Few theories have left their mark on urban studies to the extent that Actor-Network Theory (ANT) has in the last few decades. Its background in Science and Technology Studies (STS), its critique of the explanatory value of such abstractions as ‘class’ and ‘society’ and its efforts to transcend society/nature and local/global binarisms inevitably challenged conventional views on cities, urbanization and urban phenomena.Footnote 1 Economic and Marxist approaches to the city in particular have been challenged, at least to the extent that they invoke the explanatory force of the economy or capitalism as a global social system and, thus, fall back upon the binarisms under attack from ANT. The network approach questioned architectonic explanatory models (substructure vs. superstructure) and deepened our understanding of actors and agency (both emerging from networks of humans and non-humans). However, ANT has always been subject to criticism too. Because it placed the non-human on an equal footing with the human, ANT is often considered indifferent to issues of inequality and injustice, negligent of the forces of capital and capitalism (including the way they transform or destroy nature) and immodest and over-inclusive (due to the claim that nothing exists outside the networks).Footnote 2 Perhaps most disturbing of all was and is the perceived loss of explanatory force. Critics have rightfully wondered whether ANT allows for explanation at all, given the refusal to distinguish between ‘determining’ actors and processes and ‘determined’ phenomena.Footnote 3 Up to a certain degree, these criticisms explain why the concepts and approaches involved hardly surface at all in recent important syntheses of urban history.Footnote 4 Historians studying cities prior to the nineteenth and twentieth centuries are especially reluctant to use concepts of STS and ANT, partly because of ANT's perceived connection to technological and environmental issues.Footnote 5 In my view, however, the introduction of ANT can be rewarding for urban history even beyond the niches of nineteenth- and twentieth-century history of infrastructure and environmental issues. The concepts and approaches of Bruno Latour, Michel Callon, John Law and others cannot, in any case, remain absent in discussions about urban agency, the urban variable and the very usefulness of urban history as a separate sub-discipline.
In this contribution to the roundtable, I will therefore argue in favour of a more hospitable attitude towards ANT and related approaches. Taking my cue from recent attempts to transcend sterile oppositions such as the one between orthodox interpretations of Marxism and strong versions of ANT,Footnote 6 I will argue that ANT continues to provide a useful set of concepts and ideas to tackle not only the reflexive question of urban agency but also important themes in urban history as such. To that end, I will first elaborate upon the question of urban agency so as to identity the problems with current conceptual approaches. Subsequently, I will touch upon some of the recent attempts in urban history to introduce new concepts – including ANT and the broader Foucauldian approach to power of which it is part.
Urban agency as a historical problem
Most scholars in the broader field of urban studies, urban historians included, start from the assumption that cities have generated specific forms of thought and action. Imbued in good part by the pervasive dichotomy between Gemeinschaft and Gesellschaft, early twentieth-century theorists of the urban such as Georg Simmel, Robert Park and Louis Wirth from the so-called Chicago (or ecological) School of Sociology argued that the city as a spatial and material environment produces specific values, practices and mental states. Ever since, urbanism is considered to have had an impact on social relationships, political culture, criminality, taste and consumption, leisure, the production of knowledge and culture, economic innovation, the perception of environmental problems, etc.Footnote 7 However, is agency not something one should attribute to people, or at least to living beings, rather than to something as abstract and spatial as a city? Are not cities themselves the result of agency, in other words of immigration, entrepreneurial activity and the aspirations of political and intellectual elites?
The question of urban agency naturally brings with it the question of the nature and definition of a city. Many places have urban characteristics without necessarily being cities and vice versa. Is Silicon Valley a city? Where, in the absence of city walls, does the urban stop and the countryside begin? What about slums and squatter camps? Are they part of a city or do they rather resemble refugee camps? In much contemporary sociological and economic writing the term city is interchangeable with such concepts as metropolitan areas, economic agglomerations and the like. The adjective ‘urban’ can often just as well be replaced with ‘densely populated’. The issue of urban agency therefore involves isolating the urban as a specific variable. It involves finding out whether cities are more than the sum of their parts, and if so, why and how.
Economists, economic geographers and economic historians implicitly or explicitly define the city as the absence of distance. What matters for them is proximity. Proximity and spatial clustering produce economies of agglomeration through the presence of customers and suppliers of raw materials, the elastic supply of skills and specialized workers, the sharing of specialized inputs and infrastructure (including institutions), the reduction of search and information costs and – last but not least – knowledge externalities and innovation.Footnote 8 From here, we could go a long way towards explaining the growth and dynamics of cities. After all, successful cities attract immigrants of all sorts and migration in turn creates the clustering of different ideas and the proximity of a diverse set of cultural frames of reference that yield new combinations and, hence, innovation. Moreover, institutional thickness and path dependency come into play. Spatial clustering creates distinctive economic trajectories, but also durable patterns of specialization and competitiveness between countries and regions. What is significant about all this, though, is that it is spatial proximity and not urbanity per se that accounts for economic dynamism; the city dissolves into the background.
Nor is this the only way in which agency is attributed to cities. In the critical tradition of urban studies, the agency of cities is defined entirely differently. For the neo-Marxist Henri Lefebvre, every mode of production produces its own type of space, which, in turn, serves as a tool of class control – or at least produces specific types of thought and action.Footnote 9 For social geographers such as David Harvey and Erik Swyngedouw, the analysis of the city is part of the analysis of capitalism and class struggle.Footnote 10 On the one hand, the city is itself seen as the result of the forces of production and capital accumulation, which explain the physical form and layout of the cities, including spatial differentiation and segregation. On the other hand, cities help reproduce these capitalistic and neo-liberal forms of production and hegemony, to the extent that nature as well is now seen as both largely ‘urbanized’ – i.e. formed, marked and defined by the urban – and in the process of being commodified.Footnote 11
When push comes to shove, then, urbanization and the history of cities are themselves part of other, more encompassing or more fundamental, transformations. While capitalism and capital accumulation are the ultimate source of explanation in the Marxist tradition, urbanization and the development of cities were part of a process of modernization in the Chicago School of Sociology. Rather than agency emerging out of the urban as something aboriginal, cities thus convey agency produced elsewhere. They first materialize and then convey the agency of money, capital or rationalization. In this vein, urban studies and urban history are constantly at risk of reproducing narratives of modernity. While the very advent of European urbanization in the high Middle Ages has been connected to the emergence of such fundamental western values as market exchange and political representation,Footnote 12 ideas of urban development have been intimately connected to the history of capitalism and liberalism since the nineteenth century.Footnote 13
This is all the more problematic since all such approaches are fundamentally Eurocentric, as William Rowe's essay suggests. The history of European urbanization and urbanity is fundamentally intertwined with the master narrative of European modernity, whether in the Marxist or Weberian sense of the term.Footnote 14 Post-colonial criticisms about the use of abstractions such as ‘class’, ‘capital’ and ‘public sphere’ notwithstanding, we still face the risk of mistaking these concepts for explanations. In urban history, this problem is related to the fact that cities are typically addressed as bounded entities. From Walter Christaller's central place theory to recent ideas of the ‘dual’ or ‘divided city’,Footnote 15 the city is conceived of as being a political-economic actor with a kind of essence or core.Footnote 16 Something similar often applies to approaches in which the urban is seen as a ‘node’ in (global) networks or as one ‘scale’ among others (i.e. the local, the national and the global).
In the last instance, the urban is associated with ‘civilization’ through the politically charged adjective civic or civil which is derived from the Latin civis, and hence connected to an urban citizen. The intellectual roots can be traced back to Aristotle, for whom the city had a purpose or telos which was to foster happiness and virtue. Moreover, the urban and the rational were intimately connected. In the theories of Aristotle and medieval followers such as Thomas Aquinas, only people convened in the political partnership of the polis acted reasonably, i.e. in favour of the common good. While on the one hand human beings were seen as possessing a natural drive to form a political community, being involved in the urban commune or polis was considered a precondition for becoming a complete (rational and virtuous) human being.
Rewriting the history of European urbanization and urbanity inevitably entails, then, coming to terms with theories that have challenged European and western master narratives.Footnote 17 In the following, I will build upon recent historiographical transformations to re-examine critically the agency of cities.Footnote 18 Specifically, I will present some tools to re-analyse how urbanization is related to such broader societal transformations as the advent of capitalism, individualism and liberalism, while remaining critical about narratives of modernity.
From hegemony to governmentality and the city as an assemblage
In recent decades, critical social theory has experienced a shift from a Marxist to a Foucauldian approach (or in other terms, from hegemony to governmentality).Footnote 19 In a Marxist sense, power relations resulted from, on the one hand, modes of production and class relations and, on the other, ideology and hegemonic institutions. Urban historians translated this into economic forces – production, distribution and exchange, patterns of landholding and labour discipline – and the socio-political domination of urban space. In Foucault's writings, power loses its two-tier architecture (base versus superstructure). Foucault looks for the technologies and the microphysics of power, which are related less to the relations of production in a socio-economic sense than to the production of knowledge and ‘truth’. For Foucault, power is intimately related to the legitimacy of discursive practices. The latter include practices in such institutions as hospitals, asylums, jails and schools, in which material conditions such as the panopticon instantiate specific ‘truth regimes’ and vice versa.
Foucault's concepts have profoundly shaped recent views on the relationship between urban history and liberalism. Central to this endeavour is the concept of ‘governmentality’, which a number of his students and followers have singled out as central to his later thinking.Footnote 20 Elaborating upon some of his courses at the Collège de France,Footnote 21 they define gouvernementalité as the ‘art of government’ or the ‘rationality of government’, as such referring to both the techniques of governance and the ideas and mentalities implied in them. While governmentality is generally referred to as the ‘conduct of conduct’ or the ‘power to act on the action of others’, it can also be seen as a sort of ‘governing at a distance’, i.e. governing through institutions, through the organization of space and the market, through civil society and communities and last, but not least, through individuals governing themselves.
Building on these ideas, Patrick Joyce has argued in his eye-opening book The Rule of Freedom that freedom itself was a technique (rather than the absence) of rule in liberal governmentality.Footnote 22 According to Joyce, liberal governance in cities like Manchester and London aimed at the production of self-governing subjects. Urban authorities did not rule and discipline citizens in a narrow sense, rather they transformed the material form of the city in a manner that encouraged the production of the free, self-ruling citizen. In this vein, the arrival of running water in the home and the capacity to defecate in private is seen simultaneously to have generated the hygenization of the city and the creation of the moral self-regulating subject. Starting from the Foucauldian idea that liberalism as an art of governance was based on the cultivation of an autonomous and reflexive self, Joyce thus considers freedom as something that was ruled through.
In a similar fashion, Chris Otter examines the use of vision and light as part of the history of freedom and governance in his book The Victorian Eye.Footnote 23 Otter starts from a critique of the argument that the ‘expansion of illumination’ is, in western narratives, related to the rise of modernity and the disenchantment of the world. Instead, he sees the expansion of gas and electricity as part of the development of a ‘technological state’ in which the nature of governance changed along with technological infrastructure. In the nineteenth century, the liberal, self-governing subject was thus formed through the administration and agency of the material environment. The introduction of such material devices as plate glass and light switches transformed power and the self-regulation of individuals and groups. Examples include the mutual monitoring of groups within nineteenth-century libraries, museums, reading rooms and art galleries, as well as the encouragement of self-inspection and introspection in private and enclosed spaces equipped with lamps and mirrors.
Stressing how urban technologies, from water to lighting, fundamentally altered the terms on which self and society were activated in the nineteenth-century city, Otter brings to the fore another set of concepts, in which the material and the technical are even more important. While Foucault's concepts and methods have often been reduced to the analysis of discourses, and hence associated with post-structuralism and a concomitant neglect of materiality, they have also spawned a more radical cluster of theories focused on the material environment, known as Science and Technology Studies (STS) or – more specifically – Actor-Network Theory (ANT). Among other things, this body of theory has been critical of the use of abstractions such as ‘capital’, ‘class’ and ‘society’ (to which could also be added the ‘urban’) in the explanation of virtually all economic, social, cultural and political phenomena. One of the chief spokespersons of ANT, Bruno Latour, has insisted that ‘social structures’ have no explanatory power but should instead be explained themselves. While attacking post-structuralist thinking for its neglect of objects and technologies, Latour radically refuses to search for a reality ‘behind’ or ‘underneath’ objects and events and takes the ‘real stuff’ as his point of departure.
The ‘social’ or ‘structures’ are replaced with networks, which are not seen as lying beneath reality, or acting behind the scene. Networks are rather to be taken literally, as a set of real connections and associations. Moreover, the construction of networks is as much a process of non-humans as it is one of humans. ANT is often described as a sociology of association or translation in which both human and non-human actors are enrolled in the network (association) and are defined and formed by the network (translation).Footnote 24 While non-human actors (or actants) thus have an agency too, the social is formed in processes in which mediators (both human and non-human) are connected and networks are formed. As such, networks are not conveyors of causality generated elsewhere but concatenations of mediators which make other mediators do things. What ANT thus emphasizes is that it is not simply people who produce and use material artefacts or attribute meaning to them, but that the relationship between people and things is not one-way or even two-way; their significance lies in the way they work together as networks and assemblages. While material elements are what make social relations asymmetrical and durable, it is the network that produces forms of agency and identity.Footnote 25
The rationale of STS and ANT
Apart from historians like Otter and Joyce studying nineteenth- and twentieth-century cities in England and its (former) colonies, urban historians have been rather reluctant to take up these views. Undoubtedly, this is due in part to the strong influence of Marxism in urban history, some protagonists of which have presented STS and ANT as relativistic and nihilistic – i.e. part of rather than critical to capitalism and neo-liberalism.Footnote 26 But even more important may be the positivist bent among historians. Most historians look for causal mechanisms in the tradition of the social sciences in a broader sense and are thus urged to distinguish between what is to be explained and what explains. Thus, class, structures or demographic, economic, or ecological conditions explain opinions, individual choices, events or – for that matter – urban development. This is difficult when starting from ANT, owing to the radical refusal to distinguish between what is on the surface (to be explained) and what is underneath or behind it (explaining it).
However, STS and ANT may be relevant to historians for different reasons. Notwithstanding the often abstract jargon, the concepts of Latour and his accomplices may be appropriate to meet historians’ proclivity to detailed description. Latour's work eventuates in increased realism rather than in relativism, and the social (and the urban) is again given its ‘missing mass’ in a rather literal sense.Footnote 27 Starting from Latour's concepts, such multifaceted phenomena as urban markets, for example, cease to be either abstract places of economic exchange or sites of cultural ritual. They are rather seen as assemblages of physical sites, sources of information, rules and institutions, chains of supply and demand of commodities, and human actors such as traders, buyers and local officials. This perspective enables us to see both urban places and the processes that go on within them as socio-technical formations without simply privileging one domain (e.g. human organization and decision-making) or another (e.g. technological determinism).Footnote 28 It allows us to think of cities as hybrid places, complex assemblages of material and human components, in dynamic and by no means always predictable or stable relations with one another.Footnote 29
Nor does ANT eclipse power relations and ideologies. As with Foucault, power is related here to the construction of knowledge. The roots of ANT are to be found in the sociology of science, in which the construction of ‘matters of fact’ has been the central research focus for over a generation now.Footnote 30 Rather than asserting the ‘objectivity’ of matter and advocating a technological determinism or, alternatively, reducing everything to discourse and ‘construction’, Latour has tried to go beyond the (Kantian) idea that reality and objectivity are related to the physical world while perception, construction and subjectivity are to be found solely in the human mind. Refusing the position that ‘it is all in the mind’ or, alternatively, that things have an inner essence beyond human perception, he has argued that a ‘thing’ is defined by its alliances. While for traditional realists things become more real the less they are connected to the human mind, for Latour, things become real when they are enrolled and become connected to ‘allies’. They become real, in other words, thanks to their associations, and the more (human and non-human) associations, the more real they are.Footnote 31 This is clearly true for climate change and quarks, which exist thanks to the attention, rules and practices of scientists, next to a range of equipment for measurement and visualization. But looking at it historically it may indeed also apply to, say, waste and electricity, human degeneration and hysteria, civil society and, indeed, democracy and capitalism, which can all be understood as more or less durable networks of humans and non-humans.Footnote 32
The challenge therefore for urban historians (no less than for other social scientists) is to stop searching for the ‘matters of fact’ behind the ideological or discursive screen, and instead make visible the mechanisms with which matters of fact and states of affairs are constituted and maintained. This is what Michel Callon did when arguing that a modern economy is embedded not just in social relations, but also in economics and tools and programmes based on economic theory.Footnote 33 In a similar vein, in his Rule of Experts Timothy Mitchell tries to move beyond the opposition between, on the one hand, a view of modernity and global capitalism as the result of the historical unfolding of an inner logic and, on the other hand, a social constructivist approach, in which the ingredients of modernity and global capitalism – e.g. the nation, the modern consumer – are but inventions of discourses and/or the social sciences involved. For Mitchell, the emergence of the economy in its modern shape and meaning is not solely the result of language and the social imaginary. While the economy can be seen as being put in place partly by a ‘new politics of calculation’ (and hence of statistics, measuring devices, etc.) its birth involved a set of real political shifts, including the realignment of state, economy and colonial power. In the process, the making of a modern economy (which Mitchell dates from the 1930s) re-inscribed a series of binary oppositions: economy and society, the material and the ideological, the human and the non-human.Footnote 34
These binarisms are to be seen as produced by history, rather than as natural phenomena, which is a strong reminder that scientists – contemporary as well as present-day – are involved in the networks concerned as well. This is how STS and ANT eventually enable narratives of modernity and Eurocentric approaches to be tackled. For Latour, modernity is a historical process through which philosophers, scientists and others have sought to separate nature from culture (science) and object from subject in a ceaseless – but ultimately impossible – effort of ‘purification’. At the heart of modernity is the myth of Cartesian and Kantian dualism in which humans are seen as distinct from nature, although in reality they are always linked through the proliferation of ‘hybrids’.Footnote 35 In a very paradoxical way, modernity has bred the attempt to study nature objectively along with an ever-expanding range of technologies and matters-of-fact – of which diseases and environmental threats are but the most obvious ones – in which nature and culture are fundamentally intertwined.Footnote 36 While historians have only begun to understand the potential impact of these ideas, urban historians may grasp the opportunity this offers to tackle the modern European city as an ensemble made durable by the combined effects of matter, technologies and ideas.
ANT as the natural ally of history
In this vein, STS and ANT are not at odds with the so-called critical tradition. As is the case with cities, capital and capitalism can perfectly well be seen as assemblages that assume different forms through history. STS and ANT enable us to understand the development of capitalism as a material and cultural process simultaneously. They imply a refusal to reduce capitalism to either something which unfolds naturally through time or something which is constructed in discourses and ideologies first. Rather than something natural or something constructed in the realm of culture, the development of capitalism could then be seen as something layered, complex, multifaceted and radically contingent, the development of which involves material, technical, juridical, political, discursive and imaginary elements. It would be addressed, in short, as something profoundly historical.
In addition, the connection between capitalism and urbanization could be disentangled in more detail and without having to attribute explanatory value to one or both phenomena exclusively, and without excluding other factors. One such factor is the amalgam of knowledge and technology, which through the lens of STS and ANT could be addressed more adequately too. Anyone who wants to stress the importance of technology in history runs the risk of technological determinism, or of seeing technology itself as an effect of other ‘agencies’ such as rational choices, capital accumulation, discourses, etc. In reality, knowledge and technology too are radically contingent and profoundly historical, entangled as they are with myriad material and cultural elements. Addressing cities as socio-technical sites means examining the relationship of technology with other historical processes (including capitalism, knowledge formation, etc.), while allowing for the view that technology is specific in its accumulative and irreversible nature.
More examples could be added, but the history of capitalism and the history of technology are at present central to any new understanding of the ‘nature of cities’. Cities are actually and constantly produced in a process in which nature is transformed into culture – i.e. in a city as a material reality, whether via the transformation of raw materials into buildings and infrastructure or through urban dwellers consuming imported foodstuffs and energy. These processes are profoundly shaped by both capitalist modes of production and technological developments – without these elements being separable from processes of urbanization. It is the very entanglement of the urban and urbanization with capitalist modes of production which produces and reproduces the distinction between the urban and nature in the first place. This very distinction is radically contingent and historical too, and should be a central tenet of urban history for decades to come. Seriously incorporating the STS and ANT views in our work will help us to understand both these processes and our own position and involvement in them. In my view, it would lead to a more nuanced and more reflexive approach, without losing any critical potential – indeed, quite the contrary.
