convergence between historical geography and historical sociology in terms of objects of study. After years of severing links with political geography, some have now turned to questions of colonialism and the state and a few have even begun to explore questions of war and public memory. The questions of modernity are also becoming topical. Similarly after limiting discussions to human transformation of the land in reaction to the determinist interpretations, some in historical geography are beginning to look more carefully at questions regarding the social construction of nature. The hesitancy to focus on events continues, so that rather than addressing particular environmental catastrophes, such as famine and floods, historical geographers have sought more broadly based environmental narratives (Demeritt, 1994; Williams, 1994). Although historical geography has come under attack for its failure to incorporate a feminist perspective (Kay, 1900, 1991; Rose and Ogborn, 1988), that too has begun to change.

Another way to explore the relationships between these fields is to compare the Journal of Historical Geography with the Journal of Historical Sociology, both of which are international journals with strong Anglo-American slants and an aim to encourage open discussion about the nature of their respective fields. In part due to its expressed interest in publishing substantive rather theoretical pieces, the latter is probably less representative of historical sociology than the former is of historical geography, but that aim does make it a good site in which to look for any convergences between these fields.

The journals differ in both format and the character of their substantive articles. Reflecting its field’s emphasis on a critical examination of sources, the Journal of Historical Geography has an extensive book review section, often exceeding thirty pages -a contrast to the Journal of Historical Sociology, whose editors declared that it would not carry ‘regular book reviews.’ In the Journal of Historical Geography, books are often reviewed on the basis of their scholarship (particularly if it challenges stereotypes) and the effective use of primary sources, with a sizeable number of reviews devoted to books seen as research tools. Other criteria used for evaluation typically include the layout, maps and other visuals, and the writing style—again qualities that are deemed very important within the field. Another standard feature of the Journal of Historical Geography, lacking in the Journal of Historical Sociology, is the conference reports. In many ways, it is a discipline-building project intended, in the initial editors’ words, to monitor historical geography’s ‘undoubted and rapid progress’ (Clark and Patten, 1975: 1). By contrast, there is no direct equivalence to the ‘Schools and Scholars’ section of the Journal of Historical Sociology within the Journal of Historical Geography. Disciplinary creation within historical sociology still seems to hinge on the identification of particular thinkers as precursors, rather than the chronicling of growth and ‘progress’ that is so characteristic of historical geography. Both journals publish debates, with greater coverage and more theoretical discussions in the Journal of Historical Sociology in its ‘Issues and Agendas’ section. By contrast, in the relatively few debates published in the Journal of Historical Geography, discussion has often hinged on evidence and historical specifics, rather than on discussions of theory.

The disciplinary affiliations of the contributors to these two journals also differ. For the Journal of Historical Sociology, the number listed as associated with sociology departments has been surprisingly small, and historians and anthropologists have been well represented. When it first started, geography was also indicated as a contributing discipline, and David Harvey’s work was cited as a good example of the blurring of disciplinary boundaries. Harvey was listed as an associate editor from 1988 to 1991, but has not used the journal as an outlet, although he has published a few short pieces in the Journal of Historical Geography. In fact, very few geographers have contributed. Contributors to the Journal of Historical Geography have been less interdisciplinary, being predominately geographers, with a good representation of historians, but very few sociologists. The section of the Journal of Historical Geography that does include a much wider disciplinary representation is the book reviews, which include reviews from and about many different disciplines, including a few on historical sociology.

For both journals, the bulk of the space is devoted to substantive articles. In the Journal of Historical Sociology, after noting that ‘occasional pieces of a theoretical nature’ would be published, they announced, ‘The bulk of what we publish will focus on the particular and concrete—specific peoples, in specific places and times’ (Corrigan and Sayer, 1988: 4). Nevertheless, their articles would include more theory than typical of historical geography, and fewer visuals. Although both aimed at wide coverage in terms of place, the sociologists were to be more successful. In the Journal of Historical Geography, with some exceptions, the majority of places covered were Anglo-American, European or British outposts such as South Africa, Australia and New Zealand.

To give a more concrete sense to this contrast, I will compare seven recent articles on South Africa published between 1995 and 2000 written by those associated with geography or sociology departments. This selection was made both because of a recent shared interest in South Africa and because those writers have explored questions of colonialism, postcolonialism, the state and feminism, all topics with demonstrated potential for convergence between the fields. Most of these authors have academic links to both South Africa and Britain or Canada, and four hold positions at the University of Witwatersrand in Johannesburg: the geographer Charles Mather and the three sociologists (Belinda Bozzoli, Jonathan Hyslop and Deborah Posel). The sociologists have also been associated with the History Workshops there, workshops with strong ties to the History Workshop in Britain (J. Brown et al., 1990: 6)—particularly to the work of E.P. Thompson, whose writings have been so influential in historical geography.

Despite their common interests and ties, these geographers and sociologists differed in the objects studied and in how those objects were constituted and investigated. In general, the historical geographers relied more heavily on archival sources, often supplemented by oral histories. As for the sociologists, there was a greater tendency to rely on secondary sources, including many theses, but Posel (1995) did provide extensive references to archival records and governmental reports and Hyslop (1999) used some archival records from Britain in addition to many secondary sources. Less use was made of oral histories by the sociologists, although both Bozzoli (2000) and Posel (1995) drew on sessions from the History Workshops.

Three of these authors tackled issues of gender relations. For the sociologist, Deborah Posel (1995), this was approached through an examination of the conflict over the registration of customary marriage and framed in terms of the relations between the state, power and gender. Exploring the competing constructions of male authority between tribal chiefs and state officials, she argued that state power should be disaggregated into different ‘styles of governance’ corresponding to contradictions between urban and rural native administrations. To conceptualize these conflicts further, Posel turned to Barrington Moore’s ‘theorization of authority,’ stressing the contesting and renegotiation of the social contract. Both the geographers chose to approach gender relations through an examination of women’s active roles, rather than conflicting constructions of patriarchy. Jennifer Robinson (1998) investigated the Octavia Hill Women Housing Managers and their relationships with their tenants in South Africa, stressing the active mediation between masculinity and femininity informed by specific historical experience linked to nineteenth-century female philanthropy. Accordingly, she criticized those who assumed state power and citizenship to be masculine, suggesting greater complexity. In her article on the migration of the Bechuanaland women to South Africa, Camilla Cockerton (1996) stressed their migration strategies, motives and agency, which came to subvert both colonial and Tswana male authority. Rather than explore theoretical links to governance and authority, she sought specific social and economic causes for the migration, categorizing it into three spatial types.

Both Belinda Bozzoli and Charles Mather expressed an interest in the state and space which they explored through case studies at a community level: the sociologist Bozzoli studying revolts in Alexandra Township in Johannesburg, and the geographer Mather, the forced removal of the Ngomene in the Transvaal lowveld. Bozzoli argued that her case strongly resembled both urban and rural revolts elsewhere in South Africa but not elsewhere in Africa (2000: 82-3, 109). By contrast, Mather (1995) stressed the regional variability of the struggles over forced removals within South Africa and their resistance to broad generalization. Accordingly Bozzoli framed her discussion within such broad concepts as the rise of modernism and governability, writing of two ‘ideal typical eras’ of welfare paternalism and racial modernism, and drew on such writers as Charles and Louise Tilly and George Rudé as she explored questions of ‘style and repertoire.’ Her central object was the township as a social system and the revolts were linked to the replacement of a system of patronage and clientism with one of bureaucratic racism, which removed the bases of stability. Mather took a more internal approach focusing on agents other than the state, including land companies, white farmers and African communities, as well as divisions within the state bureaucracy, and stressed the economic and social bases of the struggle rather than the social system. In contrasting treatments of space, Bozzoli argued that spatial factors, such as ‘architectural and spatial decisions’ to exert control, were only a contributing cause acting in concert with the paternalistic nature of government, and for Mather space was depicted in terms of its redefinition and reconquest in the struggles and as roughly equivalent to rights over the use of the land in terms of access, hunting and agricultural practices.

The final two articles examined constructions of race: the sociologist Jonathan Hyslop (1999) through a study of the ideology of White Labourism and the geographer Alan Lester (1998) through one of discourses of racial otherness. In response to an earlier article in the Journal of Historical Sociology by the social geographer Alastair Bonnett, Hyslop argued that White Labourism was not specific to Britain but instead could be best understood through a historically connected comparison of the imperial working class in Australia, South Africa and Britain. To demonstrate this, he tracked the movements of ideas and specific people focusing on three ‘vectors’ associated with Australians, the Cornish and the Amalgamated Society of Engineers. Shying away from broad theoretical statements, he drew parallels to other writers in labour history who had argued that empire not the nation-state was the relevant geosocial framework and, reflecting his links with the History Workshops, argued for a history from below. Rather than undertaking a comparative imperial study of ideology, Lester focused on developments in the Eastern Cape Colony at the periphery of empire and on the interaction between peripheral and metropolitan discourses. Showing some willingness to discuss theory, he shaped his article as a reaction to overly simplistic postcolonial approaches, which he criticized for abstraction of conceptions of otherness from their historical contexts and for a focus on metropolitan discourse. In addition he argued that the peripheral discourses were ‘informed by material conflicts over land and other resources’ and accordingly examined the ‘spatial strategies’ of the British to separate the colonists and the Xhosa, first through expulsions and later through the incorporation of their labour while maintaining a racially segregated space.

These articles illustrate both convergence and remaining differences between the geographers and sociologists in terms of objects and methods. The legacy of ‘historical geographers’ links with economic history, their insistence on scholarship, and their reluctance to become overly theoretical is evident. Harking back to links with economic history, the geographers were more likely to seek economic and material causes, often tied to the ability to use the land, in contrast to the sociologists, who tended to stress community structures a contrast which shaped their treatment of space. The geographers’ historical approach was often that of an internal examination of competing groups and agents whereas the sociologists were more likely to stress structures of authority. Thus the sociologists tended to frame issues related to the state in terms of governance and citizenship and the geographers in terms of limits to state power associated with agents either within or outside the state—be they housing managers, land companies, Tswana migrants or settlers. All wrote of the importance of historical specificity, and yet the sociologists seemed more willing to generalize, if only within a South African context for a particular period. As one might expect, the sociologists were quicker to draw on the theories of others related to authority, rebellion and empire, while the geographers’ references to theory served primarily to demonstrate the limits of pre-existing theories such those associated with feminist state theory or postcolonial theory.

Despite some convergences in interests and methods, a gap remains which is difficult to bridge. The nature of this gap can be illustrated by an exchange that took place in 1987 and 1988 between David Harvey, depicted at that time as the historical sociologist’s geographer, and Richard Dennis, who was among those pushing for a more theoretical approach to historical geography. Dennis reviewed two of Harvey’s books that attempted to add a spatial dimension to Marxist theory and, in Harvey’s words, to ‘bring theory and historical-geographical experience together in such a way as to illuminate both’ (1985a: xvii). Criticizing Harvey for relying too heavily on secondary sources, ignoring relevant work within historical geography inspired by Harvey, and an ‘unnecessary obsession’ with validating Marx, Dennis concluded, ‘What matters are Harvey’s own ideas, prompting us to probe beneath superficial patterns, to seek connections and marry his perception to our own honest toil’ (1987: 315). In his indignant reply, Harvey objected to the implication that ‘[t]he only honest’ form of academic labour in historical geography is that applied to the digging up of such primary materials and all other forms are (presumably) unproductive of value and therefore in some respect ‘dishonest…’ (1988: 305). Furthermore, he charged that he had used primary materials in the form of novels and had dwelled in the ‘archive of Marx’s thought.’ In turn, Dennis answered that his review had supported ‘an introduction of theory’ into historical geography, and that, rather than being a defence of unfocused primary research, it had been a plea for ‘empirical research tailor-made to the argument’ (1988: 308). The gap between toil and theorization remained even for the geographers Dennis and Harvey—one seeking to make historical geography more theoretical and the other to link theory to historical-geographical experience.

The Weight of the Past

The legacy of historical geography’s struggle against simplistic explanations lives on. The historical geographers of the 1930s fought environmental determinism by turning to historical sources to demonstrate the underlying complexity. As J.H. Andrews observed in a 1980 review:

What the ‘descriptive’ pioneers opposed was the now-forgotten doctrine of environmental determinism, and they did it by showing (with some relish) that the same putative causes had produced different results in different historical periods. Space does not allow a closer analysis of this anarchic, anti-theoretical strain, but it certainly should not be underestimated. It survives in many a present-day scholar as an urge to make the world look more complicated than had previously been thought…. (1980: 205)

Historical geographers strove to distance themselves from political history, leaving aside questions of the state, and cultivated scholarly research instead of theorization, generalization and comparisons. By contrast, in their fight against grand theory and abstracted empiricism, historical sociology turned to political questions at the level of the state and, as a form of sociology, attempted to make history more theoretical. For some micro-sociologists, there was also a temptation to treat history as a laboratory in which to test their theories. Demonstrating competing visions of geography, the historical geographers have identified their field as geography by stressing past geographies, landscape, region and place and have been sceptical of involvement with spatial analysis, whereas historical sociologists, have turned to space, stressing its social production, showing little interest in the concepts of landscape, region and place.

By their very refusal to draw strict boundaries around their respective fields, historical geographers and historical sociologists continue to encounter each other around topics such as urban and labour history, feminism, colonialism and the social construction of nature. Yet even there the degree of convergence has been limited. While the historical sociologists strive to compare, find similarities and explain, the historical geographers strive to poke holes in their arguments, highlighting the complexity with their scholarship and eclecticism. It remains to be seen whether their common interests will be enough to overcome their very different everyday work habits—whether in their turning and toiling, practitioners of these two fields can find common ground.