In an article first published in July 1968 in New Left Review, Perry Anderson gave an analysis of a critical weakness of British intellectual culture. His diagnosis is remarkable and surprising. One of the key problems,

Anderson argued, was that Britain has failed to make any contribution to the classical sociological tradition; moreover, this failure was indicative of a wider failure to be European and to be modern. Sociology, he suggested, was one of the great achievements of the ‘European bourgeoisie at the end of the nineteenth and twentieth centuries’. By contrast, British intellectual culture was marked by a lack: ‘why did Britain never produce either a Weber, a Durkheim, a Pareto or a Lenin, a Lukács, a Gramsci?’ [1] Today, while one could point out many weaknesses in British intellectual culture, few would argue that the lack of an indigenous sociological tradition is one of them. But whatever its strengths and weaknesses, something like Anderson’s analysis was certainly influential. From the 1970s onwards, social theory developed rapidly in British universities. Anthony Giddens, who had published a textbook on Marx,

Weber and Durkheim, [2] can be taken as indicative of this trend and was, for a time, a leading light. What was taken to be traditional British empiricism was rejected, and British social scientists read and reread the European sociological canon they should have been reading all along, rather than through the mediation of its American interpreters. Social theory rapidly became something of a meta-discipline, effecting a radical transformation of not only sociology but also English, in the form of the emerging interdisciplinary field of cultural studies. The political influence of social theory arguably reached its height in 1997 when Giddens himself proposed the Third Way as the project of New Labour. New Labour, in Giddens’s work, became a neo-Durkheimian project of moral renewal. [3]

But what was the new British social theory? How did it come to relate to ‘French social thought’? And why would the idea of the network come to be significant for its further transformation?

I begin with four observations. First, what came to be called social theory in Britain was largely devoid of any specific empirical content. The superstructure of social theory was not built on a base of empirical research, although it could make use of empirical examples, but only in so far as they illustrated more general theoretical claims. In this context, Marx’s interest in the reports of Her Majesty’s inspectors in Capital could be taken as of incidental importance to his thought. Weber’s concepts, such as rationalization and status, could be easily stripped of any relation to historical investigations of religion and economy, or the comparative analysis of culture. And the use of statistics by Émile Durkheim in Suicide was of much less importance than his methodological prescriptions, even if his use of statistics in practice did not necessarily follow the prescriptions laid down in Rules of Sociological Method.

The purification of theoretical analysis from empirical content was enormously powerful. It enabled social theorists to develop wide-ranging accounts of modernity and postmodernity. It established a clear hierarchy between the value of theoretical and empirical labour.

It made it possible to blur the boundaries between the social sciences and philosophy in a particular way.

Foucault, for example, who had insisted on the importance of an attention to detail in Discipline and Punish, could be read as a social theorist of subjectification and power, whose treatment of historical materials was of no particular interest. His account of discipline could be taken to sum up a whole society and then criticized by those who thought that he had failed to give any account of, for example, consumption or the media.

He, along with Bourdieu and Baudrillard, was added to the canon of European social theory. French social thought, which was believed to be intrinsically more theoretical than the indigenous product, as Anderson had argued, was vital to the reinvention of British sociology as a theoretically driven enterprise.

A second observation that can be made about Anderson’s essay is that, in his view, the natural sciences and the arts were not a critical part of the problem of British intellectual culture. They did not have any special political significance. After all, the natural sciences were objective, and the arts were too subjective to provide any rigorous analysis of global society. The social and the natural sciences were clearly distinct. In effect, Anderson understood ‘the social’ as divorced from the domain of the material, the organic or biophysical, the non-human, the vital or the aesthetic. The development of an indigenous social theoretical tradition, in Anderson’s account, would disrupt the boundaries between the various social sciences and humanities, but would leave the integrity of the borders between the natural and social sciences pretty much untouched.

Third, Anderson had surprisingly little to say about geography and anthropology in Britain, two indigenous forms of explicitly social thought. Perhaps part of the reason for Anderson’s lack of interest in these disciplines was that, in so far as they focused on specific regions or peoples or cultures, they failed to address the kinds of general problems of the ‘global reconstruction of social formations’ that he thought to be clearly lacking in Britain. They were too focused on the study of particular territories or cultures to be of more general significance. Anderson cited the absence of a chair of sociology in both Oxford and Cambridge at the time as a sign of the weakness of British intellectual life, although both universities had already established posts in geography and anthropology in the late nineteenth century. Britain still had an imperial intellectual culture, which was clearly not suited to the development of the kind of wide-ranging account of society that had developed in France and Germany: ‘suppressed in every obvious sector at home, thought of the totality was painlessly exported abroad, producing the paradox of a major anthropology where there was no such sociology. In the general vacuum thus created, literary criticism usurps ethics and insinuates a philosophy of history’. [4]

Fourth, Anderson had little to say about the rich history of applied social research in Britain which could not be regarded as merely administrative. If Giddens later made social theory part of the intellectual apparatus of New Labour, there was a prejudice against anything that seemed too practical or governmental, or too rooted in a particular local or national context. [5] Thomas Osborne and Nikolas Rose’s efforts to write to the history of social research in Britain provides a necessary corrective to Anderson’s prejudice, for ‘social thought owes as much to the machinations of people like doctors and bureaucrats as it does to the erudite reflections of quasi-philosophers’. [6]

It is in this context, all too briefly sketched, that we can understand some of the significance of the idea of the network in French thought, and its translation in Britain, to which I now turn.

The actor-network

A specific paper can be taken as a starting point.

In 1981 two French sociologists, Michel Callon and Bruno Latour, published an essay that laid the basis of what later came to be called actor-network theory – an approach which has now become extraordinarily influential across the social sciences in Britain. Why has this been the case? The paper’s title, with its direct attack against sociologists and a certain style of social theory, is suggestive: ‘Unscrewing the Big Leviathan:

How Actors Macro-structure Reality and How Sociologists Help Them to Do So’. [7] For Callon and Latour, the problem for social theory, in its ambitions to be something of a transdisciplinary metadiscipline, was that it helped to create macro-structures. It did not so much analyse how the social was assembled in practice as contribute rather too rapidly, enthusiastically and uncritically to its constitution.

The origins of actor-network theory in French thought were transdisciplinary. Latour and Callon’s 1981 paper cited the semiotics of Greimas, from which they took the concept of the actant. [8] Another source of inspiration was the philosophy of Michel Serres, whose work had been deeply influenced by information theory and thermodynamics, from which actor-network theory borrowed the concept of translation. [9] And they were also influenced by the Anglo-American tradition of micro-sociology and, although they were not aware of it at the time, its unacknowledged debts, via the Chicago school, to the late-nineteenth-century sociology of Gabriel Tarde. [10] But what was implied by the concept of the network in actor-network theory? There are three main ideas, none of which is adequately conveyed by the idea of the network in English.

First, networks are about relations. But the relations do not exist between distinct entities, such as individuals, institutions, classes and so on. Nor are networks something like structures within which individuals are located. Rather, networks are mobile sets of relations within which actors are progressively formed and transformed. Take a typical actor (actant), such as a drug molecule. A given drug molecule is not one thing. It changes its properties depending on whether it is found in a pure form in a lab, exists in a solution, is part of a tablet, mixed with water, or interacts with the body of the patient or the recreational user. You can’t say that the molecule has given properties. A drug molecule is not a thing which then has relations with other things – its properties are in process. Indeed, drug companies don’t make pure isolated substances.

They make networked things, with multiple properties and forms. [11]

Freee, Everyshop Window is a Soap Box, Liverpool, 2010I use the example of the drug molecule deliberately.

The notion of the actor-network brought with it a particular relation to science, which takes the form of neither critique nor celebration. While Anderson reckoned that any study of the natural sciences was irrelevant to the analysis of political culture, actornetwork theorists have argued that the study of the natural sciences is of central importance for all those concerned with contemporary politics. And the question of the distinction and relation between the natural and social sciences is necessarily a political matter. [12]

Second, the concept of the network is about scale.

Networks can, of course, exist at any scale. One can have a network between molecules, between persons or between states. Indeed, sociology, following Comte and Durkheim, had tended to imagine that social forms exist at a series of different levels or scales. At the bottom was the individual, then there were institutions, then there were national societies and economies, then there was global society. Each higher scale could be understood as the context or frame for the one below.

The micro was simply part of the macro, and the global was the most fundamental scale of all. All too quickly, social theorists lost any concern with the specificity of things, whether regions, industries or materials.

But as Callon and Latour saw it, the idea that there was a hierarchy of scales was a mistake in the first place: ‘too often sociologists – just like politicians or the man in the street – change their framework of analysis depending on whether they are tackling a macro-actor or a micro-actor.’ [13] Seen in terms of an analysis of networks, this scalar order was disrupted.

On the one hand, one could trace the constitution of so-called macro-actors. [14] On the other hand, what might appear to be small objects – a drug molecule for example – could be understood as points of interference between multiple relations: economic calculations, desire, biophysical processes, moral norms and research programmes. In short, the micro was not just irreducible to the macro, it could also include an equally complex constellation of elements.

In subsequent years, what was called actor-network theory continued to develop. One could break down Latour’s work, for example, into a number of stages, influenced both by movements in French thought and by Anglo-American history and sociology of science. First, there was an engagement with semiotics, manifested both in the 1981 paper and in his 1987 textbook Science in Action, which became the principal introduction to actor-network theory in the English-speaking world. Second, the 1993 text We Have Never Been Modern was strongly influenced by Shapin and Schaffer’s 1985 account of the debate between Hobbes and Boyle over the question of the relation between politics and natural philosophy in the seventeenth century and the constitution of the space of debate about matters of fact about the natural world. [15]

Third, in the 1990s, Latour’s work took an ontological turn, influenced by the philosophy of Isabelle Stengers and A.N. Whitehead. At this time, the concept of the actor-network almost disappears from view altogether, as the term ‘network’ had become, by then, too closely associated with notions of instantaneous electronic communication and the virtual society. Fourth, there was the discovery of the sociology of Gabriel Tarde, as a historical antecedent to actor-network theory.

In English, the link between actor-network theory and Tarde’s sociology is most clearly articulated in Assembling the Social: An Introduction to ActorNetwork Theory (2005). Latour resurrects the idea of the network, but understands it not so much as an analytical concept as a sign of a kind of methodological and ethical commitment. The network theorist is attentive to the empirical complexity of relations between actors, manifested in the ‘quality of the text’:the network does not designate a thing out there which would have roughly the same shape of interconnected points, much like the telephone, a freeway, a sewerage ‘network’. It is nothing more than an indicator of the quality of a text about the topics at hand. [16]

Although Assembling the Social is explicitly a work of social theory, actor-network theory was not, at least initially, understood as a contribution to social theory at all in Britain. Rather it was read, much more narrowly, as a contribution to the emerging field of science studies, which, in 1981, included the sociology of scientific knowledge, associated with the Edinburgh school of Barry Barnes17 and David Bloor, and the Marxist analyses of the Radical Science Journal. In effect, actor-network theory established itself in opposition to both these approaches, although its principal terms, ‘actor’ and ‘network’, were often misunderstood in translation. The actor was all too often equated with the individual agent, and the network with the idea of the social network. Yet it is perhaps not surprising that, at least initially, it was difficult to view actornetwork theory as a form of social theory. After all, it did all the kinds of things that social theory wasn’t supposed to do, at least in the account developed, in different ways, by Anderson, Giddens and others. For a start, it both problematized the boundary between the social sciences and the natural sciences and raised the question of the significance of the natural sciences in political life.

As David Edgerton has argued, the natural sciences have been quite central to British intellectual culture, particularly through their links to war and empire.

In The Warfare State, Edgerton is scathing in his criticism of Anderson’s thesis, arguing it is complicit with scientists’ own self-serving account of Britain as an anti-scientific culture, uncritically accepting C.P.

Snow’s account of British scientific decline. [18] Edgerton himself does not discuss the significance of a strand of thought in Britain that might include Whitehead and Gregory Bateson, for example, which sought to interrogate and reconfigure the relations between the natural and social sciences. In the British context, actor-network theory can be seen as a contribution to the extension and renewal of this heterodox tradition.

But another reason why it was difficult to read actornetwork theory as social theory, at least initially, was because of its apparent reliance on detailed empirical case studies. The importance of case studies to actor-network theory was common enough in science studies, for good reason. For sociologists of scientific knowledge wanted to distance themselves from the overgeneralized accounts of scientific method that had been typical of the philosophy of science in the second half of the twentieth century, a field that had become increasingly decoupled from any engagement with or awareness of contemporary scientific practice.

For actor-network theorists and others, science had to be understood as a set of localized practices, and the generalities that scientific practice produced were the product of local circumstances.

In the UK, a clearer engagement with actor-network theory developed in social anthropology. In Marilyn Strathern’s work, in particular, the notion of the network was brought into critical dialogue with the anthropologist’s interests in kinship relations. Drawing on her fieldwork in Papua New Guinea, Strathern had already recognized that relations should not be conceived as existing between persons, but could be taken as models for complex phenomena involving persons and things. Strathern had no problem with the idea of non-human actors that had been so problematic for some sociological readers of actor-network theory, although she differed from Latour about how one might rethink the question of scale and the problem of where networks ‘are chopped off’ through ownership. [19]

Here I do not want to dwell on Strathern’s work in detail, but note one connection between Strathern and actor-network theory. For Strathern, and more generally in British social anthropology, fieldwork can never Freee, Everyshop Window is a Soap Box, Liverpool, 2010be conceived of as the application of social theory:

the idea of social theory as a kind of metadiscipline, which cuts across the differences between the social sciences, is a mistake. In this sense, Strathern’s work is rooted in anthropological fieldwork, and is always in dialogue with the anthropological tradition. But at the same time, Strathern’s approach has been continually inventive, drawing in and reworking sources – including actor-network theory – from outside of social anthropology. The practice of theorizing is not exterior to or prior to ethnography, but should be understood as an integral part of ethnographic practice. Indeed one of the challenges of reading Strathern’s work is the way in which she weaves between ethnographic observations, comments on the social relations of the University, including such matters as research assessment and departmental management, and apparently theoretical claims. [20] From this angle of vision, the idea of a hierarchical relation between social theoretical and empirical work and between studies of particular cases and wider contexts was problematic.

So it is within social anthropology that the network initially found its disciplinary home in Britain. Contemporary anthropologists are used to the idea that theoretical invention comes through empirical engagement. The challenge that fieldwork poses for theory often emerges out of an attention to detail, which doesn’t mean any detail is relevant. Fieldwork always generates an excess of research materials. Moreover, anthropologists have had little difficulty with the idea that it is necessary to interrogate the distinction between persons and things. Some of the same considerations apply to geography, for geographers have also become increasingly concerned with the problem of the relations between the natural and social sciences, as well as rethinking the question of the significance of the particular case, as situation or event.

Physical sociology

In a lecture given as part of a series organized by the British Journal of Sociology to mark the new millennium Latour called for the formation of a physical sociology. This problem for sociology was not, as Anderson saw it, a lack of theory per se, but an on-going dialogue with the physical sciences and an interest in the ‘physical’. As Latour noted, this already existed within anthropology and geography: ‘until the advent of STS [science and technology studies], each social science was confronted by its disciplinary boundaries by the issue of what a “thing” is. Only sociology had escaped such a fate. There is a physical and a human geography and a physical and a social (or cultural) anthropology.’ [21] Actually, physical geography and anthropology are not particularly good models for the new field of physical sociology, for there is not that much dialogue between physical and social anthropologists or even between human and physical geographers. But one area where there is ongoing collaboration between human and physical geography is in relation to what might appear to be applied areas of environmental research, related to such matters as climate change and conservation. And indeed some of the most inventive work in geography has been in relation to the study of environmental policy and environmental politics.

However, in French thought the possibility of a physical sociology had already been raised in the 1890s. [22] In Suicide Durkheim had sought to demonstrate, against Tarde, that neither the study of psychology nor physical geography were at all relevant to the sociological project. For Durkheim the social was devoid of both mental and material elements. The possibility of a transdisciplinary form of social thought was precluded from the very beginning. Tarde, by contrast, had insisted on the possibility of a sociology of animals, cells and atoms as well as persons. [23] Part of the importance of actor-network theory in the British context has been to reopen the question of how to bring the study of materials into the social sciences.

In Britain, French social thought has all too often been understood simply as theory. Indeed, the view that French thought is of interest primarily because of its theoretical orientation has led to an industry of secondary commentary. But the purification of social theory cuts theory off both from the diverse traditions of thought that have informed it and from the empirical and experimental research which has been essential for its continuing vitality. Actor-network theory has played a part in something of a revitalization of empirical social research in Britain, and formed part of the basis for a rethinking of the empiricist tradition. [24] Anderson was wrong to say that Britain lacked an indigenous tradition of social thought. There were several such traditions, but their renewal was to be achieved not by the purification of social theory, but through a process which perhaps can best be described as transdisciplinary.


1. ^ Perry Anderson, ‘Components of the National Culture’, New Left Review 50, July/August 1968, p. 10.

2. ^ Anthony Giddens, Capitalism and Modern Social Theory, Cambridge University Press, Cambridge, 1971.

3. ^ Anthony Giddens, The Third Way: The Renewal of Social Democracy, Polity Press, Cambridge, 1998.

4. ^ Anderson, ‘Components of the National Culture’, p. 56.

5. ^ Gregory Elliott, Perry Anderson, Minnesota University Press, Minneapolis, 1998, p. 53.

6. ^ Thomas Osborne and Nikolas Rose, ‘In the Name of Society, or Three Theses on the History of Social Thought’, History of the Human Sciences, vol. 10, no. 3, 1997, p. 89.

7. ^ Michel Callon and Bruno Latour, ‘Unscrewing the Big Leviathan: How Actors Macro-structure Reality and How Sociologists Help Them to Do So’, in K. KnorrCetina and A.V. Cicourel, eds, Advances in Social Theory and Methodology: Towards an Integration of Microand Macro-Sociologies, Routledge & Kegan Paul, London, 1981, pp. 277–303.

8. ^ Algirdas Julien Greimas, Dictionnaire de sémiotique, Hachette, Paris, 1979.

9. ^ Michel Serres, La Traduction, Hermès III, Éditions de Minuit, Paris, 1974.

10. ^ Terry Clark, ‘Introduction’, Gabriel Tarde, On Communication and Social Influence, Chicago University Press,

Chicago, 1969, p. 68.

11. ^ Andrew Barry, ‘Pharmaceutical Matters: the Invention of Informed Materials’, Theory, Culture and Society, vol. 22, no. 1, 2005, pp. 51–69.

12. ^ Bruno Latour, We Have Never Been Modern, Harvester Wheatsheaf, Hemel Hempstead, 1993.

13. ^ Callon and Latour, ‘Unscrewing the Big Leviathan’, p. 280.

14. ^ See, for example, Michel Callon, ed., The Laws of the Market, Basil Blackwell, Oxford, 1998.

15. ^ Steven Shapin and Simon Schaffer, Leviathan and the Air-Pump: Hobbes, Boyle and the Experimental Life, Princeton University Press, Princeton NJ, 1985; Latour, We Have Never Been Modern.

16. ^ Bruno Latour, Assembling the Social: An Introduction to Actor-Network Theory, Oxford University Press, Oxford, p. 129; emphasis in original.

17. ^ E.g. Barry Barnes, Scientific Knowledge and Sociological Theory, Routledge & Kegan Paul, London, 1974.

18. ^ David Edgerton, Warfare State: Britain 1920–1970, Cambridge University Press, Cambridge 2006, p. 225.

19. ^ Marilyn Strathern, Property, Substance and Effect, Athlone, London, 1999, p. 135.

20. ^ Marilyn Strathern, The Relation: Issues in Complexity and Scale, Prickly Pear Press, Cambridge, 1995.

21. ^ Bruno Latour, ‘When Things Strike Back: A Possible Contribution of Science Studies to the Social Sciences’, British Journal of Sociology, vol. 51, no. 1, 2000, pp. 120–21.

22. ^ Eduardo Viana Vargas, Bruno Latour, Bruno Karsenti,

Frédérique Aït-Touati and Louise Salmon, ‘The Debate between Tarde and Durkheim’, Environment and Planning D: Society and Space, vol. 26, no. 5, 2008, pp. 761–77.

23. ^ Gabriel Tarde, Monadologie et sociologie, Les empêcheurs de penser en rond [1893], Paris, 1999, p. 58.

24. ^ Georgina Born, ‘On Tardean Relations: Temporality and Ethnography’, in M. Candea, ed., The Social after Gabriel Tarde: Debates and Assessments, Routledge,

London, 2010, pp. 230–47.