The regional survey movement and popular autoethnography in early 20th-century Britain

Abstract

This article’s subject is the theory and practice of ‘regional survey’, the method of social and environmental study associated with Scottish thinker Patrick Geddes (1854–1932). Despite being overlooked or dismissed in most accounts of early 20th-century social science, regional survey had a wide influence on the development of the nascent disciplines of anthropology, sociology, and human geography. Emerging from late 19th-century field biology, the regional survey came to typify a methodological moment in the natural and social sciences that favoured the holistic analysis of geographically delimited areas. By the interwar period, the kinds of projects that went under its name can clearly be seen as forerunners of the post-Second World War tradition of community studies. Additionally, in its self-presentation as a civic, participatory exercise, the regional survey can be read as a form of popular autoethnography that contrasts with other, more familiar social-scientific ventures in the first half of the 20th century, and defies the dichotomy between ‘gentlemanly’ and ‘technical’ modes of social science. As a result, this article argues, the regional survey provides an alternative point of departure for thinking about the origins and development of the modern social sciences in Britain.

According to what is now a large and varied body of literature, the interwar decades formed a period in which the British Islanders found themselves increasingly subjected – and willing to subject themselves – to the kind of exoticising, anthropological gaze they normally reserved for far-flung corners of the globe. Opinion polling, the ‘documentary turn’ in film and literature, the photo-magazine Picture Post, and even the late careers of the Bloomsbury writers have all been examined as part of a culture of autoethnography (Anthony and Mansell, 2012; Cunningham, 1988; Esty, 2003; Hall, 1972; Igo, 2007). So too, in a different way, have the interwar period’s new ‘observational cultures’: citizen science projects like Max Nicholson’s ‘new ornithology’, or Dudley Stamp’s Land Utilisation Survey, that placed new emphases on ordinary citizens becoming close observers of their own immediate environments (Allen, 1994: 228; Macdonald, 2002; Rycroft and Cosgrove, 1995; Toogood, 2011). Above all, though, it is the social research organisation Mass-Observation, the movement with a mission statement to create an ‘anthropology of ourselves’, that has been the ur-specimen for this body of literature as a whole (Buzard, 1997; Harrison, 2014; Hinton, 2013; Hubble, 2006). M-O, more than most, has been seen to have epitomised a mid-century moment when observing and being observed, knowing oneself as well as one’s place in the collective, became a vital ingredient in the making of the democratic citizen.

Without wishing to deny the importance of this cultural moment, in this article I contend that the historiographical attention given to modernist projects like Mass-Observation risks obscuring the extent to which alternative models of autoethnographic practice persisted and prevailed in the first half of the 20th century. I pursue this line of argument via an examination of the ‘movement’ for regional surveys that took shape between the 1890s and the 1930s. Most commonly associated with the Scottish thinker Patrick Geddes, a regional survey was a holistic, ecological form of study, involving its practitioners in a tireless pursuit of data on almost every conceivable aspect of the locale: its climate, vegetation, geology, history, architecture, and more. In many respects, the rhetoric surrounding regional surveys was remarkably similar to M-O’s. Both promoted scientific observation of one’s own world as a social good, and both promoted themselves as a democratic autoethnographic project. One proponent of regional surveys even spoke of making ‘every citizen his own sociologist’ (V. Branford, 1921).

Where the regional survey departed, however, from the more familiar culture of interwar autoethnography was in its practice of a markedly different spatial politics. Running throughout M-O and allied projects was a distinctive mode of spatialisation that sited their mass publics within an abstracted yet integrated British nation. Documentarists invoked the ‘spaces of British identity’ (Anthony and Mansell, 2012; emphasis added); modernist writers contributed to ‘national culture’ (Esty, 2003; emphasis added); citizen science projects busied themselves with ‘mapping the modern nation’ (Rycroft and Cosgrove, 1995; emphasis added). Projects oriented towards reclaiming the cultural wholeness of the workaday nation made sense, too, in a country experiencing imperial contraction, limited democratisation, and state expansion. The regional survey’s spatialisations, by contrast, owed more to Victorian cultures of natural history than the high-modernist culture of scientific governance. In their emphasis on the specificities of place, regional surveyors carved out an autoethnographic practice that put the ecological community, rather than the massed nation, at its centre.

Apart from providing an interesting counter-example to an established historiography on interwar autoethnography, this article also argues that attending to the regional survey movement allows us to reframe some influential narratives about the development of the social-scientific methods in 20th-century Britain. That argument proceeds on two fronts. First, I aim to show that, contrary to some more traditional accounts that have sought to downplay its significance – mostly from within disciplinary histories of sociology: a point I will return to at the end – the regional survey had a wide and deep impact on the 20th-century social sciences. Second, I argue that the distinction I have just drawn between the way a project like Mass-Observation and one like the regional survey movement spatialised their objects provides us with a better grasp of the tensions inherent in the ‘politics of method’ than the now influential contrast between ‘gentlemanly’ and ‘technical’ social science offered up by Mike Savage. Savage, in fact, uses both sets of distinctions, but as he appears to acknowledge, the two do not always map on to each other. If, on one side, there is a social science that displays a ‘locationless logic’ (Savage, 2010: 62), a science of abstracted social groups lifted out of their specific location, then it is one that includes the idealist sociology of Hobhouse and Marshall (gentlemanly), Mass-Observation (technical), and John Goldthorpe’s Affluent Worker study (technical). On the other is what Savage calls a ‘landscaped’ form of social analysis, rooted in place (ibid.: Chapter 6), and it includes the surveys of Booth and Rowntree (gentlemanly), post-war community studies (technical) – and, I would argue, regional survey. Paying attention to the influence of the regional survey movement thus allows us to position it as a missing link in a story less about the conflict between gentlemanly and technical modes of social science than about a tension between rival forms of spatialisation.

The first section of this article traces the origins of the regional survey to late 19th-century natural history, contextualising its peculiar form of landscaped analysis within a wider turn to the field in the natural sciences. The impact of that turn on a related discipline, anthropology, has been richly documented: I thus place Geddes’s development of the regional survey alongside the work of his friend and interlocutor, Alfred Cort Haddon, seeing them both as participating in a new project of ‘ethnographic holism’. The second section then attends to the politics associated with the regional survey’s methodological commitments, revealing how its spatialisation along the site of the regional community offered an alternative to the managerial, statist vision of society implied in contemporary forms of social science. I then turn to the specific fieldwork and observational practices of regional survey, showing how they mapped onto democratic imperatives towards self-knowledge. I discuss these practices with the aim of demonstrating which parts of them did and did not survive into later social investigation as the social sciences professionalised, the subject of the final section. I explore in particular the links between regional survey and post-war community studies, viewing the latter as the culmination of a trend towards localised social research that had been set in motion by the regional survey.

Ethnographic holism

1892: Edinburgh, the Arans

Patrick Geddes first put the regional survey into practice in 1892, at the series of University Extension Courses he had been running annually while employed as Professor of Botany at the University of Dundee. The Edinburgh Summer Meetings, as the courses came to be called, provided multidisciplinary instruction as well as the opportunity to acquire knowledge first-hand (Sutherland, 2009: 361–2). On field excursions around the city, students took photographs, mapped distributions of flora and fauna, collected meteorological data, tabulated demographic statistics, and recorded customs in neighbouring fishing villages. ¹ Back in the classroom, they constructed relief models, pored over historical documents, and traced maps charting the growth of the city from its medieval beginnings (P. Geddes, 1906: 17–18). Students would supposedly emerge from the meetings able to discern the region holistically, to ‘read its evolution on every side’ (V. Branford, 1893: 218). Having begun ‘from everyday life’, students would strive towards ‘a scientific analysis of its many conditions, social, biological, and physical’. ²

The year 1892 also saw the launch of the British Association’s Ethnographic Survey of the United Kingdom, an ultimately ill-fated attempt to settle the question of Britain’s ‘racial history’. Ostensibly quite a different project to the one Geddes was developing in Edinburgh, the Ethnographic Survey nevertheless aligned with Geddes’s projects in its commitment to what James Urry has called a ‘holistic’ approach to ethnographic research. Provincial naturalist societies were given the task of undertaking comprehensive studies of specific localities, collecting data on local traditions and beliefs, ‘physical types’, dialects, and archaeological finds. According to Urry, the ascendancy of this kind of ‘ethnographic holism’ was embodied above all within the figure of Alfred Cort Haddon, one of the survey’s chief architects (Urry, 1993: 88). Haddon’s own preliminary work for the survey, conducted on the Aran Islands off the west coast of Ireland, amassed a vast store of data under headings including ‘physiography’ (physical geography, zoology, botany), ‘anthropography’ (anthropometry, population density, health statistics), ‘sociology’ (occupations, customs, clothing, transport), folklore, archaeology, and history (Haddon and Browne, 1891). His work typified the ultimately influential approach to ethnographic research that he variously called ‘anthropogeography’, ‘the extensive study of limited areas’, or ‘regional survey’ (Haddon, 1900b, 1910).

The regional survey and the natural sciences

Haddon is a useful starting point for discussing the regional survey because, while much is known about his role in the development of field techniques in anthropology, less often acknowledged is his intellectual affinity with, and debt to, Geddes (though see Walsh, 2021). Haddon was a frequent visitor to and lecturer at Geddes’s summer schools in the 1890s, and the two men had a correspondence stretching over 30 years. Placing what we know about Haddon’s ‘ethnographic holism’ alongside what we know about Geddes’s career helps us to see them both as participating in a joint methodological enterprise that shaped the human sciences at the turn of the 20th century. Evident in both the Ethnographic Survey and Geddes’s efforts in Edinburgh, three core elements made up this enterprise: an emphasis on fieldwork, a tight focus on the specificities of place, and a tendency towards a methodological and conceptual holism. All three were derived from the shifting practices of the biological sciences in which both Geddes and Haddon had been incubated.

The degree of similarity in Geddes’s and Haddon’s early careers is striking: both received their initial training in the 1870s, at Michael Foster’s Cambridge physiology lab, where they imbibed a Spencerian conception of the organism as a bundle of forces that would be variously conserved and released in the course of adapting to its environment (Kuklick, 1998; Renwick, 2009). Both would then take up the opportunity of studying those organisms in situ, at some of the many field stations then popping up across the continent. At one of these stations, in Roscoff, Brittany, Geddes first made his name as a biologist (P. Geddes, 1882), while Haddon co-wrote his first publication after working on a dredging expedition in Ireland (Haddon and Shackleton, 1891). In 1879, both also spent a season at the Naples Zoological Station, where it appears likely they first met. The prominence of field study in the two men’s training also led both to emphasise the specificities of place in biological, and later social, study. Robert Kohler has referred to the late 19th century as the age of the ‘natural history survey’: a period when a fading enthusiasm for specimen collecting was being replaced by a vogue for charting and mapping species distribution within circumscribed localities (Kohler, 2006). It was precisely in this biogeographical spirit that, in 1888, Haddon first set sail for the Torres Strait (Kuklick, 1996).

On returning to Britain, the pair set about promoting natural history surveys at home. The first biogeographical surveys of Scottish vegetation were undertaken by two of Geddes’s students, Marcel Hardy and Robert Smith (Hardy, 1906; Smith, 1900; Withers, 2001: 210), both of whom he had sent to study with the French botanist Charles Flahault. Geddes also spent time, and with some success, encouraging natural history societies to conduct surveys of their local area (Allen, 1994: 217; Withers and Finnegan, 2003: 344). Haddon, too, promoted the use of surveying while working as a lecturer at Cambridge, and in the mid 1890s began taking his students out on survey trips to neighbouring villages (Schaffer, 1994: 5; Urry, 1993: 93). Evidently, Haddon’s work served as a model for Geddes: when Geddes was establishing a Naturalists’ Club in Dunfermline in 1902, he wrote to Haddon asking for a ‘brief outline of your regional survey of Cambridge’. ³ Haddon obliged, and replied that he was glad a ‘natural history movement’ was stirring in Scotland. ⁴ At the Dunfermline Club’s first meeting, Geddes reflected on the progress he felt Scottish science had made towards properly local study, listing an array of ongoing surveying projects – in botany, limnology, and anthropology – all taking place within geographically delimited areas (P. Geddes, 1903: 146–7).

Geography and human ecology

As Geddes’s reference to anthropological survey suggests, he – and, it appears, Haddon – never saw the study of living human specimens as a separate enterprise from field-based biology. This, too, was another legacy of their immersion in French scholarship, particularly in that generation of geographers who were seeking to make integrative connections between natural environment and social organisation. Geddes’s discovery and subsequent championing of thinkers like Élisée Reclus and Edmond Demolins, and through them both the work of Frédéric Le Play, is well known (Livingstone, 1992: 271–8; Meller, 1993: 38–43; Scott and Bromley, 2013: 81–5). But Haddon, too, was also for a time taken in by this ‘French school’ (Haddon, 1928: 406–7). From 1892 both men, along with Havelock Ellis, were embarked on a project to translate and publish articles from the Le Playist journal La science sociale. ⁵ One such article later appeared in the popular journal Knowledge (Haddon, 1900a).

Le Play’s emphasis on analysing functional relationships between environment and culture – Place, Work, and Folk, in Geddesian parlance – imparted two characteristics to what became the regional survey. The first was its spatialisation within the site of the natural region. Conducting a natural history survey was never simply a matter of designating a field site to study at random, but an attempt to show how specific natural environments could be distinguished from each other, and how each one generated constellations of evolutionary possibilities for the organisms within them. The second was the regional survey’s methodological holism. Tackling the complex of a region, from its geology to its social organisation, required the tools of a wide range of disciplines, and it is notable that both Geddes and Haddon displayed an essentially Comtean fascination with how those disciplines ought to be ordered in relation to each other. Haddon’s first presidential address to the Anthropological Institute in 1903, for instance, submitted a plea for a new anthropological ‘synthesis’ that could comprise ‘anthropography’ (physical anthropology as well as the ‘inter-relations between the physical and biological environment’), ‘ethnology’, and ‘psychology’ (Haddon, 1903: 11–13). In preparing it, he turned to Geddes for advice, and although the two differed on the relative meanings of ‘ethnology’ and ‘sociology’, both agreed that ‘Ethnography’ ought to be understood as ‘a description of a single group of man’, combining ‘Anthropography, Ethnology (Sociology) and Psychology alike’ (ibid.: 12). ⁶

Admitting of continuity between the ‘branches’ of science also entailed admitting of continuity between their objects. Regional survey in its Geddesian ideal type was intended to produce less an inventory of the region, than an analysis of the interconnections between the ‘physical’, the ‘organic’, and the ‘social’ (P. Geddes, 1927: 83) – later theorised as ‘cosmosphere’, ‘biosphere’, and ‘sociosphere’ (P. Geddes and Thomson, 1931: 646). Foregrounding the interdependencies underwriting the natural and social worlds, Geddes often liked to claim that a regional survey was an exercise in ecology, an attempt to understand the ‘larger physiology of nature’ (quoted in Boardman, 1978: 30). If an ecology of plants sought to scrutinise ‘their relations to their environment, and to one another, and to animals’, then an ecology of humans was one that recognised ‘our own lives are not only individual, but in intricate natural and human associations, and these in our respective regions of Nature’ (P. Geddes and Thomson, 1931: 45–6, 145).

It is worth emphasising that the nascent science of ecology was itself a direct beneficiary of Geddes’s holistic mode of thinking. The leading plant ecologist of the mid 20th century, Arthur Tansley, acknowledged that his discipline’s ‘establishment here was ultimately due to the influence of Geddes’, since it was Geddes who had first encouraged Robert Smith to embark on the vegetation-mapping project that, for Tansley, was the discipline’s founding moment (quoted in Fagg, 1928; see also Tansley, 1947). Another Scottish ecologist who crossed paths with Geddes in fin de siècle Edinburgh, John William Bews, also drew heavily on Geddesian and Le Playist frameworks in his 1935 outline of Human Ecology (Bews, 1935). In it, as in much of his work, Bews found a neat fit between Geddes’s ideas and the ecological philosophy of ‘holism’ propounded by Jan Christian Smuts, Bews’s patron and mentor throughout his career at Natal University College (cf. Anker, 2002). Geddes, too, recognised the connection, and in his last major text, Life: Outlines of General Biology, quoted Smuts approvingly (P. Geddes and Thomson, 1931: 445–7).

According to Alison Bashford, the rise of ecology in the early 20th century reflected a ‘particular intellectual moment in the history of science primed to think in integrative ways about any given object of inquiry’ (Bashford, 2014: 159). If most of the new generation of professional ecologists, Bews aside, were shy about studying humans directly, they were mostly clear that ecological principles could, in theory, apply to the study of human social life (see e.g. Tansley, 1939: 529). Turning our attention to the presence of Geddes and Haddon at the founding of their respective disciplines allows us to see how this ecological ‘moment’ played out in the human sciences. Both drew on their training in biology to offer a new resource for treating human sociality as a part of natural history, and a method for studying it using a holistic, multidisciplinary methodology based on first-hand observation within geographically delimited areas.

The regional survey in the human sciences

When Geddes and Haddon made the transition into the human sciences in the 1880s, they entered an arena where disciplinary boundaries, and questions of method, were rather more up for grabs than they would later become. Sociology, at this point, had no institutions of its own and no one to teach it in the universities. Geography fared marginally better, having active learned societies in England and Scotland as well as, from 1887, its first practitioner employed by a university: Halford Mackinder. As is well established, the geographical discipline proved receptive to Geddesian ideas, and it was when Mackinder hired one of Geddes’s students and most consistent devotees, Andrew Herbertson, as his replacement for the readership at Oxford, that the decades-long dominance of geography’s ‘regionalising ritual’ began (Livingstone, 1992: 271–90; Matless, 1992; Stoddart, 1986; Withers, 2001). By the interwar period, regional geography, practised through regional survey, found adherents across the new generation of university geographers, in figures like Charles Fawcett, Percy Roxby, J. F. Unstead, and, significantly, the Haddonian anthro-geographer H. J. Fleure (Livingstone, 1992: 282–90).

Anthropology, probably the most institutionally developed discipline in the late 19th century, also found its methodological underpinnings being transformed by the new emphases that Geddes and Haddon brought to it. It is notable, for instance, that when Geddes heard that Haddon had decided to ‘go into anthropology’, Geddes enthused about the new possibilities to raise the discipline ‘from the anatomical standpoint to the physiological’: that is, the study of living bodies adapting to their environment, not the ‘skull measuring business’ he felt characterised the subject. ⁷ Clearly, both felt as if they had the opportunity to shape the discipline, and a number of their writings reveal a rhetorical strategy aimed at staking out a position for their new approach. On one side, they sought to differentiate their subject matter from what Geddes called the ‘dead and poisoned sources’ of the then dominant physical anthropology, the ‘the skulls and skeletons’ that had left the discipline in a ‘fossilised’ state as well as contributing to a ‘crude racialism’ (V. Branford and Geddes, 1917: 197; P. Geddes, 1913: 92). On the other, they cautioned against the grand, global evolutionary schema of the Tylorian ‘armchair’ scholars and the ‘comparative method’ – which a friend of Geddes described to him as ‘reading and running’ ⁸ – on which their theories were based. Central to both critiques was a distinct emphasis on studying communities in place, and not on lifting cultural artefacts out of the landscape and pressing them into global typologies (see e.g. Haddon, 1903; Stocking, 1991: 159–73). As Geddes later put it, ‘The only societies we ourselves claim to know of, or care much about, are those societies which are to us definite and real, because we have seen, observed, and interpreted them in some measure’ (V. Branford and Geddes, 1919: 120).

Despite his pronouncements, Geddes never especially thought of himself as an anthropologist, but Haddon’s contribution to the subject is now recorded in a wide corpus of secondary literature (Kenny, 2016; Kuklick, 2011; Stocking, 1984; 1996: 102–23; Urry, 1993). According to it, Haddon’s intensive, field biology-inflected approach to ethnographic research was pivotal in collapsing a long-standing distinction between the fieldworker on the ground and the anthropologist in the armchair. The largely descriptive accounts he produced – evident both in his Irish studies for the Ethnographic Survey and in the two Torres Strait expeditions for which he is best known – were not ‘ethnographies’ in the sense the term would later acquire (Stocking, 1984: 77). But they did exhibit the holistic, ecological principles that were the hallmark of his methodology. And his preparedness to swap out sweeping accounts of the rise of civilisation for the fine-grained study of single areas placed him at the forefront of a new generation of professional fieldworker-anthropologists whose careers spanned the divide between the evolutionary universalism of the 19th century and the cultural relativism of the 20th.

Lastly, when sociology did eventually emerge in Britain as a distinct arena of intellectual activity, it, too, owed much to the efforts of Geddes and Haddon. In 1894, the latter began teaching undergraduate courses on sociology at Cambridge, the first of their kind in the country, their content decidedly Le Playist (Scott, 2020: 8). Geddes drew on Haddon’s courses when preparing the syllabus for the Edinburgh School of Sociology he founded in 1902, again the first institution devoted solely to sociology in the country. ⁹ Both developments also fed into what would be another significant founding moment for the discipline: the creation, in 1903, of London’s Sociological Society, this time owing to the efforts of another Geddes protégé, Victor Branford (Scott and Husbands, 2007). Haddon also appears to have been involved with establishing the society, and sat on its Executive Committee during its first years (Scott and Bromley, 2013: 55–7). In the society’s early meetings, largely devoted to discussing what kind of subject sociology ought to be, Geddes once again saw the opportunity of a relatively inchoate discipline to carve out a space for his programme. And once again, as he encountered rival approaches to the subject, it was the regional survey’s embeddedness in place that offered him a means of differentiation. When, for instance, he later criticised the kind of sociology practised at the London School of Economics (LSE) – he famously failed to get the first university chair there in 1907 – for its tendency towards ‘colourless’ abstraction, he was making an argument above all about the way those sociologists sought to explain social phenomena without reference to their spatial context. Their conception of ‘Society’, Geddes argued, was ‘at best, a very vaguely generalized term’. ‘The ways of the field naturalist’, he went on, ‘have to be taken over by the field sociologist’ (P. Geddes, 1909: 56).

As well as distinguishing his sociology from the ‘too abstract “Society and its Members”’ of the LSE sociologists (V. Branford and Geddes, 1919: 121), Geddes and his followers also sought to distinguish the regional survey from what might have been the closest analogue at the time to ‘field sociology’: the urban surveys of Charles Booth. Booth’s surveys were, of course, like Geddes’s, embedded in specific locales, and Booth had even displayed Le Playist leanings in designing his Life and Labour (A. F. Wells, 1935: 15; Wilson, 2018: 123, 141). But Geddes and his followers encouraged regional surveyors to ‘go further’ than Booth, first by appreciating that cities could only be surveyed naturalistically if placed in the context of their wider natural region, but also by examining the region’s history (P. Geddes, 1911: 537). Regions had to be temporalised as well as spatialised. In its true Geddesian form, the type of historical analysis involved in a regional survey took a peculiar shape, employing Geddes’s ‘valley section’ model to plot relationships between physical landscape and primitive psychological types, and exploring the legacies of these in modern culture (Fagg and Hutchings, 1930: 136–41; P. Geddes, 1925; Matless, 1992: 470) – a form of analysis in which, as Haddon put it, ‘we start with physical geography and find ourselves drawn into statecraft and political economy’ (Haddon, 1898: xviii). Despite these idiosyncrasies, the regional survey’s broad attention to history was one that later social investigators would widely adopt in their own work (see e.g. Mess, 1927: 429; A. F. Wells, 1935: 25).

Ethnographic radicalism

Situating Geddes and Haddon’s methodological projects at the birth of modern geography, sociology, and anthropology, is not, however, solely an exercise in genealogical storytelling; nor is it solely an exercise in demonstrating the biological roots of the social sciences (Renwick, 2012). It is also an exercise in recovering a kind of radicalism present in social science’s beginnings. Ciarán Walsh has recently situated Haddon’s work within the anarchism of the fin de siècle (Walsh, 2021). Kropotkinite themes pervaded both his and Geddes’s writings, and the spatialisation of both men’s practice on the site of the natural region can be read as an attempt to envision a world of intrinsically co-operative biotic communities. The anthropologists Anna Grimshaw and Keith Hart, in a similar tack, argued in a searing polemic that the generation of Haddon, Boas, and Rivers is best understood as an anti-statist response to the ‘rise and visibility of ordinary people as a force in history’ (Grimshaw and Hart, 1993: 21). The ‘scientific’ methods these ethnologists pursued sought to position the subjects of their research as active participants in an ongoing deep-historical drama. In Geddes, Nikolas Rose and Thomas Osborne have found an ‘dynamic, open, ethical’ sociological practice aiming to define a procedure for ‘each citizen to find his or her place in the course of evolution and to search out the best means by which to act upon the future’ (Osborne and Rose, 2004: 220).

We need not accept all these claims to appreciate that the two men were responding imaginatively to what Grimshaw and Hart call the ‘popular forces for democracy’ at the turn of the century (Grimshaw and Hart, 1993: 25). For Geddes, explaining human difference via regional ecology advanced a kind of egalitarianism just as it was a route to a vision of global citizenship. When he wrote that ‘we can best and most fully share in the world’s culture … by developing our local possibilities to the full’, it was probably the closest he came to stating the ‘Think global, act local’ slogan with which he has long been associated (P. Geddes, 1902: 303; cf. Stephen, 2004). Other, more familiar thinkers would find themselves adopting a more managerial view of human affairs. Grimshaw and Hart, for instance, positively contrast Haddon’s work with the later generation of functionalist ethnographers around Malinowski and Radcliffe-Brown. The way these thinkers ‘banished history’ from their analysis produced only images of ‘society as a rule-bound, self-sufficient entity, an eternal anonymous collectivity’, images that could be easily hitched to the needs of expanding state and imperial bureaucracies (Grimshaw and Hart, 1993: 21, 32). Similarly predisposed to technocracy were thinkers like H. G Wells and Julian Huxley. Like Geddes, they adopted an ecological understanding of human life. Unlike Geddes, however, their formulation of human ecology was less a matter of developing ‘local possibilities’ than a matter of efficiently managing global flows of energy and resources (Anker, 2002: 204–45; H. G. Wells, Huxley, and G. P. Wells, 1931). As Bashford has pointed out, their visions of world communities, and world states, tended to encounter democracy more as an obstacle than a basis for a new social order (Bashford, 2014: 179).

If the turbulent world of the fin de siècle produced thinkers like Geddes and Haddon, whose holistic method produced models of social organisation grounded in small-scale, ecological communities, then it also produced thinkers like Wells, Huxley, and the functionalist ethnographers, whose work found a more natural home in the growth of technocratic, managerial states. The contrast between the two is not strictly one of politics, at least not of left and right. Geddes’s own politics were complex, simultaneously radical and conservative, and the kinds of projects aligned with his encompassed a broad spectrum (Renwick, 2010). Smuts’s holism was deployed in service of naturalising racial segregation, just as the regionalism of geographers like H. J. Fleure pursued a kind of co-operative socialism (Anker, 2002: 80; Rees, 2019). But all of them took place within what Grimshaw and Hart call the ‘open atmosphere’ of the early 20th century (Grimshaw and Hart, 1993: 25). It is to that atmosphere, and the regional survey’s response to it, that we now turn.

Popular autoethnography

Geddes often claimed that the regional survey was an exercise in education as much as research; in creating new citizens as much as creating new knowledge. Pivoting away from Booth’s kind of social survey, regional study was a ‘field sociology’ that privileged the perspective of the cultural insider, not the perambulating investigator. And with the shift in perspective came a commitment to a different kind of social policy, one that placed greater weight on local planning, civic betterment, and on achieving both via collective self-actualisation. In some of his more romantic moments, Geddes referred to the regional survey as a way of understanding – and enacting – the ‘mind of the community’ (P. Geddes, 1905: 72; emphasis added). Geographers like Herbertson and Fleure later wrote about developing ‘regional consciousness’ through surveying (Herbertson, 1915; Peate and Fleure, 1930). Other supporters of regional survey promoted it as an exercise in acquiring a critical sociological imagination, revealing to surveyors how the domain of the social constrained them but also gave them possibilities for action. The authors of the first textbook on regional surveying argued that its key task would be to ‘direct the attention of men and women, and still more of the rising generation, to the natural conditions and social possibilities of the world around them in their own neighbourhood, so that … a more vivid imagination be directed to the realisation of ideals’ (S. Branford and Farquharson, 1924: 8).

The regional survey literature’s emphasis on knowing your social environment, on looking at it afresh, articulated a vision of what I wish to call popular autoethnography. Auto here refers both to studying one’s own social world and to the sense of ‘automatic’: several writers on regional survey wanted the techniques associated with it to be practised almost as a matter of reflex in everyday life. ¹⁰ Victor Branford wrote about surveying as a ‘social process modifying our knowledge of our whole environment, guiding our use of it also, and thus a real piece of contemporary social evolution’ (V. Branford, 1914: 64; emphasis added). Later writers referred to it as a ‘technique for living’ (Fagg and Hutchings, 1930: 142). ‘Ethnography’, likewise, refers to both the description of a single social group – in this case the regional community – and the techniques involved in studying it. At a time when the post-Haddon generation of ethnographers – W. H. R. Rivers, Bronislaw Malinowski – were emphasising the need to understand things ‘from the native point of view’, and in the process devising new techniques for using their own bodies as research instruments (Kuklick, 1998; Schaffer, 1994), Geddes was also devoting attention to the kinds of bodily and observational practices needed to become an interpreter of regions. For Geddes, however, the problem was complementary to the one posed by what later became known as participant observation. Here the issue was not so much how to pass into the ‘native’ mindset, since the regional surveyor was already a ‘native’. The problem, instead, was how to pass outside again – to obtain a detached, alien-like perspective on the cultural totality.

Geddes and Branford, for instance, often likened the observational practices of the regional survey to those of a child or a tourist, both figures who had not yet acquired the level of cultural inside-ness that they felt tended to immunise the adult self to its surroundings (e.g. V. Branford, 1921: 145; P. Geddes, 1904: 110). Other metaphoric figures in their writings – the hillway traveller, the airman – also captured the need he felt for regional surveyors to transcend their partial, limited, perspective, to perceive the cultural order holistically (V. Branford and Geddes, 1917: 89; Matless, 1999: 210). He even figured his Outlook Tower, the civic museum cum ‘sociological laboratory’ he purchased in 1894, as a kind of machine for making regional observers: ‘an autogenetic process … education in the best sense’ (quoted in Boardman, 1978: 194). Visitors to the Tower would learn about Edinburgh, Scotland, empire, Europe, and the world, and each from the perspective (or ‘outlook’) of different sciences: astronomy, geography, botany, history, economics. The Tower problematised and relativised the observer’s gaze at every turn, and yet the point was also that visitors would emerge able to combine the different ‘outlooks’ they encountered into one, holistic synthesis: the ‘geographic and social whole, the regional and civic unity before us’ (Chabard, 2004; P. Geddes, 1915: 321).

In presenting the regional survey’s observational practices as methods for locating geographic and social ‘unity’ in the site of the region, Geddes distinguished them both from an earlier, 19th-century version of ethnography ‘at home’, and a later, modernist iteration of autoethnography. If the former, in the shape of folklore research, urban social investigation, and the like, had found cultural authenticity or ‘unity’ at home, then it was in the residual pockets of a metropolitan culture that remained largely unexamined. Investigators distanced themselves from their subjects as they constructed mutual equivalences between domestic ‘folk’ and ‘primitive’ colonial societies (Dorson, 1968; Nord, 1987). The regional survey, by contrast, was always in Geddes’s formulation a matter of locating cultural coherence in the metropole itself, and especially in its characteristic social form: the modern, industrial, city-region. But in this sense the regional survey was also distinct from the kind of modernist autoethnography that several authors have detected in the 1930s, in cultural practices like documentary, domestic tourism, and Mass-Observation. Like the regional survey, these projects all relied on a set of observational strategies that held out the possibility, or necessity, of viewing one’s own social environment with the eyes of a stranger (Buzard, 1997). But they also took root in the context of a 1930s moment of national, international, and imperial crisis, where they became emblematic practices of a culture that was learning to appreciate the urgency of restoring ‘knowability’ to England. As Jed Esty explains, their signal feature was their tendency to proclaim the nation, not the region, as the ‘seat of authentic cultural totality’ (Esty, 2003: 46).

Yet even if the specific context of 1930s crisis gave impetus to newer forms of autoethnographic practice, those forms did not entirely supplant the older kind of autoethnography involved in regional survey. Regional surveying continued to be promoted and practised throughout the interwar period, in a variety of contexts; it is for this reason that I also wish to call the kind of autoethnography involved in regional survey popular. ‘All ages and classes’ would be able to contribute, since, as Geddes put it, ‘everyone we see is more or less of a sociologist’. ¹¹ His call increasingly found traction, during and after the First World War, among sections of a citizenry increasingly seeking tools with which to diagnose and remedy itself. Geddes himself wrote of surveying as a form of ‘convalescence’ from war, for ‘the disoriented citizen’ (V. Branford and Geddes, 1919: 136–8). He also noted, during the same period, the ‘civic enthusiasms’ brewing among the ‘least municipally powerful members of the community’: workers, women, and children (P. Geddes, 1915: 337). His goddaughter, the geographer Mabel Barker, called regional survey an ‘education for a state of peacedom’ (Barker, 1915).

Consequently, it was during the war that what claimed to be an organised survey ‘movement’ emerged. In 1914, approximately 50 interested geographers, town planners, social workers, and schoolteachers formed a provisional committee for the development of regional surveys. ¹² The committee, in 1918, became the Regional Association, with premises at London’s Le Play House, the new home of the Sociological Society. In 1919, the RA’s first annual report noted that, even by this early point, surveys were already underway in Croydon, Aberystwyth, Saffron Walden, Glasgow, Letchworth, Hampstead, Wakefield, West Ham, East Ham, Bournemouth, and the Wirral. ¹³

‘Movement’ is, however, something of a misnomer when it comes to describing the variety of surveying activity in the interwar period. Befitting a broad commitment to political decentralisation found in the writings of many regional surveyors, attempts to create a national ‘centre’ for regional survey were never wholly successful. Membership of Le Play House, the institution most likely to think of itself in this way, was relatively small. David Evans has put the figure in 1923 as 435 (Evans, 1986: 34). (This was about the same number, incidentally, that sat on Mass-Observation’s ‘national panel’ in the late 1930s, a fact that is difficult to glean from the discrepancy in historiographical attention afforded to each. ¹⁴ ) Much of Le Play House’s work in the 1920s took the form of the form of a kind of survey consultancy, providing assistance to voluntary organisations like the National Council for Social Service, or else operating as a travel agency running educational tours to Europe (ibid.: 30–1). Most regional surveys in the interwar period, in fact, started up independently of Le Play House. The forces generated by the regional survey were always more centrifugal than centripetal, thus providing a point of contrast with some of the better-known citizen science projects of the interwar period (cf. Benson, 2017). Rather than relying on a ‘technocratic hand’ gathering scattered local observations into a tidy national archive (Jardine, 2018: 71), regional surveys were intensive exercises relying on small groups operating in concert with each other, but not with a national centre.

In 1918, for instance, the geographer Percy Roxby formed a Liverpool and District Regional Survey Association, drawing in members from across Liverpool University as well as representatives from local civil society. ¹⁵ A number of similar projects, operating through civic and associational culture, started up in the 1920s, in Manchester, Huddersfield, Aberdeen, Exeter, and elsewhere. ¹⁶ Naturalists’ societies, field clubs, and even local photography clubs also began attaching the label ‘regional survey’ to their activities in the 1920s. ¹⁷ The new era of town and country planning, embodied in figures like Patrick Abercrombie (much influenced by Geddes), adopted the survey of the city-region as a core tenet (Meller, 1993: 294–300). And the Geographical Association, especially under the leadership of H. J. Fleure, established surveying as a pedagogical exercise in schools (Ploszajska, 1998). Its success in this respect was reflected in a number of high-profile educational ventures throughout the interwar period: in the ‘scheme for the collection of rural lore’ run out of the Welsh Department of the Board of Education in the 1920s (Gruffudd, 1996); in a 1928 Board of Education pamphlet offering guidance for teachers wishing to conduct local surveys with their pupils (C. V. Butler and Simpson, 1928); and in a long-running series of BBC schools programmes on British regional geography (published as Simpson, 1931).

Regional surveys thus suffused interwar British culture, and the wide variety of activities that went under the name makes calculating the exact extent of the ‘movement’ impossible. Broadly, however, the most archivally visible practitioners and adherents of regional survey were drawn from the socially minded sections of the middle classes: educators, social workers, and town planners. In their commitment to what the Regional Association described on its mission statement as ‘social betterment based on … an intensive and scientific knowledge of natural and social life’, they displayed many of the elements of what Savage finds, in Mass-Observation, as a protean ‘groping’ towards a ‘technical’ middle-class identity (Savage, 2010: 62, 64). ¹⁸ This was, Savage explains, an identity grounded in the ‘rational’ language of science rather than the ‘gentlemanly’ humanities, and in the interwar period, it was one that found expression in support for progressive ‘planning’ (ibid.: 67). Yet regional survey was also ‘gentlemanly’ in its own way, too: it concerned itself with mapping whole populations, and some of the more high-profile surveys conducted by Le Play House, such as at Chester, displayed many of the ‘moralising’ tendencies associated with an older vision of social investigation (Leach, 2017).

If regional survey thus troubles the distinction between gentlemanly and technical modes of social science, a different means of characterising its distinctiveness from contemporary autoethnographic projects lies in its orientation to what several writers called ‘bigness’. ‘The fatal device of too many state organisers’, wrote H. J. Fleure, ‘has been that of arousing enthusiasm for bigness of area and of statistics. The results have been dire enough, and we need to restore and develop conditions which shall be more favourable to spiritual inspiration’ (Fleure, 1917: 44). (The founders of Mass-Observation, by contrast, were rather more at home in state bureaucracies; see Harrison, 2014.) Sybella Branford, another authority on surveying, explicitly contrasted the place-based approach of the regional survey with prevailing Liberal-Idealist thought. That ‘German influence on thought’, she argued, had produced only the ‘abstractions’ of state and nation. ‘Citizenship’, she wrote, ‘is now taken to mean not a man’s or woman’s relation to the city or actual community in which he lives, but something so vague that it becomes quite indefinite’ (S. Branford, 1921: 228).

The regional survey and the professional social sciences

It was along these lines that the interwar regional survey began to depart from the mainstream of a social-scientific establishment increasingly centred in London. The Geddesian framing of regional survey as a collective exercise, in which citizens of all kinds could come together and study, first-hand, their own geographically bounded social world, ran up against the move towards professionalisation in interwar sociology and anthropology. As we have seen, academic sociology under Hobhouse and later Ginsberg tended to avoid fieldwork altogether. In other, more biologically oriented figures, such as Alexander Carr-Saunders, sociology made its claims to distinction not from the study of localised biotic communities, but from the rigorous statistical analysis of less geographically delimited ‘populations’ (Osborne and Rose, 2008). Professional anthropologists, by contrast, did still find themselves focusing on small-scale communities, but it was increasingly a disciplinary norm that these should be located far away from the British mainland. Moreover, professional anthropology also increasingly came to be nourished by stories of the lone fieldworker heroically performing the ‘ethnographer’s magic’ (Stocking, 1984). Even as early as 1914, W. H. R. Rivers was maintaining distinctions between ‘survey’ anthropology, conducted by a team (the kind he had himself performed with Haddon in the Torres Strait), and ‘intensive’ study, performed by a solo ethnographer (Urry, 1993: 28). In Geddes’s scheme, by contrast, the trained fieldworker, operating alone, could claim no special status.

Yet there was a different logic of professionalisation also available. Academic geography, as already mentioned, continued to promote and perform regional surveying, including as a participatory endeavour. Roxby’s Liverpool survey is a good example: organised out of the university’s geography department, it drew in participants from across the university as well as from civil society (Hewitt, 1922: 5). By 1926, one writer noted that similar projects were ongoing at university geography departments across the country, including at Aberystwyth under Fleure, at Exeter under W. S. Lewis, at Leeds under C. B. Fawcett, at Sheffield under R. N. Rudmose Brown, at Manchester under W. H. Barker, at Bristol under W. W. Jervis, and at the LSE under Hilda Ormsby (Russell, 1926: 440). As well as geography, the work of conducting social and sociological surveys also professionalised in ways that left space for the core tenets of regional survey to be adopted. Henry Mess’s survey of Tyneside openly acknowledged its debts to Geddesian regional survey, through its deployment of history as well as its spatialisation on the site of the wider city-region (Mess, 1927, 1934). It also framed itself as a participatory exercise, utilising the assistance of the residents of Bensham Grove, the local educational settlement (Freeman, 2002: 255). Other interwar provincial social surveys approached their task along similar lines: here we can include the ‘civic survey’ of Southampton (Ford, 1934), the social survey of Merseyside (Jones, 1934), and the ‘community self-study’ of Brynmawr (Jennings, 1934). As A. F. Wells pointed out in his 1935 survey of surveys, all four could properly be called ‘regional’ or ‘community’ surveys because of the way they attempted to study areas holistically, and to link geographical with social conditions (A. F. Wells, 1935: 24, 27).

If professional geography and local social surveying represented the two continuing traditions within the social sciences (education and town planning aside) that remained influenced by regional survey, then they were traditions that appeared to develop without much acknowledgement of each other. Baudry Rocquin suggests that what had started out as a unified ‘movement’ in the 1920s came, by the 1930s, to be split between geographical and sociological camps (Rocquin, 2019: 48–53). That split was visibly played out in 1930, in the rather acrimonious divorce at Le Play House between what became the Institute of Sociology and the more geographically oriented Le Play Society (Evans, 1986: 37). How much we ought to read into this split is arguable, however, since after the Second World War the geographical and sociological traditions of regional survey once again found themselves uniting in new ways. What became the post-war wave of ‘community studies’ owed much, at least initially, to the fusing of the geographical, the anthropological, and the sociological that defined much of the regional survey.

Indeed, it was notable that, before the expansion of sociology within British universities, some of the earliest ventures into the study of community after the Second World War emerged from those university geography departments most aligned with Geddesian regional survey. It was, for instance, at Aberystwyth, where H. J. Fleure and his successor Daryll Forde both taught, that Alwyn Rees’s Life in a Welsh Countryside (1950) first appeared. The same department also produced Bill Williams’s early community studies of Gosforth and Ashworthy. Another early ‘study in community’ issued from the pen of Patrick Geddes’s own son, who taught geography at the University of Edinburgh (A. Geddes, 1955). And Ray Pahl, another influential sociologist of community, also started life as a geographer, recalling early influences on his work, once again, as the ‘anthrogeography’ of Forde and Fleure (Pahl, 2011: 170).

Later iterations of community studies adopted a wider range of influences: Chicago sociology, functionalist anthropology, Max Gluckman’s ‘Manchester School’. Many of them, too, shed the aspirations to ‘ethnographic holism’ associated with the earlier vision of regional survey, preferring instead the study of a single sociological theme – class, kinship, race relations – within a bounded locale (Lawrence, 2019: 9). They also tended to drop the evolutionary, historical methodology, as well as the commitment to local study as a collective, democratic enterprise (although some elements of this also persisted, as when Margaret Stacey employed her Workers’ Educational Association classes to conduct her Banbury study). But even in these later, more strictly sociological versions of community studies, something of the older tradition persisted. When Michael Young founded the Institute for Community Studies, for example, he did so at least partly out of the same political critique of state ‘bigness’ found in interwar regional survey (L. Butler, 2020: Chapter 2). Young’s own politics had been honed in that tradition of interwar utopian thought whose members had also included Geddes. And as Savage shows, what the first wave of community studies all tended to share was a commitment to a ‘landscaped’ vision of social analysis, in which the social relations they studied were embedded within the peculiarities of place. It was only in the 1960s, when sociologists began to abstract their analyses from the specifics of the location they studied, that this landscaped vision was effectively closed down (Savage, 2010: 163).

Conclusion

Of the historians of sociology to have written about the moment of community studies, it is only John Scott who has attributed any influence at all to Geddesian regional survey (Scott, 2020: 86). Others have stressed the importance of functionalism; or in one case claimed Mass-Observation’s ‘Worktown’ as the first ‘full-blown’ community study (Foks, 2023: Chapter 5; Lawrence, 2019: 245). Mike Savage has read community studies as an ultimately failed quest to define an English ‘Middletown’, an ordinary location that could be rendered a stand-in for the nation as a whole (Savage, 2010: Chapter 6). Yet placing regional geography, nourished by the regional survey, back at the origins of community studies complicates this picture: as I hope to have shown, this was a tradition oriented less towards reading any one locale as typical of a broader ‘Englishness’ than towards examining, holistically, the specificities of each place.

The reluctance among historians of sociology to acknowledge regional survey’s influence speaks to a more general wariness about attributing significance to Geddesian ideas. In most accounts Geddes appears as an eccentric, marginal figure who spawned a largely ineffectual ‘movement’ between the wars that petered out as soon as sociology professionalised (Abrams, 1968; Bulmer, 1985; Bulmer, Bales, and Sklar, 1991; Kent, 1981). There is, of course, some element of truth in this: the ‘movement’ at Le Play House was institutionally weak, and Geddes was marginalised, deliberately so, by the sociological establishment at the LSE (Lybeck, 2013). But I have argued in this article that we should see regional survey less as an institutionally bound ‘movement’ than a methodological spirit present at the birth of the social sciences, which animated a wide range of research across the first half of the 20th century. Its distinctive approach emerged from a particular juncture in the late 19th-century natural sciences, in which a renewed emphasis on fieldwork, surveying, and the holistic study of place combined with a broadly organicist, anti-statist politics to give rise to new conceptions of ecological communities as the spatial units of social organisation. That approach had a wide influence on the practice of the social sciences at the turn of the century and beyond, and was evident in disciplinary traditions including Haddonian ethnographic survey, local social survey, and the new human geography.

The regional survey thus supplied an alternative tradition to what we know as a more familiar story of British social science in the first half of the 20th century. In it, the centres of gravity have tended to be professional sociology and anthropology at the LSE, or else projects like Mass-Observation, whose methods were less oriented to the geographic, the ecological, and more oriented to the politics of the state. Attending to the kind of alternative tradition embodied in regional survey, for historians of sociology, will therefore mean looking beyond London. It will also mean continuing to venture into the histories of neighbouring disciplines, in the mould of Renwick’s work on sociology’s biological roots, or Foks’s on sociology’s anthropological inheritance (Foks, 2023; Renwick, 2012). As I have merely begun to sketch out here, we also need to take geography seriously as a disciplinary player in the making of social-scientific thought.

There is also, I think, a deeper reason for some historians’ reluctance to acknowledge Geddesian regional survey as part of sociology’s history. Many of the first histories of sociology and social research were produced in precisely the post-1960s moment that Savage describes, one of the major achievements of which was to separate the social from the spatial – as well as from the biological. In this context, the regional survey’s emphasis on ecological, holistic analysis seems out of place within disciplinary genealogies. One assessment of Geddes’s contribution to community studies, for instance, dismissed it on the basis that it ‘tended not to produce sociological material’ (Crow, 2014: 379; emphasis added). Another claimed that Geddes’s approach to sociology ‘increased suspicion’ of the subject among the ‘conventionally minded’ (Bulmer, 1985: 11). Now that we appear to be in a moment when we are less certain about what ‘conventionally minded’ sociology entails – sociologists are re-examining their discipline’s relationship to biology, for instance (Meloni et al., 2018) – we have more opportunities to reassess our relationship to the discipline’s history. Doing so would not mean having to rehabilitate Geddes’s methods, or his politics (Fuller, 2007; Studholme, 2008; Studholme, Scott, and Husbands, 2007). But it might give us a different perspective on the trajectories and forces in operation in the making of social knowledge. As Chris Renwick recently pointed out, it is the sociology of the post-war era, not the early 20th century, that may eventually come to be seen as anomalous in the discipline’s history (Renwick, 2016: 153).

As well as arguing that the regional survey ought to be seen as a significant research intervention, I have also suggested that it can be viewed as an episode in popular autoethnography. Its self-presentation as a method that could, in theory, be performed by any citizen seeking to understand their social environment was one that failed to survive the professionalisation of the social sciences. But attending to the moment in which it was viewed in this way reveals something of the level of demand within a democratising culture for a set of intellectual tools, however blunt or unformed, with which to examine itself (Mandler, 2019). The kind of autoethnography involved in a regional survey had its practitioners understand themselves as part of a functioning social collectivity concretised in the space of the natural region. It thus filled an often overlooked interregnum between the ‘lost social totality’ of the 19th-century metropole and the recovery of an insular national culture in 20th-century modernism. Reconstructing this moment helps us to see early 20th-century Britain as a place not only increasingly devoted to observing itself, but one also questioning what it should observe, where, and how.