The material form of global cities: nodes, regions and networks

The global city refers to both a theoretical object and a distinctive urban form. Beginning in the 1970s, significant transformations altered the urban fabric of many cities around the world, producing a set of historically distinctive urban formations. Although there are variations to the specific forms that individual cities are taking, based upon their particular histories and cultures, it is a feature of global cities that they exhibit a set of core parallel material changes to their form.Footnote ³ Distinctive global city morphology takes the shape of the reinvigoration of the central business district, the taking up of a nodal position within trans-territorial global city networks, and the parallel emergence of transnational global city regions.

A number of trends are identified. Firstly, the flight from the inner city that accompanied de-industrialisation in the developed world, and also contributed to the breakdown of the urban fabric in the rioting and unrest of the late-1960s, has been reversed in many cities, their centres reinvigorated and gentrified. Such centres have grown vertically, often reaching unprecedented levels of density in their business districts. Such processes are equally at work in developing cities, such as Shanghai, where over five thousand towers over eighty stories tall have been constructed in the last twenty-five years.Footnote ⁴

This reinvigoration of the city centre is a direct result of the need for businesses to agglomerate in certain physical locations, and the need for centres of coordination and control in a technologically dispersed global economy. The gentrification of the inner city around these centres has been directly linked to the high wages available to those employed in the sectors of the economy related to such control: the financial and insurance sector, corporate headquarters and advanced producer services.Footnote ⁵ Such cities also display great polarisation of wealth, leading to segregation and privatisation, the growth of gated communities and the citadelisation of corporate space. Increasing wealth disparities, a shrinking middle class, immigration and slum production, uneven access to services, the creation of private spaces and networks, these processes pull at the cohesion of such cities and threaten their social reproduction.Footnote ⁶

At the same time as the central business district has grown vertically, certain cities are also being stretched horizontally, as information communication and transportation technologies allow them to reach across ever greater expanses of physical space. Networks of air and high-speed rail transportation have shrunk selected stretches of global and regional space. This has led to debates about the formation of global city regions – integrated and polycentric urban agglomerations of sometimes quasi-continental size, where improvements in transport infrastructures connect huge numbers of people to core cities. Such regions may be viewed as the driving force of globalisation, offering a size and scale more appropriate for the contemporary global economy than the historical city or the nation-state.

Over twenty global city regions have been identified with populations of over ten million; some arranged around a core city, such as London or Mexico City, others taking polycentric form, such as the Dutch Randstad.Footnote ⁷ Regional urban formations, such as the New Jersey-New York-Long Island-Rhode Island-Connecticut metropolitan region, the Hong Kong-Shenzhen-Canton-Macau-Zuhai-Pearl River Delta region, or the London-Paris-Lille-Brussels-Netherlands-Frankfurt-Cologne network, contain vast populations that often far surpass the size of nation-states. Manuel Castells puts it this way: ‘the entire planet is being reorganised around gigantic metropolitan nodes that absorb an increasing proportion of the urban population, itself the majority of the population of the planet’.Footnote ⁸ Driving the growth of these dense economic clusters is the heightened competition that has accompanied economic globalisation: such clustering brings with it greater operational flexibility and enhanced learning, creativity and innovation.Footnote ⁹ Such density is also a result of the great late-twentieth century migration flows, which have swelled global city regions and brought with them complex mixtures of cultural and ethnic diversity. A focus for the future is likely to be on how the levels of political integration, and the institutions of governance and representation, lag far behind the levels of economic integration.

These regions form around valued global city nodes, which are themselves further linked together by an increasingly sophisticated material infrastructure that transcends the limitations of physical contiguity. The growth in network connections between global cities is enabled by the information technology revolution: it allows distant, non-contiguous city spaces to develop patterned and durable interactions of a density and scope previously unavailable. Global cities are linked together through the creation of a selectively sited digital infrastructure: a vast project of infrastructure construction comparable to the construction of the railways, mass transit systems and motorways of previous periods.Footnote ¹⁰ The social ramifications of this technological infrastructure, which is housed predominantly within global cities, are investigated at length in Castell's notion of the ‘space-of-flows’: a new form of social space offering a qualitatively different experience of space and time for those participating within it.Footnote ¹¹ This is a subject I return to. Recent work has collected relational data and mapped empirically the types and intensities of network interactions between different sets of cities, focusing upon, for example, Internet pathways, airline routes, and global office location strategies.Footnote ¹²

Global cities and capitalist restructuring

Global city theorists have progressed a long way in developing the conceptual tools and much needed non-state-centric relational empirics that have allowed purchase on these urban developments. Before discussing their findings, however, I want to argue that their lack of familiarity and engagement with the tools of IR has narrowed their ability to draw out the wider systemic importance of their many insights. On the other hand, the state-centrism of much IR theory has blinded IR scholars to the historic importance of such developments. The changing historical relationship between cities and states has often been of epochal significance.Footnote ¹³ Cities formed the bedrock for the first ancient city-states. In the modern international system, cities were subjugated to the state, and, with industrialisation, became the growth engines of national economies. Taking a long historical perspective reveals to us that, ‘compared to cities, nation-states are “young” enterprises that have yet to prove their viability’.Footnote ¹⁴ One question I want to pose here is whether the emergence of global cities, and the new forms of social space that they represent, signals a moment of fundamental transformation in the nature of the modern international system.

To begin to answer this question I go back to the emergence of global city theory to show the nature of the intellectual problems that it was developed to solve, and trace the symbiotic development of the concept and the objects that it was designed to explain. The story begins in the late 1960s, when the urban fabric of many cities in the Western world was under increasing strain. The decade witnessed a series of breakdowns in the functioning of cities, with riots in New York and Paris, repeated around the world.Footnote ¹⁵ Important work at this time focused upon the ways in which the capitalist city produced, and continued to reproduce, patterns of poverty and inequality. Castells argued that the nation-state had destroyed the historical territorial integrity and functional viability of cities as autonomous units, leaving them at the mercy of capitalist logics.Footnote ¹⁶ Harvey noted how capitalist logics of accumulation demanded the constant restructuring of the city's built environment, through successive waves of creation and destruction. These problems, it was argued, were at the heart of the urban crises of the 1960s and early 1970s.Footnote ¹⁷

The urban crises were eventually deferred, though not resolved, with the economic restructuring of the global economy in the 1970s, which was accompanied by waves of deindustrialisation that reshaped many formerly dynamic industrial metropolises. It is at this point that the global city discourse begins to emerge, as the relationship between city and state is rescaled. The specific way in which this relationship has been rescaled was shaped by historical context. In particular, three interdependent and historically situated mechanisms have been identified that interacted during the 1960s and 1970s: crisis-led economic restructuring, the technological revolution associated with microelectronics and digitalisation, and the cultural revolutions of the late-1960s.

Castells has argued that these developments interacted reflexively to shape social practices, economic rescaling and the direction of technological development. Castells sees technology and society as mutually constituted. This perspective is often referred to as the ‘social shaping of technology’ – it refuses to view technology as an independent variable, and emphasises the role of historical context in influencing the forms that technological development takes.Footnote ¹⁸ In this sense, ‘technology is society, and society cannot be understood or represented without its technological tools’.Footnote ¹⁹ This way, the cultural proclivities of the time (cultures of individual freedom, resistance to centralisation, the post-modern sensibility) informed the neo-liberal philosophical basis of economic restructuring and the rise of the multinational corporation and networked enterprise. These new social and economic forms were, in turn, facilitated by, and reliant upon, networked technologies that allowed them the ability to be operationally flexible, less hierarchical, to have greater access to market information, and to rescale and reconstitute their activities at the global level. All three strands worked in symbiotic fashion, shaping the economic and social restructuring of the last four decades. Of course, the neo-liberal state has been central to this process by fostering the type of cultural and business environments in which these developments have emerged. By embracing neo-liberal policies it may be argued that the modern state has encouraged developments that alter the historically specific environment in which it evolved. One way to gain purchase on this issue is to examine how global cities are at the centre of this restructuring, both as centres of corporate decision making and as physical sites for the networked technological infrastructure of the global economy.

These rescaling processes began to be reflected in those physical changes to many cities that began in the 1970s and were outlined above. Analysts began to link the physical changes to structural changes in the world economy: the breakdown of the post-war Bretton Woods system and the formation of global financial markets, the emergence of offshore banking and export processing zones, the rise of multinationals and the emergence of a new international division of labour. The first clear statement making this connection between cities and economic restructuring was John Friedmann's The World City Hypothesis.Footnote ²⁰ This piece kick-started a concerted research programme in urban sociology and political geography that attempted to understand how certain core cities both shape and are shaped by the changing nature of the global economy. Friedmann's work drew upon both traditional urban theory, which examined hierarchies of national systems of cities, and world systems theory, whose neo-Marxian framework sought to show how states were tied together in unequal structural core-periphery relationships.Footnote ²¹ Friedmann's key contribution was to place cities back on the agenda of international political economy, showing how the internal life of cities, and the form that such cities take, in terms of their built environment and morphology, could only be understood by reference to their connections at the international level and the functions that they fulfil for the global economy. At the same time, the global economy could only be properly understood by reference to the role that certain cities play within it.

Such functionalism was later expanded into the idea of urban specialisation.Footnote ²² Particular cities, based upon their historical lineage, and upon particular functions that they fulfil for the global economy, become intrinsic elements of global networks. London becomes a key strategic and nodal point for processing capital, Houston a key site for the concentration of skills and expertise in the energy sector, Chicago the node for commodities exchange. In this sense, a global city develops its specialisation in relation to other global cities, and also in relation to its own particular history, resources and geographical location.Footnote ²³

Global cities and globalisation

In the 1990s these theoretical foundations became entwined with the acceleration of economic globalisation. Technological advances in information and communications developed to higher levels of sophistication, and their effects became more visible. In this context Saskia Sassen produced The Global City, which became a valuable theoretical statement of how cities are central to globalisation. Sassen's work pushes beyond Friedmann's earlier formulation in the sophistication of its analysis of the linkage between cities and economic globalisation, and in its capacity to account directly for many of the physical changes to the urban fabric. Another factor that sets Sassen's contribution apart is her emphasis upon how digital networks are producing new capabilities within global cities. She highlights how cities central to the command and control of the world economy develop significant transnational links, often expanding at the expense of the economic performance of cities within the same national space.

Under globalisation, a seemingly paradoxical trend in the economy was becoming apparent during the 1990s: the increasing spatial dispersion of economic activity around the world, as manufacturing relocated to areas of low cost labour, while, at the same time, the global economy became ever more integrated. Sassen's theory of the global city provides an explanation for this double movement of dispersal and integration. She argues that the advent of the new international division of labour, the rise of the multinational corporation, and the emergence of digital networks and new working practices have created a need for new forms of strategic command and control. The technologically enabled decentralisation of the economy undercuts the traditional controlling and organising function of the state in economic life. At the same time it opens up both a space and a need for new forms of global economic governance. This strategic function is being fulfilled by global cities. The form of decentralised decision-making they offer can match the speed and flexibility of the flows of the global economy in a way that the centralised state cannot. It is the decentralised and fragmented nature of global cities that makes them appropriate for governance in a world of flows.

Sassen argues that global cities have developed in the contemporary period in ways that go beyond the traditional role of cities as nodes in international trading and banking systems, marking global cities as a qualitatively new historical development.Footnote ²⁴ Multinational corporations, now operating global networks and assembly lines of dispersed factories, offices and sales outlets, require centralised command. At the same time, the drive towards efficiency has led to the outsourcing of key operations, making leading corporations reliant upon what Sassen terms ‘advanced producer services’, such as management consulting, legal services, public relations, accounting, financial services, design and real estate. Firms providing these services agglomerate in global city centres.Footnote ²⁵ This has reinvigorated the city as a ‘creative milieu’, offering the innovation, value creation and synergies of face-to-face contact. It is such developments that are behind the dense physical concentration and vertical growth of new state-of-the-art office construction in global city centres.

Sassen also outlines how the requirements of both firms and their wealthy elites bring low paid work for unskilled labour, drawing migrants into global cities and constructing their hybrid social character. She argues that the middle class thus becomes squeezed, and that the extreme economic polarisation that results from income disparities is given form in the social production of a type of city space where inequalities and segmentation are highly visible. This focus on cities also highlights the places and actors that construct globalisation: the highly mobile international corporate elite, but also the migrants and work cultures transforming the composition and character of global cities.

Networked and relational approaches to global cities

Where these early attempts to understand the rise of the global city were centred on its economic functions, latterly critical scholars have gone on to investigate the types of social relations that tie global cities together, the material infrastructure of these new transnational spaces, and their implications for our geographical imaginaries of the contemporary world. There arises from the work of thinkers such as Manuel Castells, Peter Taylor and Doreen Massey the sense that the meta-geographies and spatial imaginaries of modernity are being challenged by these developments, and that the emerging outline of a new spatial arrangement is slowly being revealed.

Castells work on the network society is a central text for many of those analysing global city networks. Castells has shown empirically how the material infrastructure for the new forms of social practice he describes is located and maintained primarily within global cities. Global cities do not simply benefit from the coming of post-industrial forms of economy and society: they are intrinsic to their creation and existence. Castells shows empirically how the geography of the Internet is located within global cities, where the provision of Internet content, a highly specialised activity, is overwhelmingly produced by the very global service firms that Sassen showed locate in the centre of global cities; finance firms, insurance, consulting, accounting, legal services, advertising, marketing, and the cultural and creative industries such as media, art, publishing, fashion, museums.Footnote ²⁶ Global cities are also central to the information communication revolution's technical geography. Although almost universally urban, the Internet's technological geography is extremely uneven on a global scale. The quality of technology infrastructure is constantly being upgraded, as competitive advantage accrues to those locales that have the best telecommunications infrastructure.Footnote ²⁷ Sometimes this upgrading is carried out by private firms, sometimes by metropolitan authorities, sometimes by state governments.Footnote ²⁸ And it is not just on the global or national level that inequality in the technological infrastructure of the Internet can be found: at the metropolitan scale certain districts or areas within a city may have significantly better technological infrastructure than others.

The uneven distribution of superior technological infrastructure allows the creation of selective global networks of value. This raises questions about the new patterns of inequality that are being generated: a growing ‘digital divide’ that separates the information rich from vast numbers of rural poor, the slums of mega-cities, or the disconnected neighbourhoods of global cities.Footnote ²⁹ In some ways this new pattern matches the growing wealth gap between the rich and poor countries of the world, but in other ways it is considerably more complex, because the digital divide exists not just between countries, but also within countries, and also within cities themselves. Such a situation suggests that a networked understanding of social relations cleaves closer to the new realities than the old meta-geography of a world of nation-states.

Castells work has, then, provided the basis for moving beyond Friedmann's hierarchical approach, to a networked view of global cities. Taylor's work innovates theoretically, drawing on Castells' ‘space of flows’ to take a processual view of cities.Footnote ³⁰ He argues that in the 1990s the theoretical literature on global cities had outrun its empirical foundations, with most data on cities being accumulated by states seeking to measure city attributes rather than their global relationships. Such data measured the nodes, but not their networked relations. Continuing efforts to map these networks should eventually show the rise and fall in the connectivity of various global cities.

One of the key tensions that remains in this formulation is over the question of whether global cities refer to a particular class or set of cities, or whether globalising processes effect all cities in similar ways: whether, in effect, all cities are globalising cities. There is an intrinsic ontological issue here involved in the problem of identifying a boundary that defines any particular city. It may be a mistake to try to characterise the nature of cities as bounded entities in this way. As Taylor has argued, under globalisation all cities and urban regions are responding to similar pressures, integrating parts of themselves into technological networks and circuits of value, while other parts remain switched off. Behind his work is the ontological conviction that cities should not be viewed simply as bounded places, but as ongoing processes, comprised of various flows. This ontological possibility will be picked up in the following section as one route into linking the global cities literature and IR theory.

Taylor's approach has tended to deal narrowly with the economic relations of global cities, but the way in which such places draw in material and social flows from around the globe, and extend their influence outwards, should also be a feature of a relational framework. Doreen Massey, in a recent investigation of the nature of London as a global city, argues that from the centres of global cities a ‘geography of dependencies, relations and effects’ radiates around the globe. This relational perspective clearly challenges many of the unified and essentialised notions of entities found in much of IR, and offers a profound redrawing of our geographical and geopolitical imaginations. Massey argues that

So, Massey argues, London, building on the historical legacy of the British Empire, has been a crucial site of agency for the construction of the institutional and cultural infrastructure of neo-liberalism and market deregulation. The global is thus produced and maintained within particular local nodes, which can provide an analytic bridge between the elusive global scale and specific localities. At the same time, Massey wants to argue for a relational politics of responsibility that recognises the effects of London's global reach, both on far-flung locales and on London's urban fabric.Footnote ³² She argues that if global cities are to reap the benefits from their privileged positions there must also be a consideration of their global responsibilities: their extensive environmental footprints, their impact on other economies and regions, and their responsibilities to those who come there. These issues of accountability challenge the current scalar imagination and institutions of representation, which operate at the national level.

This brief tour through key developments in the global cities literature indicates the continued evolution of the concept as it seeks to come to terms with the changing nature of global economic and social relations. I would draw out four interwoven strands of particular note from this literature that will feature in the following analysis. The first is the historically variable relationship between cities and states, which now seems to be undergoing an important shift. Secondly, the global cities literature, in line with a century old tradition of thinking about cities and Friedmann's pioneering world systems approach, has contained a strong Marxian element, which continues to be relevant. The literature points to how global cities were born of a crisis-induced restructuring of capitalism, and how they continue to produce and reproduce a logic of exclusion and inclusion that was evident in the urban riots of the 1960s, but has now been rescaled in complex ways. This logic of inclusion and exclusion is linked to, indeed facilitated by, the two other strands. The third is the logic of neo-liberalism, which, I will go on to argue, has become inscribed in the physical spaces of global cities, particularly in the continued privatisation and decline of public space. The fourth is the dependence of global cities on post-industrial forms of economic activity and the technologies that are linked with such practices. These, I will go on to argue, are related to the creation of novel spatial and temporal structures that indicate transformative tendencies. These four strands will be picked up in the following section, which uses them to discuss a number of ways in which global cities may be seen to indicate a transformation in the very nature of the modern international system, and how IR scholars may contribute to the analysis of such transformations.

Units, structures and international systems

Thinking about the role of cities within international systems demands a historically sensitive approach. The important changes to urban form represented by the global cities discourse are indicative of fundamental changes in the nature of the modern territorial state.Footnote ³⁶ An international systems framework allows analysts to clarify and conceptualise the wider impact of such change.

Such a framework has tended to begin with the analytic building blocks of ‘system’, ‘units’ and ‘structure’. At its most basic, the international system may be conceptualised as a set of interacting units organised by a structure of some kind. Despite neo-realism's limiting starting point of thinking about structures as either anarchical or hierarchical, Buzan and Little draw upon the wing of the ‘English School’ that advocated the historical comparison of international systems to argue that the historical record reveals a great diversity of units and structures.Footnote ³⁷ They identify city-states, city-leagues, various forms of empire, and nomadic tribes as some of the many viable units for international systems. Building on the legacy of Adam Watson, they also outline a variety of organising structures for international systems, arranged on a spectrum from pure independence of the units to total domination by a particular unit.Footnote ³⁸ Such a spectrum would include systems of independent states, hegemonies, suzerainties, dominions and empires. In this longue durée world historical perspective, the modern state system, with its structurally and functionally similar units, may come to seem like a unique and short run historical moment. The important question becomes: do global cities herald the re-emergence of unit diversity in the contemporary world, and thus a transformational shift of epochal significance?Footnote ³⁹

One of the core issues here, and a problem that has been central to IR in recent decades, is the relationship of units to structure. Any transformation of the international system would require a form of agency that overcomes the structural effects of anarchy that realists have long argued must inevitably shape both the form of units and their behaviour. The constructivist project in IR has primarily been an attempt to show that the anarchical structure of the modern international system need not remove the possibility of purposive change from the repertoire of state behaviour.Footnote ⁴⁰ Ruggie was the first to probe the limitations of the neo-realist conception of structure. He argued that as ‘Waltz's model of structure contains no transformational logic […] only a reproductive logic’, it cannot satisfactorily match up to the historical record, being unable, for example, to properly identify the European feudal period.Footnote ⁴¹ Other scholarship has also emphasised how the Waltzian model is undermined by its pre-theoretical attachment to the modern state, and is of little use in accounting for historical periods that contain diverse mixtures of units and institutional variations, such as city-states and empires.

One response to these critiques might be that, whatever the historical story behind its origins, the development of the modern territorial state represents an evolutionary logic (involving the centralisation of power and resources and the accumulation of the monopoly of legitimate violence) that has left other institutional forms unable to compete with it, bequeathing us a world where units have converged to be functionally and structurally alike.Footnote ⁴² Such state-centrism and inherent conservatism about possible future trajectories for the international system afflicts not just realism, but also what has emerged as the mainstream of constructivism in IR, which remains heavily state-centric. Wendt defends the state-centrism of his theory on just such grounds, arguing that his project is concerned with the big problem of regulating violence in international affairs.Footnote ⁴³ He further argues that, although there are other important actors in world politics, all significant change in the system must occur through states. It is difficult to disagree with this assessment. However, there remains the possibility that the agency of states may alter the environment in which their institutional form was historically derived, thus bringing about changes in the state itself, and the parallel emergence of new units. In recent decades the collective agency of states has been employed to pursue a project of integration for the global economy along neo-liberal principles. Among the results of this agency has been the growth of complex interdependence, experiments in pooled sovereignty, the emergence of powerful non-state actors and the privatisation of a series of previously state monopolised functions, including the privatisation of violence.

In sum, such state-centric constructivism offers significant theoretical progress in conceptualising change, but fails to push its own logic far enough because of its commitment to the state as the object of theory. Its theoretical progress is in showing how structures and units must be mutually constituted, and how it is impossible for structures to have effects separate from the interactions of agents.Footnote ⁴⁴ State-centric constructivists such as Wendt argue that states under an anarchical structure may develop among themselves different ‘cultures of anarchy’, as in their interactions with each other over time they may come to view each other as enemies, friends, or rivals.Footnote ⁴⁵ It is also possible, however, that they may embark upon collective projects that alter international systemic structure in even more radical ways, as the neo-liberal project seems to have done, and that the effects of agency may have unforeseen consequences and take on emergent logics. One result of neo-liberal policies has been the restructuring and consequent weakening of the state as the unchallenged unit across all of the domains of the international system, as new transnational structures develop. This has allowed the emergence of other units, which have begun to take on some of the modern state's functions. The global city is one such example, which fulfils the requirement for command and control nodes within a decentralised global economy, but also brings with its emergence a site and space containing new potentialities for political and social transformation.

So it would seem that in order to appreciate the significance of the emergence of global cities, the most promising avenue is that which emphasises the comparative historical sociology of international systems, in the sense that this tradition is comfortable with a diversity of units. However, even the most theoretically sophisticated attempt to conceptualise international systems in this way contains a significant weakness: despite being non-state centric, it retains an essentialist view of units. Units are viewed as entities that are sufficiently cohesive as to exhibit qualities of conscious decision-making and self-directed behaviour.Footnote ⁴⁶ However, as we have seen, this idea of unproblematic boundaries, whether we are thinking of cities or states, may be a hangover of the social science of the modern period of bounded national communities. As Colin Wight argues, such an approach commits the error of ‘methodological structuralism’: giving the powers and attributes of humans to social collectives.Footnote ⁴⁷ Recent developments in the field, however, have argued for a processual and relational perspective on units that may take us beyond this formulation.

A processual approach sees the state, and other forms of unit, as the transient solidification of the processes that have created it, which have a historical origin, and will most likely have a historical endpoint.Footnote ⁴⁸ This would apply, for example, to the process of the accumulation of the monopoly of violence, a core mechanism of state formation in the early-modern period, which may well now be receding with the privatisation of certain aspects of warfare and the rise of mafias and terrorist networks. It also applies to the continued dynamics of capital accumulation, which lie behind the emergence of the global city and the rescaling of the state-city relationship, and also to processes of territorialisation, deterritorialisation and bordering. A processual conceptualisation of the international system, which takes as the objects of analysis the processes, social transactions and relations themselves, rather than the reified units that result from such processes (‘states’, ‘cities’, etc.), is comfortable with analysing such change. Process ontology would see the state, or the city, as thus ‘more instructively and adequately understood as instantiations of certain sets of process-complexes’ than as reified objects.Footnote ⁴⁹ Just as states have formed and endured as the result of particular historical processes, so they may dissolve, or take new forms.

One of the advantages of an ontology which links form and process together is that it enables us to view different types of units, such as cities and states, in an ontologically consistent fashion.Footnote ⁵⁰ As some theorists of global cities have argued, cities are more instructively considered as sets of processes than as bounded entities.Footnote ⁵¹ Under globalisation, conceptualising states in this way also seems necessary, as the processes that gave form to them are being rescaled to create new linkages. Theorising international life in this way requires us to investigate how the components of international systems are related and organised, what the linkages are, what dynamics of the system operate to allow these relationships to persist over time, or to be transformed. Viewed in this way, the differences between global cities theorists and a new generation of IR scholars are not so great.

‘Territorial unbundling’ and global assemblages

One of the keys to theoretical progress in thinking about the transformation of the modern international system lies in its particular configuration of territoriality. Ruggie argues that ‘the modern international polity embodies a historically specific form of political space: distinct, disjoint, and mutually exclusive territorial formations’.Footnote ⁵² This is the inscription in space of the principle of sovereign territoriality. Any transformation in the nature of the modern international system would therefore require some change to this historically specific form of political space: an ‘unbundling of modern territoriality’. I want to argue here that the analysis of global cities provides concrete examples of the mechanisms by which territoriality is being rescaled in the late-modern period, thus moving Ruggie's earlier arguments forward. Global cities are the local sites through which many of the processes that construct the global scale operate, and through which trans-territorial capabilities are produced.

One way to understand the ways in which the territorial unbundling of the modern states system is proceeding is through the language of ‘assemblages’, which offers an important grip on transformation because it enables scholars to trace how the capabilities and institutions of one historical era are disassembled and reconstructed in ways that produce new capabilities and institutions. It has become particularly attractive to those that wish to add theoretical and empirical rigour to some of the debates about globalisation, which have often existed on a plane of high generality.

Assemblage theory has a sophisticated philosophical heritage in the work of Gilles Deleuze, whose ontology has been carefully reconstructed in the work of Manuel Delanda.Footnote ⁵³ Assemblages have also been approached from other perspectives to refer broadly to the ways in which the material and social elements of systems are combined to form historically particular hybrids of technical and administrative practices, capabilities and configurations of space. Different historical periods will contain different mixtures and arrangements of such elements. Recent efforts in this direction have focused upon the reconfiguration of the elements of the territorial nation-state into new global assemblages. I want to focus here upon two emblematic figures: Bruno Latour and Saskia Sassen.

Latour's contribution, variously described as actor-network-theory or the ‘sociology of associations’, first emerged from studying the sociology of technology. Latour's sociology, like the relational/processual theorists, seeks to circumvent the reification of social groupings as stable and permanent entities. He argues that social theory has forgotten the original etymology of the word ‘social’, which was about how things are connected or assembled together. Thus, national societies are not eternally stable entities: the ‘units’ that IR theorists in the systems tradition speak of are simply shorthand for something that has been assembled from networks of people, ideas and material objects at a particular historical moment. And, unlike the social constructivists, Latour argues that the presence of material objects within these assemblages, such as technology or architecture, is essential to allow such collective entities to endure across time and space.Footnote ⁵⁴

Where traditional social theory starts with society and uses it to explain certain phenomena, Latour traces the connections and processes that form the social and maintain it, or reform it. He argues that in times of relative stability it might be possible to get away with reifying a grouping such as the nation-state. But in times of change, when boundaries are shifting quickly and the sense of belonging to a particular group is in crisis, when new groupings of ideas, people, and material objects are being formed, it will not suffice. The tension between the hitherto relatively stable groupings of nation-states, and the emergence of global scale capabilities and institutions, represents a point at which it is useful to look for new assemblages of people, ideas and material objects. Global cities, and the technological networks that are connecting them, offer just such an opportunity to study the fraying boundaries of territorial nation-states and the larger Westphalian international system that they form. The empirical work of global cities scholars is a key resource in any attempt to map out the material and ideational networks that are unbundling modern territoriality and replacing it with something distinctively post-modern.

Sassen's recent contribution to the debates about the rescaling of territorial formations presents a case for understanding the epochal significance of ‘globalisation’ through a comparative analysis of different historical assemblages of territory, authority and rights.Footnote ⁵⁵ Her argument is that the modern nation-state embodies a set of assumptions about how to configure these three elements, but that the contemporary period is witnessing the renegotiation of this relationship. The unbundling of modern territoriality is dislodging and rescaling national capabilities. Ruggie's earlier arguments moved along parallel lines, particularly with his assessment of how medieval trade fairs, although encouraged by local feudal lords because of the tax revenue they generated, eventually undermined feudal rule by allowing the development of new forms of thinking about authority and rights, ideas later transferred to the growing towns and cities. The rise of global cities and their relationship to the problem of extra-territoriality seems strikingly analogous.

Sassen's centrality to the evolution of the global cities debates is important in this move into a broader ‘global sociology’: global cities are a vital element in this rescaling of the national territorial state-system. As we have seen, the agency of the state is directly related to the rise of global cities. The neo-liberal state is implicated in this renegotiation of territorial scale through the augmentation of global city capabilities, both in the logical outcomes of its political philosophy, and as a major investor in global city infrastructure. Graham and Marvin have outlined how the liberalisation and privatisation of the markets for networked infrastructures has led to the ‘splintering’ of both urban and national space.Footnote ⁵⁶ The old social democratic concept of universal access to national infrastructure has been superseded by privatised networks: premium spaces and infrastructures tailored to the needs of specific users. Such networks are not, as in the past, tying the nation-state together, but are instead tying different pieces of global space together. Infrastructure networks can simultaneously be ‘unbundled’ locally whilst being integrated internationally. This reversal of over a century of national integration fundamentally challenges the modern notion that a ‘city’ or ‘nation’ necessarily has territorial coherence in its own right as a spatial container for economic activity that is somehow ‘naturally’ separate from surrounding spaces.

Globalisation does not signal state decline, but rather the complex rescaling, reorganisation and re-territorialisation of the state. Global city formation and state rescaling are ‘dialectically intertwined moments of a single dynamic of global capitalist restructuring’.Footnote ⁵⁷ States recognise that investing in and upgrading global city infrastructure can work to attract global capital flows into their territory, engaging in ‘supply-side entrepreneurialism’. At the same time, such states can no longer contain the process of capital accumulation and urbanisation within their borders. They become both complicit in this rescaling, but also increasingly constrained by it, while global cities and their regions emerge as the fundamental territorial capitalist infrastructure of economic globalisation.

Such developments point to a fundamental mismatch between the territorial state and the global economy, the contradictions of which have become apparent in the current financial crisis, as global financial practices have outrun institutional regulatory capacities. This tension between global scale capitalist logics and the scale of the representative institutions of the nation-state had heretofore been partially addressed by the emergence of regimes of private authority designed to cope with specific economic developments at the global level. Sassen charts how private authority, in a move to fill the global economic governance responsibilities left vacant by the nation-state, has created a ‘global web of bordered spaces’, consisting of different specialised self-regulatory regimes (commercial arbitration, for example) exhibiting autonomy from national law. By disembedding selective elements that were previously located at the national level, non-state economic actors have knitted together a distinct space that ‘assembles bits of national territory, authority and rights into new types of specialised and highly particularized fields’ that ‘destabilise conventional understandings of national borders’.Footnote ⁵⁸

The dematerialisation of national territoriality requires the construction and maintenance of a ‘state-of-the-art built environment’, which, as we have seen, forms at the core of global city nodes. The splintering of the national spaces of modernity has the effect of redefining the context of these valued physical sites. Thus, Sassen argues,

There are two juxtaposed implications of this type of development. Firstly, it illustrates how the global scale is produced at the local scale, and how such production works to fragment and disrupt national space. At the same time, it shows how the type of political and economic regime that has backed the creation of global cities is endangering the reproduction of the city itself – the issues of the 1960s have been deferred but, it seems, not solved, and now the national polity is itself under question. Investment in developing global city infrastructure is aimed at tying valued privatised global spatial fragments together. Less valued physical spaces of such cities may decay, diminishing the prospects of social justice and inscribing systemic contradiction into the morphology of social space, as we see with the rise of mega-cities, slums and deprived metropolitan districts.

The spatial and temporal structures of international systems

The disruptions to the territorial frameworks of modernity wrought by global cities also offer the opportunity to compare the spatial and temporal structures of the modern international system with an immanent set of new structures. The historical specificity of spatial and temporal structures was well understood by Martin Wight when he famously argued that the space inside the state enclosed the boundary of political progress in time, marking off the space between states as an apolitical no-mans land of repetition and recurrence.Footnote ⁶⁰ Rob Walker outlined how mistaken it is to think that this peculiarly modern arrangement of space and time in the international system is a transhistorical fact.Footnote ⁶¹ Changing spatial and temporal structures are thus a dimension of international transformation.

Ruggie's work represents a rare attempt to think through signs of immanent transformation in the temporal and spatial structures of modernity. He shows how these structures are social in character, and how the key to manipulating them lies in the emergence of new forms of culture, and the technologies that reflect them. He examines medieval European perceptions of space and territoriality as a benchmark against which to judge the possible movement to a post-modern form of spatial organisation in the contemporary international system. Ruggie sees the feudal system in Europe as one without fixed borders and territorial boundaries, where authority was exercised over individuals rather than territory. He argues that the emergence of single-point perspective in the European Renaissance enabled the development of an individual, fixed, subjective point of view that opened the imagination to new ideas about social space and political community.Footnote ⁶² This period also saw the emergence of a modern appreciation of temporality and contingency, as the eternal certainties of a divinely ordered universe began to dissolve.

In many ways, Ruggie's emphasis upon the social character of space echoes Henri Lefebvre's earlier philosophical investigations into the subject. For Lefebvre, space is not simply an external environment, but a social product, the creation of human agency. Lefebvre's strategy was to link space and language, and to show how political struggle is carried out and inscribed in space. With much the same motivation as Latour, Lefebvre reconnects ideas, ideologies and discourse to the material world, which is seen as essential to their durability across time and space. He asks,

Seen in this way, space becomes a product of different historically situated societies. Every historical period produces its own ‘spatial code’: a common language, which can then be read or decoded by the analyst. Thus, the European world of the Renaissance and Scientific Revolution was characterised by a spatial code that included classical perspective and Euclidean geometry: hence, Ruggie's argument that these mental representations of space were a crucial pre-condition for the political space of the bounded territorial state. The formation and dissolution of spatial codes offers a way to understand historical change:

Lefebvre argues that there is a history of social space inscribed in the physical world. To understand historical transition points we should look to see how the particular spatial code of a given historical society undergoes crisis and collapse. Lefebvre argues that all periods of revolutionary social upheaval are accompanied by the pre-conditions for a new form of space. This observation lies at the heart of my argument about global cities and their associated transnational spaces – global cities represent the inscription in space of a certain set of historically contextualised political, social and economic practices.

In looking for the possible emergence of new spatial structures in the contemporary world, Ruggie was drawn to the global economy. Following on from his single-point perspective insight, Ruggie looks for signs of the emergence of a multi-dimensional perspective, where actors might hold new understandings of space as non-distinct and overlapping. He finds it in the logic of deregulated transnational economic relations, and the developments in information, communications and transport technologies that integrate transnational economic spaces. The global market, despite its activities taking place at different physical locations around the planet, exists in functional terms as if its elements were operating in the same place:

For an understanding of the related temporal structures of the international system, Ruggie leans on the Annales historian Fernand Braudel for conceptual tools. He links Braudel's notion of la longue durée, the deep epochal structural layer of time, to systemic transformation. For Ruggie, epochal systemic transformation can be expected to occur only when the temporal structure that underpins modernity ‘dissipates’.Footnote ⁶⁶ One way in which a possible break in the temporal structure of modernity has been characterised is in the movement from sequence to simultaneity.

The trend towards simultaneity has long been a staple of theorists of post-modernity and globalisation. Between them, they have drawn out the relationship between multinational capitalism and space-time compression, where the pursuit of more efficient markets for capital drives technological developments towards the elimination of sequential time.Footnote ⁶⁷ The capitalist restructuring of the 1970s, it has been noted, was dependent upon information technology, as the basis of the production of value moved from manufacturing to the creation and manipulation of information, images and signs.Footnote ⁶⁸ Post-modernism, as a cultural and aesthetic expression, can be directly linked to this post-industrial phase of capitalism, which signalled both the continuity of capitalism as an economic form, and an experiential and cultural rupture its organisation and perception. The culture of post-modernism has been argued to perform the task of coordinating the practices and ‘structures of feeling’ of individuals with the new requirements of flexible, fragmented and decentralised forms of economic production and organisation, and with a new globalised division of labour. Post-modern culture creates post-modern people, adapted to the functional requirements of the socio-economic environment of late-capitalist societies, who begin to experience space and time in qualitatively novel ways.Footnote ⁶⁹

As outlined earlier, the non-substitutable technological infrastructure that underpins this electronically mediated digital environment of signs and symbols is located within global cities and their networks. In this sense, global cities provide an infrastructure of simultaneity, and the material supports for a qualitatively new form of dematerialised social space. Within this environment ‘the contiguity of space is separated from the simultaneity of time’. For Castells, this ‘space of flows’ is beginning to dominate physical spaces, which do not, of course, disappear, but now receive their meaning from their functioning as nodal points within specific networks. Castells argues that the most important social functions and processes are now organised around networking logics. Financial markets, global governance, transnational production, social movements, for example, are all seen to be organised through the space of flows. Such logic also gives rise to those that reject it. Resistance to the power of the space of flows often centres on traditional forms of identity: the rise of religious fundamentalism, the return to the local, the reinvigorated appeal of nationalism and ethnicity, may all be seen as expressions of resistance by those excluded from the networks of social power.

Where these forms of identity look to the past for their appeal, Castells also identifies a new culture emerging for those within the ‘network society’, which he calls ‘real virtuality’. It results from people sharing the new forms of space and time made available by technology. This is a world constructed from ideas, a world of imaginary representations stored and communicated through computer memory and networks. The social exclusiveness of these networks of power and value exacerbates existing inequalities, which appear both within states and between states. Castells describes black holes of poverty that exist just blocks away from areas of cities that are home to firms centrally involved in global flows of power. Members of these firms are likely to identify, culturally, more closely with others located in similar parts of cities across the globe: within the space of flows forms a novel type of transnational identity. Castells also identifies how the technological creation of a space of flows enables the temporal structure of simultaneity that he calls ‘timeless time’. Whereas the culture of modernity has been strongly informed by the notion of progress in time, in the network society temporal sequence is first compressed, then blurred, and, eventually, dissolved. For those privileged actors within the space of flows, ‘being cancels becoming’, while in the ‘multiple space of places, fragmented and disconnected […] devalued activities and subordinate people endure life as time goes by’.Footnote ⁷⁰ The ways in which time is sequenced becomes a battleground for political projects, just as the construction and reconstruction of social space does for Lefebvre.

In this sense, Castells formulation offers something qualitatively new from those earlier core-periphery models that have influenced IR.Footnote ⁷¹ Technology has created a qualitatively new transnational social space, where disconnection and connection are expressed through physically non-contiguous networks that operate at different scales. Such developments crosscut the boundaries of the territorial state, which, under modernity, were seen as containing the progressive potentials of domestic politics. These networks rest upon the privatisation and citadelisation of corporate space, and the further erosion of the public sphere. In this way, the systemic issues that have attended the neo-liberal project of the last three decades have become written into the material fabric of global cities. Such tensions, Appadurai has noted, indicate how many features of global politics have become ‘telescoped’ into the contemporary global city: an ‘implosion of global and national politics into the urban world’.Footnote ⁷² Global cities become key sites of political contestation, amplifying both systemic contradictions and historical possibilities. All of the many contradictions that globalisation theorists have discovered in the dialectical interplay of global and local are present in the global city today: the great disparities of wealth and poverty, the great heterogeneous mixing of cultures, ethnicities and diasporic diversity. Global cities are thus exhibiting nascent temporal and spatial structures with logics that are in fundamental contradiction to the bounded spaces and linear temporalities of modernity.