GaWC Research Bulletin 352

(A)

Flow-Place: Reflections on Cities, Commutation and Urban Production Processes

L. Halbert and J. Rutherford^*

Abstract

This paper contributes to ongoing debates around the nature, composition and production of cities in an era of increasing globalisation. Arguing that we need to move beyond relational urban conceptualisations, we sketch out a view of cities as dynamic and unstable flow-places in which the respective qualities of flows and places cannot be meaningfully disarticulated. We use the notion of ‘commutation' to capture this idea of cities as a communicational nexus in which multiple actors perform, manage, undergo and contest a variety of socio-technical operations which interconnect flows and places. Three constitutive components and implications of this commutation process are identified through which the socio-technical practices and outcomes of urban production processes are constructed and contested. We work through the heuristic potential of the flow-place notion through a focus on the shifting practices, power relations and transcalar circulations in commercial office space development.

Introduction

“Heavy trucks went downtown bouncing, headed to the garment district or the meatpacking docks, and nobody saw them. They saw the cockney selling children's books from a cardboard box, making his pitch from his knees. Eric thought they were the same thing, these two, and the old Chinese was the same, doing acupoint massage, and the repair crew passing fiber-optic cable down a manhole from an enormous yellow spool. He thought about the amassments, the material crush, days and nights of bumper to bumper, red light, green light, the fixedness of things, the obsolescences, going mostly unseen... How things persist, the habits of gravity and time, in this new and fluid reality” (DeLillo, 2004, page 83).

Developing understandings of the diverse nature, composition and production of cities is a crucial task in an age of increasingly interrelated globalisation and urbanisation. Their incorporation into intensive, global circuits of transmission and exchange appears to problematise traditional conceptualisations of the city as a functionally bounded territory which structures its immediate hinterland (Sassen, 1991; Castells, 1996; Taylor, 2003). At the same time, the very factors behind their development as centres and agglomerations of activities plead for the persistent importance of their physical, territorially-based and, in many cases, historically constituted advantages (Storper, 1997; Beauregard and Haila, 2000). In short, all research on urban development is inherently caught up in what appears to be a kind of conceptual spectrum stretching from a focus on the place-based qualities of cities at one extreme to a flow-based vision of cities at the other.

In our research however, we do not see a space of places diametrically opposed to a space of flows, but more a mutually constitutive logic in which places are only ever constructed through processes of ‘fixing' flows, while flows are only ever produced (and ‘consumed') from more or less ‘fixed' bases/places. Indeed, as flows always have a start and an end point, they are meaningful only if and when they become ‘(re)territorialised' (in cities) at their end points. For example, information or capital can only be used once it reaches its destination. The paradox of the network society is that its main focus (flows) proves to be useless whilst in transit and always requires place-based fixity for utility. From this perspective, and contra some recent strands of urban studies, one could argue that there is essentially nothing in between cities¹. In turn, however, places are never self-bounded entities but are open social and material constructions (Massey, 1994; Latham and McCormack, 2004). They are always constituted by networks of flows. Globalisation and digitisation processes contribute to a vectorisation of places, that is to their increasing engagement in a geography of exchanges and distant links.

This paper sets out to discuss a view of cities as dynamic and unstable flow-places, in which the respective qualities of flows and places cannot be meaningfully disarticulated. Following Latour (1993, page 122), we seek a way into cities that is based more on exploring their “intermediary arrangements” which are always constituted by flows² and places³ in sync rather than necessarily by one more than the other. Departing from Castells' argument (1996), the main thesis of this paper is that if cities are better understood as a process it is not because they constantly mix flows, but because they are a communicational nexus in which flows and places are always being interconnected or commutated. We argue that this implies studying urban production processes⁴ as inherently socio-technical undertakings, which notably demands two methodological steps. First, we need to follow more closely the actions of, and relations between, the diversity of actors involved in producing and constantly reproducing built environments. Second, it requires focus on the complex ways in which these actors mobilise different tools, instruments, norms, regulations and technologies in their work to create variable spatial outcomes.

The paper proceeds in four stages. First, we critically review two current strands of the urban studies literature which conceive of cities in distinctive relational terms, either as deterritorialised functions of global circulation and exchange, or as dynamic meeting places for all kinds of flows. Critiquing their implicit idea of urban assemblage, we sketch out a conceptualisation of cities as flow-places to help understand the simultaneous and mutually constitutive nature of the territorialising/deterritorialising dynamics at work in the production (and reproduction) of the urban. We situate the notion of ‘commutation' as at the heart of urban production processes to capture this idea of cities as a communicational nexus in which flows and places are always (being) interconnected. In developing a socio-technical approach to urban (re)production, we then identify three constitutive components and implications of this commutation process around which the intersection between the social practices of multiple actors and the built environment they work with is produced and contested. We then work through the heuristic potential of this approach in a brief case study of urban office space development. The conclusion summarises our main arguments and discusses their methodological, theoretical and policy relevance for contemporary urban research.

The rise of relational theories of cities

Cities have been increasingly analysed as predominantly relational entities in the last twenty years. Rather than stressing their place-based assets and qualities, urban theorists have refocused their efforts on exploring how the meanings and functions of cities are the produced outcomes of flows, processes and relations which commonly circulate or originate beyond traditional urban boundaries. This reorientation of urban theory (which we hesitate to call a ‘relational turn') can be viewed as having produced two principal strands of research which we discuss briefly in turn.

Dissociating Places and Flows: Cities in Networks

The first strand has been concerned mostly with conceptualising cities as nodes whose primary function is to process the multifaceted flows of an increasingly networked society. Cities here are merely the function of dominating global flows, and the most developed cities (the global and the world cities) are consequently those which process not just the most flows, but also the most qualitatively diverse sets of flows, thus becoming obligatory points of passage across multiple domains or sectors (Sassen, 2001). Cities as places count less here than the flow-based logic which both permeates most if not all elements of urban development and renders irrelevant the traditional territorial boundaries of the city. In short, cities are the outcome of deterritorialised flows which territorial actors have relatively little capacity to regulate and check.

Emblematic of this strand of work has been the research initiated and driven by the GaWC (Globalization and World Cities) group which has refocused the analysis of urban systems upon inter-city relations rather than on cities as central places, with the consequence of replacing the study of hinterlands by the analysis of (world) city-regional hinterworlds (Taylor, 2003). The central importance of global cities to the contemporary political economy is the key evidence for Beaverstock et al. (2000) that the world spatial order is no longer organised around states but as an ‘interlocking' network of cities. Economic globalisation and the widespread use of information and communications technologies (ICTs) have led to a thorough re-evaluation of cities' distinctive properties, notably loosening their traditional ties to specific territories, or at least redefining the nature of their relations with national or regional environments. Leaving aside the hypothesis that globalisation has not fully disconnected cities from their wider regional and national environments (as exemplified in the debates on state rescaling and new regional economies: see Storper, 1997; Swyngedouw et al., 2003; Brenner, 2004), cities are now seen as dynamic entities both emitting and receiving multiple types and levels of flows, such that they are no longer primarily organised as ‘spaces of places' but as ‘spaces of flows' (Castells, 1996).

In this light, the global economy is thus founded on a network organisation, or more precisely, on a structure of innumerable intertwined networks (Dicken, 2007) which 'interlink city-regions' (Taylor, 2003)⁵. This results in the formulation of a geography of city-regions that takes the shape of a worldwide archipelago of metropolises (Scott, 1998; Veltz, 2005) which are interconnected on an everyday basis. One might argue that there is nothing new about this reconfigured spatial organisation: cities now exchange and form an urban system on a global scale rather than the national/regional scale of Christaller. The stakes may have merely been ‘upscaled' to take globalisation into account (ESPON 1.4.3., 2007). According to the world city network perspective, however, the transformation is more important than this as cities have taken on new meanings and identities, being more defined by their exposition to, and production and management of flows, than by the assets they contain within their boundaries (Beaverstock et al., 2000). In other words, cities count for less in the global economy than the flows that are exchanged between them: NY-LON space is more important and valuable than the individual urban spaces of New York and London.

Although all this work has greatly contributed to shifting our understanding of cities away from an overly static (and ‘statised') vision of separate, unrelated, bounded territories, we wish to argue that its conceptualisations of the nature, composition and production of cities (as predominantly the outcome of deterritorialised flows) have three limitations which all still serve to overly dissociate places and flows.

First, the ‘world city network', as “an interlocking network wherein the nodes (cities) are connected through the activities of trans-nodal agents” (Taylor, 2003, page 13), is presently rather narrowly defined as being shaped largely by the locations of the offices of major producer service firms. ‘Defining' power and control of world cities comes from the “actual inter-city social practices” apportioned to global banking, law, insurance, accounting and consulting firms (Taylor, 2003, page 30). Yet the actual detailed strategies of these firms and the concrete social practices of their employees in individual cities are mostly passed over (see Jones, 2002, for a critique) in favour of very general analyses of global office networks which are used as a proxy for inter-city flows (of information, people, investment...) produced by firms (for an exception see the qualitative analyses developed in Hall and Pain, 2006). While it is pertinent to argue that cities need inter-city flows to function (economically at least), we believe that the specificities of urban places and the social practices of their actors should be central to analysis of urban production processes.

Second, although this view helps to clarify the multifaceted ways in which flows shape places, the focus is very rarely, if at all, on the complex ways in which places (through their contingent mixes of social actors and institutions) may or do shape flows (Storper, 1997; Hubbard and Hall, 1998; Smith, 2001). At the same time as flows shape cities, urban actors directly or indirectly contribute to shaping the flows of globalisation in at least three ways. First, they contribute to emitting flows, e.g. in terms of processes of emigration, capital circulation or knowledge and ideas. Second, they operate a filtering or selection process upon ‘incoming' flows, e.g. according to legal or regulatory norms, but also for linguistic, cultural, socio-economic or political reasons. Third, they shape flows by altering content or form to fit particular urban specificities. In other words, flows must be seen as having to adapt to at least some extent to different forms of regulation, resistance and friction that urban actors impose, and which prevent flows from ‘landing' in places anywhere, anytime and in any form.

Third, this kind of approach contributes to a continuing more general rigidification of geographical scales in current urban studies research. This has led to crude oppositions made between ‘global' and ‘local' development processes, or to arbitrary distinctions between the ‘urban' and the ‘extra-urban'⁶, which promote both a rather artificial differentiation of endogenous (inside) and exogenous (outside) dynamics, and a persistently hierarchical and intransigent view of scalar organisation. These problems are not forsaken either by adopting a reverse perspective which collapses scales into fuzzy concepts such as the ‘glocal'.

In summary, we need a more nuanced perspective on urban development to complement that driven by grand notions of the space of flows and the structuring ‘power' of inter-city economic relations, which arbitrarily oppose flows and places in urban analysis as if contemporary globalisation has suddenly split up the mix of processes and practices that collide to make and remake urban space. David Harvey's classic conceptualisation of the inherent contradictions in capitalism between fixity and mobility would appear to be more relevant than ever (see Brenner, 1998). The focus on the “rising power to overcome space and the immobile spatial structures required for such a purpose” (Harvey, 1985, page 150), which lead to dominant forms of territorial organisation underpinned by ‘spatial fixes', offers a powerful view of (capital) flows materialising the urban, and vice versa, which is largely absent in the ‘cities in networks' literature. His more recent identification of the importance of ‘cogredience' in urban life extends this view beyond the political economic realm in propounding a focus on “the way in which multiple processes flow together to construct a single, consistent, coherent though multi-faceted time-space system” (Harvey, 1996, pages 260-61; see also Swyngedouw et al., 2003, on ‘the world in a grain of sand'). In short, while world cities research concentrates on the increasing mobility of the urban, Harvey is more interested in the continuing need of mobile flows for (temporary) territorial fixity. While we are sympathetic to both approaches, it is clear that neither embodies what we see as the more unitary nature of flows and places, with a high degree of sensitivity for place-based context and more focus on the actual (socio-technical) processes through which flows are ‘territorialised'.

Connecting to Relational Urban Theory

The second strand of research offers a way of working towards a more place-sensitive and processual perspective by building on the work of writers such as Doreen Massey, Ash Amin and Nigel Thrift, and Richard Smith, who have contributed to rethinking the ontology of the city around what can be termed a more ‘relational' urban theory (see also MacLeod and Jones, 2007). This emerging loose body of work has taken issue with both heavily contextualised and more generalised urban research which is seen as overly focusing on local specificities or globalising trends respectively (“privileging or dismissing nearness” in Amin and Thrift's terms, 2002, page 37). Instead, relational urban theory seeks to unchain analysis from a bounded conceptualisation of the city and to move it towards a recognition of the city as a dynamic process, without going so far as to envision cities purely as placeless nodes in globalising flows. The city is, therefore, still a local place where people live, work and play, business is conducted, etc., yet its place-based qualities are not founded on a single internal, contextualised identity, but on the fact that it is a site where all kinds of flows meet (albeit temporarily and thus in constantly changing configurations) and are experienced by people in varying ways and to varying extents (see also Urry, 2000, on the forms of mobility which are reconfiguring social life in profoundly diverse ways). As Massey (2007, page 10) argues: “Conceptually, it is important to recognise that the global is as much locally produced as vice versa... that the very fact of specificity (that places vary) both opens up the space for debate and enjoins us to invent”. The city is thus not a pre-existing entity into which flows are attracted. Without its flow-based relations, the city cannot be seen to exist. The key task from this perspective is then not to understand the city (which would imply adopting a fixed definition of the object of study), but to understand the processes which constitute and continuously reshape it.

This more relational perspective on the city can be traced back at least to Doreen Massey's plea for ‘a global sense of place' which might constitute a progressive social differentiation of the processes and outcomes of globalization and time-space compression (Massey, 1994). She outlines here a vision of ‘place' which is processual rather than static, open and extroverted rather than enclosed and introverted, constituted by a constellation of varying and conflictual identities rather than by one unique identity, but which continues to recognise the specificities of individual places. A city is, from this perspective, a ‘meeting place', the particular intersection of social relations which are all “constructed on a far larger scale than what we happen to define for that moment as the place itself” (Massey, 1994, page 154; see also Amin and Thrift, 2002, page 30). In other words, a place is a site of interconnections of differing flows, which are, importantly, differently experienced and understood by people. As the flows and the interconnections change and evolve, so does a place (and the experience of it). For Massey it is this that gives a place its specificity and “not some long internalized history” (Massey, 1994).

Likewise, Amin and Thrift observe “the everydayness of spatially stretched and distant connections” of the city (Amin and Thrift, 2002, page 27). To them, spatial propinquity is no longer the defining feature of the city, as its porosity exposes it to all manner of more ‘distanciated' rhythms, imaginaries and potentialities. As economies, cities are “assemblages of more or less distanciated economic relations which will have different intensities at different locations” or “sites in near-far networks” (Amin and Thrift, 2002, pages 52 and 64; see also Amin, 2002). They finally aim to “redescribe the city as an ecology of circumstance, as an ordering of uncertainty, so producing lines of power, and as a political arena full of potentialities able to be mobilized and fought/thought through” (Amin and Thrift, 2002, page 77).

In short, cities are seen as social arenas of interaction, rather than as places containing ‘things' such as buildings. They are open intensities of social relations (Smith, 2003, page 30), which can be seen to involve a series of processes, events, performances and practices in the making of places as networks. Beyond their representation and interpretation, places are constituted by relations that are constantly being performed (Thrift, 2008). The city is thus an open theatre or stage upon which all kinds of visible and less visible actors perform a multitude of constantly changing lead roles and bit parts (Latour and Hermant, 1998).

This focus on the social relations that produce cities is welcome but too quickly skims over the technical elements of the urban (i.e. buildings, infrastructures, materials, documents, plans, etc.) which are important both in shaping social relations and in forging how specific actors contribute to building and rebuilding cities. Smith (2003) sees actor network theory (ANT) as offering something approaching this kind of socio-technical perspective on cities, in terms of theorising how short and long networks made up of humans and non-humans alike are the basis of urban production processes – processes of ‘heterogeneous engineering' which are always making, maintaining and remaking urban space. ANT also provides the advantage of bypassing fixed and hierarchical visions of geographical scale: Smith quotes the famous statement of Latour on how “the words ‘local' and ‘global' offer points of view on networks that are by nature neither local nor global, but are more or less long and more or less connected” (Latour, 1993). The local is not somehow more concrete than the global out there. The length and connection of networks do not correspond to scales. There is only a continuum of relations in networks of varying length. In this perspective, globalisation is not leading to a rescaling of, or a multiscalar dimension to, whatever activity. Both ‘relational' urban theory and Smith through actor network theory propose more of a transcalar vision of long and short networks, close and distant relations, etc. (Smith, 2008).

This work helps us on two points: the more open approach to the city greatly advances our understanding of the transcalar nature of urban development as it sees cities as simultaneously drawing influences from and extending influences to the farthest corners of the globe. It begins to get beyond an artificial disjuncture between flows and places by envisioning cities as porous places which continuously attract in-flows from near and afar and produce their own out-flows which diffuse over varying distances. It also offers a means to analyse intra-urban differentiations as reflecting the variable levels of openness, attractiveness, speed, rhythms, etc. of different parts of each city: “cities hold a diversity of power arrangements, with different parts of the city locked into different sets of rhythms and connections, with different degrees of influence and authority” (Allen, 1999, page 186). The focus on actors and their practices, and on performances in relation to other actors, opens up a way of exploring how cities are made and understood by social actions and relations, which are sometimes evident but more often implicit, hidden and undocumented (Latour and Hermant, 1998; Pile, 2000).

However, we believe that this ‘relational approach' can be extended in several ways which may help to mobilise or equip it further for urban analysis. This involves first an analytical shift from seeing cities as assemblages of places and flows to observing the process of commutation that urban actors operate between various types of flows and their place-based contexts, and second the development of a more precise socio-technical framework for ‘operationalising' the flow-place notion. It is to these tasks that we turn in the next section.

Conceptualising cities as flow-places: towards a socio-technical approach

From Relational Urban Assemblages to Cities as the Sites and Outcomes of a Commutation of Flows and Places

Many of the authors who have contributed to the emergence of a relational urban theory use the notion of ‘assemblage' to capture the unstable, informal juxtaposition between different types of flows and between flows and places in the city. We see the notion of cities as assemblages of diverse forms of flows and place-based logics as potentially problematic as this still seems to imply some form of arbitrary, mechanistic separation of flows and places⁷. We argue that because it is impossible to separate out functionally or meaningfully flows and places from the start, then it is therefore impossible for actors to combine or even recombine what has never been separate⁸. Instead, we prefer to highlight how flows and places are co-produced in cities through socio-technical processes of commutation.

It is this operation of commutation of flows and places that we argue is always at the heart of the production and reproduction of the urban. Commutation is a communicational process through which relations between two or more systemic entities are (temporarily) established, maintained, modified and/or interrupted (Guillaume, 1999). Traditionally used in the telecommunications sector to describe the technical process which permits the connecting of a telephone call between two people⁹, it has been used metaphorically to illustrate the role of cities in agglomeration economies as maximising the possibility of meetings between individuals, goods, etc. (Claval, 1981; Crague, 2004). Here, the city becomes the commutator as it attracts and brings together various flows on its territory. In other words, it is flows which are commutated in places through social and technical networks (Crague, 2004, page 9). Our conceptualisation is slightly different insofar as what is important is the commutation operated between flows and places, and not (just) that between different kinds of flows. Rather than seeing the city as a vast commutator, we argue that urban actors use their differing tools, instruments and technologies to commutate flows and places in multiple, sometimes conflicting ways. Rather than operating purely in the interstices of various networks to interlink flows¹⁰, they are more often to be found interpreting, translating and adapting their own dynamic flow-place contexts. In much the same way as a telephone exchange exists to operate the link between the network and a place-based (although increasingly mobile) user, so can the work of many urban actors be seen to almost solely involve the commutation of flows and places. The city is therefore at once the site and the outcome of these diverse, ever-changing socio-technical commutation operations.

We see three advantages for such an approach. First, this co-production/commutation places an emphasis on the non-mechanistic ways in which cities as flow-places are constructed by multiple actors, their practices and their strategies. Second, an important part of this, which is less fully conceptualised in relational urban theory, is the technical dimension of the production of the urban. Commutation of flows and places in cities is a socio-technical process because not only is it initiated, managed, resisted and undergone by diverse sets of social actors, but it always relies on and works through the tools and instruments that these actors have at their disposal and that they use on an everyday basis. Third, we argue that this socio-technical co-production/commutation approach begins to offer some practical, heuristic purchase on how we might go out and study cities in a more correlative sense, i.e. as flow-places.

Unpacking the Flow-Place Commutation Process

The notion of flow-place is valuable both for the holistic approach it allows, and its privileging of the multiple and diverse ways that actors contribute to urban production processes. As commutation operations are inherently founded on the intersection between the social practices of multiple actors and the built environment and materials they work with, it follows on that the conceptualisation of cities as flow-places demands a socio-technical perspective. We identify and outline here three constitutive components and implications of the flow-place commutation process through which cities are produced and contested, and around which a socio-technical perspective might thus be constructed (Table 1).

Table 1: Components and implications of the flow-place commutation process.

First, urban commutation of flows and places occurs primarily through a series of everyday practices, constructed by actors in relation to the tools, instruments, norms, regulations and technologies that they use in their actions. Nevertheless, our current understanding of urban development and flows remains constrained by difficulties in ‘mapping' and assessing the changing roles and interactions of the many public and private actors involved. Increasingly multifarious and transcalar flows, reinforced by globalisation dynamics, simultaneously create new roles for actors, suppress existing ones, and reinforce or restrict others. It is here that ‘following the actors' might bring in the mobile practices and performativities of non-representational theory to study quite simply what actors do (and how what they do evolves over time). If performance is “the art of producing the now” (Thrift, 2000, page 577), then it is fair to assume that we can already learn a great deal about how cities are produced and reproduced simply by unpacking the agency and socio-technical ‘performances' of the different actors on the urban stage, in terms of how and with whom they interact and how they make use of the various props (tools, instruments, norms, regulations and technologies) at their disposal. This first component of commutation demands then analysis of how urban actors (commutators) are engaging concretely with different types of flows on an everyday basis. This is very much about understanding how the various elements of the commutation process function and are managed in cities:

"Commutation is also a multitude of everyday operations, most of them done without us realising. Moving around a city, using a remote control to channel hop, flicking through a book, making a phone call or even doing a simple calculation on a computer are all examples of operations largely based on commutators (Guillaume, 1999, page 18).

The everyday practices/operations performed by actors change or evolve too through time, necessitating temporally dynamic analysis focusing on different moments or periods throughout short- and long-term commutation processes. In some ongoing research on urban energy and climate policy issues in Stockholm, for example, one factor pinpointed by officials as important in shaping the objectives and actions of this policy was simply the extent to which, at a given point in time, they were allowed (by their hierarchy) to work on these issues (as opposed to other environmental or urban issues). When the issues were political priorities, officials were given the capacity to actively develop policy responses (and to construct their day-to-day work practices accordingly), whereas at other times, they were forced to do other work involving a quite different performance of commutation operations. Understanding, therefore, what people do at different times within administrations and companies sheds considerable light on how cities work (differently) as the sites and outcomes of flow-place commutation.

Second, and following on from this, urban commutation of flows and places inherently involves and (re)produces the complex and dynamic systems of power relations between sets of actors with more or less diverging or conflictual objectives. The presence of powerful and less powerful actors throughout the flow-place commutation process ensures that urban development is shaped within a constantly unstable context of tensions and harmonies between different actors: “a field of multiple actors, trajectories, stories with their own energies – which may mingle in harmony, collide, even annihilate each other” (Massey, 2007, page 22). Whether it is the transnational social practices of Guatemalan refugees (Smith, 2001), the concurrent ‘flexibility' and ‘coercion' offered by the car to drivers (Urry, 2000, pages 59-63), negotiations and decisions about the most effective forms of urban policy for tackling climate change, or the simultaneous networked-territorial operations of illegal or semi-legal commerce and trade of arms, narcotics, sex and laundered money (see Castells, 1998, chapter 3; Zook, 2003), the simple juxtaposition of urban actors always promotes contested forms of place-making. The commutation practices and performances of individual actors always have consequences for other actors and groups. It is crucial therefore to explore h ow actors situate themselves and their actions in relation to others, and to what extent their use of particular tools and instruments materialises relations of power and influence in urban development (c.f. Moulaert et al., 2003). We suggest that this political economy-oriented perspective on urban development is necessary in order to separate out who operates (and/or commands) processes of commutation, for whom and for what overarching purpose, and with what social and political ramifications in terms of winners and losers (c.f. Heynen et al., 2006). Uncovering the power relations underpinning flow-place commutation processes is crucial therefore to building understandings of the social and political tensions inherent to urban development, with a view to promoting alternative, more inclusive forms of city-building.

Third, and again related to the first two points, urban commutation of flows and places inherently involves a transcalar, systemic set of processes and practices from multiple scales, thus requiring a conceptual shift beyond scalar fixity to viewing space more as a soft continuum of communications between the here and the distant (Massey, 2007). In other words, the spatial arena of flow-place commutation always extends well beyond any political or functional city boundaries. Going beyond global-local debates that tend to rigidify geographies in a series of fixed scales, even within a multiscalar approach (Smith, 2008), we thus understand flow-place commutation to be inherently transcalar, invoking (continually changing) links and connections between socio-technical practices and processes on a priori differing geographical levels. In describing the restructuring of legal services in Singapore, for example, Smith (2008) explains the evolving actor-networks of ‘domestic' lawyers and law firms as they interact with ‘foreign' firms not as a narrative of ‘local' firms becoming ‘globalised', but simply as a process of the gradual lengthening of always local actor-networks which reach further without any shift between scales. In a similar manner, we see the commutation of flows and places as mixing multiple scales into a soft continuity (though not necessarily without rupture) of spaces. Such a transcalar continuum avoids the trap of ‘pigeon-holing' or ‘compartmentalising' actors and socio-technical relations within fixed and boxed-in scales, recognising instead that “There is a vast geography of dependencies, relations and effects that spreads out from here around the globe” (Massey, 2007, page 13) and also arrives here from across the globe. This involves a ‘softer', more flexible and open vision of the spatial continuum of flow-place socio-technical relations, which might create more potential room for manoeuvre for urban actors, but can just as easily become commandeered by powerful elites for their own purposes.

In the next section, we work through our socio-technical approach to cities as the context and outcome of flow-place commutation by applying it to analysis of the complex relations between global capital circulation and one dimension of the (re-)production of cities, the development of commercial office space in large cities.

Mobilising flow-place: the example of office space development in cities

Commercial real estate, including office space, is widely recognised as being necessary for any local/regional form of economic development with the quantity and quality of available buildings being crucial to the overall efficiency of all firms (Henneberry and Roberts, 2008). However, with the trend of many companies increasingly focusing on what they consider to be their ‘core' activities, the ownership of offices, and indeed most commercial real estate buildings, has been transferred from occupying firms to real estate-focused investment funds (institutional investors, pension funds, hedge funds, Real Estate Investment Trusts, etc.). These investors are developing global capital allocation strategies at world-level both to benefit from the growth of so-called emerging cities¹¹, for example in India, China and South America, and to reduce their geographic risks. As a result, an increasing proportion of office developments in most emerging and developed world/global cities are owned by these ‘international' or ‘global' investors. Capital flows from North American, European, Middle-East and Asian investors thus contribute to shaping the landscapes and spatial organisation of cities by selecting and funding the development of particular districts / buildings (Fainstein, 2001; Halbert and Diziain, 2006; Fainstein, 2008). One can see this as a clear-cut example of global flows shaping cities. However this is but one part of the story.

Real estate investments do not ‘land' accidentally in cities. It is the representations and attributes of these cities (and of the countries in which they are located) that influence, at least to some extent, investment decisions. There are (at least) two ways in which this occurs. First, investment managers base their capital allocation models on a series of hypotheses (expected rental and selling prices, vacancy durations between two occupiers) that are said to reflect the dynamics of a given city (growth curve, sectoral specialization, quality/quantity of the workforce, etc.). To do this, they interact with a number of ‘local' actors who play the role of commutators in the decision-making process. These are lawyers and legal consultants that help investors to select among the multiple investment pathways available to them (direct investments in projects or portfolios of projects, investments at company level in the form of private equity in a real estate developer, wholly owned subsidiary firms). This is also the role of real estate agencies, be they local realtors or branch offices of the global majors, which all have their own database and tacit knowledge on the local real estate micro-markets. These agencies critically contribute to simplifying the complexity of a city-region into a limited number of real estate and/or financial indicators. In this task, both the use of cartographic and statistical techniques (often coupled in GIS systems) and the mobilising of diverse social and professional networks are important. Second, potential investors rely on the ability of the local property industry to access land and to develop the actual buildings. In the case of Bangalore in India, our findings show that a group of local actors have developed a niche situation in sourcing, aggregating and, above all, 'clarifying' title deeds so that, as a result of their commutational work, the city-region becomes ‘usable' for investors that have a series of pre-requisites in terms of transparency and costs. This chain of commutation often starts with real estate developers that are the primary link to investors, but the chain involves a myriad of more or less formally organised intermediaries (real estate agencies, freelance consultants, 'middlemen', ‘mafia') that have the social and technical know-how to assemble and obtain legal permissions for a given land. This involves the ability to sign market deals with very fragmented (and often rural) landowners or with large landlords that are (politically) well-connected, but it also necessitates the use of other forms of commutation, from what is known as 'speed up' money and bribery to the use of physical threat or violence. Once the land is incorporated into the market, a second strand of the property industry conceives, develops and, after completion, maintains and manages the buildings that support investments from fund managers. This obviously demands the work of developers, architects and planners, and asset management companies (security, cleaning, catering, etc.), which allows the value for investors to be guaranteed at the development stage and throughout the entire life of the building. Thus, from the crane to the broom, there is a series of human and non-human actors that contribute to ‘fixing' capital flows (which nonetheless circulate on a quarterly basis between investment opportunities) into the materialities of city-regions. Cities can therefore be conceived here as inherently socio-technical systems constituted by buildings and multiple actors (firms, elected representatives, planners, residents) that each bring their own geographies, temporalities, tools, norms, institutions and regulatory environments to bear in their relations with the investment industry. In this way, flows and places are interlinked to the point of being analytically inseparable. The place-based attributes (e.g. geographies, norms, institutions, etc.) are used by the financial sector to calculate the differential risks between locations (within or between cities) which are critical in defining the nature and size of investments. In other words, it is because of the urban specificities and frictions that investors can expect certain rates of return in certain parts of the city. Analysing the real estate industry in cities therefore requires exploring how the circulation of capital both shapes and is shaped by the fixed assets or materialities of those cities. Cities thus exist in this case as sites and outcomes of processes of commutating flows and places.

Figure 1: Brand-new office buildings and labourers' tents (Bangalore).

To separate out how flows shape places from how places shape flows would lead in this case to potentially misleading and simplified conclusions wherein either cities are considered purely as the material productions resulting from the circulation of capital or global capital circulation is overly determined by the resilience of heterogeneous geographies (between and within cities). The example of commercial real estate highlights instead the analytical pertinence of the notion of flow-place in at least two ways. First, it invites us to analyse flows of capital and urban office development not as two distinct objects facing off to each other, but as one mutually constitutive process in which flows and places define and reinforce one another. Second, it encourages us to shift from macro-level analyses to address the everyday practices and performativities of a complex socio-technical system that links together in a long distance network continuum the investors in Manhattan, national policy makers defining the regulatory framework for real estate-related Foreign Direct Investments, ‘international' architects and ‘local' developers, unskilled migrant workers surviving in campsites near the constructions and city planning authorities, but also non-humans such as buildings, cranes, tents, and financial spreadsheets used by investment managers. The key focus is thus to analyse the socio-technical commutation processes all along the chain that links capital circulation and the buildings. Our aim must be to look at how the investment decision is practically made and results from negotiations between heterogeneous agents in a continuum of scales that runs from so-called ‘global' investors to the street level work which goes into developing a building.

Recent work we have conducted on the office market in the Paris region backs up these observations, highlighting how ‘external' investors have developed numerous contact points either within or outside their firm to engage with understandings of local Parisian geographies (Auvray et al., 2008). The opposition between the local and the global often mobilised to explain the circulation of capital is thus irrelevant in this case, because of the commutators drawing the local and the global into one systemic continuum. Although financial calculations/anticipations are increasingly important in understanding the intra- and inter-city uneven development of offices (Henneberry and Roberts, 2008), we have also observed the development of a commutational metropolitan 'knowledge arena' that informs investment strategies. Investment decisions are being made mostly by comparison with existing buildings in the vicinity, while the ability of private but also public actors to create, analyse and circulate knowledge on regional submarkets is critical to guide the final investment decisions. In this sense, the differentiating factors between investment/development actors are now mostly related to the knowledge economy. The increasingly professionalised knowledge production chain within the investment and real estate development industry has become a commutator between capital circulation and the construction and reconstruction of commercial buildings in cities. This explains the growing role of advisory companies, including the majors (Jones Lang Lasalle, CBRE and Cushman & Wakefield), in modelling city-region trajectories (Figure 2). They contribute to the production of representations of flow-place cities which then directly/indirectly influence how cities are constructed and reconstructed.

Figure 2: The Property clock model by Jones Lang Lasalle (2006).

The description of this changing socio-technical system should lead on to analysis of the reconfiguration of urban power relations. The development of cities as illustrated by commercial real estate dynamics is shifting from a metropolitan system where local / national actors (promoters and public bodies) have been playing a critical role in shaping a city's development to a situation where international investors and their numerous metropolitan commutators gain new power positions in the material development of urban space (Fainstein, 2001, 2008; Torrance, 2008). This leads nonetheless to very different local configurations. In the case of the Paris region, owner-occupier firms and developers which have been the drivers of commercial real estate development since the 1960s must now negotiate with another couple formed by fund managers and their accompanying real estate consultant agencies. This leads to changing relations vis-à-vis the materiality of the Paris city-region which is increasingly gauged in financial terms. It is in this sense that the Excel spreadsheet determining the major financial characteristics of a building is as important as the crane that builds it. In the case of Bangalore, the arrival of international capital has had direct impacts on the local and Indian property development industry, but in a different manner. Traditional real estate developers (often families) as well as newcomers that hold some knowledge of the local workings of the political system and real estate industry have managed to reinforce their position by playing the major commutating role between international investors in search of predefined financial characteristics (transparency, liquidity, good appreciation of risks, appropriate return on investment ratios) and a local system of actors that would have failed to meet these prerequisites otherwise, if only because of the complexity of the local scene. In both the Paris and Bangalore cases however, the new realignments crucially question the place, role and legitimacy of public authorities in urban production processes. In the North, this may be used as a hypothesis to explain both the shift from traditional planning policies to a more strategy-oriented public sector agenda (Swyngedouw et al., 2003), and the rise of 'progressive community planning' in reaction to the growing role of 'global' investment and property development sectors (Angotti, 2008). In India meanwhile, the strength of the demand for investments coming from international fund managers is such that not only have the national policies been altered to welcome them, but local and State policy-makers are urged to adapt their policies to meet their expectations in terms of infrastructure development, master planning rules, etc.

This application of a flow-place approach to office development explores only one aspect of the urban production processes that make and remake cities. A similar socio-technical approach could be adopted to study, for example, the development of infrastructure networks in cities. Indeed, recent work on telecommunications network provision in European cities has highlighted the presence of a similar flow-place commutation process whereby ‘global' operators have been investing and deploying their fibre networks in cities according to the mix of historical, regulatory, economic and physical specificities/‘territorialities' of each city. As in property development, infrastructure investment decisions depend on negotiations with and knowledge provided by numerous ‘local' actors/commutators (local authorities, consultant firms, engineers…) which again reconfigures the power geometries enveloped in the production of the urban (Rutherford, 2004). The notion of flow-place thus has potentially wide application to help disentangle the complex social and technical processes through which cities are materially made and re-made.

Table 2: Summarising a flow-place approach to real estate development in cities.

Conclusion

In this paper, we have argued that we need to move beyond relational conceptualisations of cities to focus on the socio-technical practices and implications of urban production processes in which multiple actors perform, manage, undergo and contest a variety of commutation operations interconnecting flows and places. We see two main methodological-analytical advantages to this flow-place commutation approach. First, it invites us to analyse flows (of capital, people, goods, ideas, etc.) and urban development neither as distinct objects or moments facing off to, or superimposing, each other (as in work on cities as disembedded nodes in globalising networks), nor as mechanistic urban assemblages, articulations or combinations (as is described in recent relational urban theory), but as one mutually constitutive, cogredient process in which flows and places are co-produced, continually defining and reinforcing one another. Second, it encourages us to shift from generalised macro-level or ‘metageographical' analyses to address the contingent practices and performativities of a complex socio-technical system that links together multiple actors and technical instruments in one long distance network continuum (investors in New York, national policy makers defining the regulatory frameworks, ‘international' building or network architects and ‘local' developers, workers and city planning authorities, but also pipes, cables, buildings, cranes, diggers, economic models and financial spreadsheets). The key focus must be to analyse the socio-technical commutation processes all along the chain that link flows, actor negotiations and decisions, and the concrete production of the urban. This approach is therefore theoretically conscious of the centrality of technical elements of the urban in giving or denying agency to actors. Yet it is also concerned with ‘operationalising' the approach through empirical observation of ongoing urban production processes, and analysis of the performance and functioning, political meaning and implications, and spatial arenas of the diverse commutation operations which constitute these processes.

In policy terms, the centrality of flow-place commutation to urban development is highly ambivalent in its socio-political implications and outcomes. Flow-places, as sites and outcomes of socio-technical relations present within and extending beyond political and functional city boundaries, concurrently reinforce to some extent territorialised/institutionalised politics (through attention to the specificities of place) whilst conceiving of the contested arena of place-making as porous and transcalar (see MacLeod and Jones, 2007). While this may be quite constraining in the ways that some actors are more able than others to dominate these processes and (implicitly or explicitly) bypass or ‘disconnect' less powerful interests, our approach also envisions/enables all kinds of actors to possess the (potential or actual) capacity to contribute to and contest the multiple and varied commutation processes that construct and reconstruct cities. This might therefore open up new spaces for a material urban politics which recognises the sheer diversity of actors centrally involved in the production of the urban, and in which local and city authorities can grasp more concretely the stakes of everyday urban development rather than being allocated less room for manoeuvre in the current paradigm of depoliticised competitive/collaborative policy-making (see Veltz, 2008). Rather than being forced (willingly or resentfully) to adapt to logics of networking and connectivity at all costs, the work of many actors in urban governance is still largely concerned with the regulation and management of a given territory. While we argue that this work is increasingly constituted by the initiation, performance and management of tasks which commutate (‘extra-territorial') flows and place (c.f. the tasks of urban actors in commercial office development discussed in the previous section), and therefore concretely opens up the work of cities to wider (or longer) communicational dynamics, we contend that flow-place making should continue to be as much, if not more, attuned to the contested adjustments of urban inertias, sediments and persistencies as to full reconfigurations of the spatial arenas of urban production dominated by a purely relational or aterritorial politics of place. This recognises that, for the majority of urban populations, meaningful commutation of flows and places may actually involve a multiplication of quite small-scale, everyday practices and routines that incrementally make and remake cities in distinctive ways (see Varsanyi, 2000). An authentic politics of flow-place would mesh together these practices and routines to develop alternative, more empowering ways for minority groups and interests to use their own commutation capacities in a more inclusive form of city-building. In whatever form this might take, there must be recognition that what counts is not the duality or double meaning of cities as (global) spaces of flows and (local) spaces of places, nor even the articulation or assemblage between the two, but that cities are above all the socio-technically constructed and contested sites and outcomes of multiple and diverse processes of commutation through which flows and places are always being interconnected.

ACKNOWLEDGMENTS

Previous versions of this paper were presented to audiences at the Regional Studies Association conference in Prague (27-29 May 2008), the 'Paris, global city in 2020' working group at the Chamber of Commerce and Industry of the City of Paris, and the Networks, Institutions and Territories research group in LATTS. We would like to thank the participants at these events for their interest and stimulating questions. Jonathan Rutherford acknowledges financial support from the Agence Nationale de la Recherche (project ANR-05-BLAN-0344).

REFERENCES

Alderson A, Beckfield J, 2004, "Power and position in the world city system" American Journal of Sociology 109 811-851

Allen J, 1999, "Cities of power and influence: settled formations", in Unsettling Cities: Movement/Settlement Eds J Allen, D Massey, M Pryke (Routledge, London) pp 181-227

Amin A, 2002, "Spatialities of globalisation" Environment and Planning A 34 385-399

Amin A, Thrift N, 2002 Cities: Reimagining the Urban (Polity, Cambridge)

Angotti T, 2008 New York for Sale: Community Planning Confronts Global Real Estate (MIT Press, Cambridge)

Auvray A, Costantino R, Hervé F, 2008, "La filière d'investissement en immobilier d'entreprise et ses impacts sur le territoire en Ile-de-France", (Ecole des Ponts - Université Paris-Est - Latts)

Beauregard R A, Haila A, 2000, "The unavoidable continuities of the city ", in Globalizing Cities: A New Spatial Order? Eds P Marcuse, R van Kempen (Blackwell, Oxford) pp 22-36

Beaverstock J V, Smith R G, Taylor P J, 2000, "World-City Network: A New Metageography?" Annals of the Association of American Geographers 90 123 - 134

Brenner N, 1998, "Between fixity and motion: accumulation, territorial organization and the historical geography of spatial scales" Environment and Planning D: Society and Space 16 459-481

Brenner N, 2004 New State Spaces: Urban Governance and the Rescaling of Statehood (Oxford University Press, Oxford)

Carroll W K, 2007, "Global cities in the global corporate network" Environment and Planning A 39 2297-2323

Castells M, 1996 The Information Age, Volume 1: The Rise of the Network Society (Blackwell, Oxford)

Castells M, 1998 End of Millennium (Blackwell, Oxford)

Castells M, 2000, "Materials for an exploratory theory of the network society" British Journal of Sociology 51 5-24

Cattan N, 2007, "Students mobility, gender and polycentrism in Europe", in Cities and Networks in Europe: A Critical Approach of Polycentrism Ed N Cattan (John Libbey, Montrouge) pp 139-148

Claval P, 1981 La Logique des Villes: Essai d'Urbanologie (Litec, Paris)

Crague G, 2004, "Commutation. Essai sur l'économie de l'agglomération" Géographie, Economie, Société 6 9-21

Curien N, Dupuy G, 1996 Réseaux de Communication, Marchés et Territoires (Presses de l'Ecole Nationale des Ponts et Chaussées, Paris)

DeLillo D, 2004 Cosmopolis (Picador, London)

Derudder B, Devriendt L, Witlox F, 2007, "Flying where you don't want to go: an empirical analysis of hubs in the global airline network" Tijdschrift voor Economische en Sociale Geografie 98 307-324

Dicken P, 2007 Global Shift: Mapping the Changing Contours of the World Economy (Sage, London)

ESPON 1.4.3., 2007, "Study on Urban function. – Final Report, ESPON 1.4.3. programme ", p 253

Fainstein S, 2001 The City Builders: property development in New York and London 1980-2000 (University Press of Kansas, Lawrence)

Fainstein S, 2008, "Mega-projects in New York, London and Amsterdam" International Journal of Urban and Regional Research 32 768-785

Guillaume M, 1999 L'Empire des Réseaux (Descartes & Cie, Paris)

Halbert L, 2007, "L'horizon des entreprises franciliennes vu par les appels téléphoniques", in Données Urbaines, Tome 5 Eds M-F Mattei, D Pumain (Economica, Paris) pp 269-282

Halbert L, Diziain R, 2006, "Débordement et immobilier d'entreprise", in Effets de débordement, polarisations économiques et stratégies d'acteurs en Ile-de-France et dans le Bassin Parisien Eds M Berger, R Diziain, L Halbert, M Rousset-Deschamps, P Thiard (Minsitère de l'Equipement, des Transports et du Logement, Paris) pp 81-136

Hall P, Pain K, 2006 The polycentric Metropolis. Learning from mega-city regions in Europe (Earthscan, London,)

Harvey D, 1985, "The geopolitics of capitalism", in Social Relations and Spatial Structures Eds D Gregory, J Urry (Macmillan, Basingstoke) pp 128-163

Harvey D, 1996 Justice, Nature and the Geography of Difference (Blackwell, Oxford)

Henneberry J, Roberts C, 2008, "Calculated Inequality? Portfolio Benchmarking and Regional Office Property Investment in the UK" Urban Studies 45 1217-1241

Heynen N, Kaika M, Swyngedouw E, 2006 In the Nature of Cities: Urban Political Ecology and the Politics of Urban Metabolism (Routledge, London)

Hubbard P, Hall T, 1998, "The entrepreneurial city and the ‘new urban politics'”", in The Entrepreneurial City: Geographies of Politics, Regime and Representation Eds T Hall, P Hubbard (John Wiley, Chichester) pp 1-23

Jones A, 2002, "The ‘global city' misconceived: the myth of ‘global management' in transnational service firms" Geoforum 33 335-350

Latham A, McCormack D, 2004, "Moving cities: rethinking the materialities of urban geographies" Progress in Human Geography 28 701-724

Latour B, 1993 We Have Never Been Modern (Harvester Wheatsheaf, London)

Latour B, 2005 Reassembling the Social: An Introduction to Actor-Network Theory (Oxford University Press, Oxford)

Latour B, Hermant E, 1998 Paris, Ville Invisible (La Découverte, Paris)

MacLeod G, Jones M, 2007, "Territorial, scalar, networked, connected: in what sense a 'regional world'? " Regional Studies 41 1177-1191

Massey D, 1994 Space, Place and Gender (Cambridge)

Massey D, 2007 World City (Polity, Cambridge)

Moulaert F, Rodriguez A, Swyngedouw E, 2003 The Globalized City: Economic Restructuring and Social Polarization in European Cities (Oxford University Press, Oxford)

Musso P, 1997 Télécommunications et Philosophie des Réseaux (Presses Universitaires de France, Paris)

Pile S, 2000, "The un(known) city… or, an urban geography of what lies buried below the surface", in The Unknown City: Contesting Architecture and Social Space Eds I Borden, J Kerr, J Rendell, A Pivaro (MIT Press, London)

Rozenblat C, Pumain D, 2006, "Firm linkages, innovation and the evolution of urban systems", in Cities in Globalization: Practices, policies and theories Eds P J Taylor, B Derudder, P Saey, F Witlox (Routledge, London, New York) pp 124-149

Rutherford J, 2004 A Tale of Two Global Cities: Comparing the Territorialities of Telecommunications Developments in Paris and London (Ashgate, Aldershot)

Rutherford J, Gillespie A, Richardson R, 2004, "The territoriality of pan-European telecommunications backbone networks" Journal of Urban Technology 11 1-34

Sassen S, 1991 The Global City: New York, London, Tokyo, (Princetown University Press, New York)

Sassen S, 2001 The Global City: New York, London, Tokyo (Princeton University Press, Princeton)

Short J, Kim Y, Kuss M, Wells H, 1996, "The dirty little secret of world cities research " International Journal of Urban and Regional Research 20 697-717

Smith M P, 2001 Transnational Urbanism: Locating Globalization (Blackwell, Oxford)

Smith R, 2003, "World city actor-networks" Progress in Human Geography 27 25-44

Smith R, 2008, "Urban studies without ‘scale': localizing the global through Singapore", in Urban Assemblages: How Actor-Network Theory Changes Urban Studies Eds T Bender, I Farias (Routledge, London)

Storper M, 1997 The Regional World: Territorial Development in a Global Economy (Guilford Press, London and New York)

Swyngedouw E, Moulaert F, Rodriguez A, 2003, "‘The world in a grain of sand': large-scale urban development projects and the dynamics of ‘glocal' transformations”", in The Globalized City: Economic Restructuring and Social Polarization in European Cities Eds F Moulaert, A Rodriguez, E Swyngedouw (Oxford University Press, Oxford) pp 9-28

Taylor P J, 2003 World City Network: a Global Urban Analysis (Routledge, New York)

Thrift N, 2000, "Performance", in The Dictionary of Human Geography Eds R Johnston, D Gregory, G Pratt, M Watts (Blackwell, Oxford) pp 577-578

Thrift N, 2008 Non-Representational Theory: Space, Politics, Affect (Routledge, London)

Torrance M I, 2008, "Forging Glocal Governance? Urban Infrastructures as Networked Financial Products" International Journal of Urban and Regional Research 32 1-21

Urry J, 2000 Sociology Beyond Societies: Mobilities for the Twenty-First Century (Routledge, London)

Varsanyi M, 2000, "Global cities from the ground up: a response to Peter Taylor" Political Geography 19 33-38

Veltz P, 2008 Des Lieux et des Liens: Essai sur les Politiques du Territoire à l'Heure de la Mondialisation (Ed. de l'Aube, La Tour d'Aigues)

Zook M, 2003, "Underground globalization: mapping the space of flows of the Internet adult industry " Environment and Planning A 35 1261-1286

NOTES

^* Ludovic Halbert and Jonathan Rutherford, LATTS (Laboratoire Techniques, Territoires et Sociétés), Université Paris-Est, France, Email: ludovic.halbert@enpc.fr, jonathan.rutherford@enpc.fr

1. As Latour suggests: “There exists no place that can be said to be ‘non-local'. If something is to be ‘delocalized', it means that it is being sent from one place to some other place, not from one place to no place” (Latour, 2005).

2. We understand ‘flows' to refer to forms of circulation (of ideas, people, capital, goods, etc.) in space which are more or less interrelated and more or less spatially extensive.

3. We understand ‘places' to loosely refer to functional and/or meaningful spatial entities which may be more or less bounded and are experienced in differing ways by different people.

4. We understand this term to refer to sectoral and cross-cutting processes involving a variety of human and non-human actors that contribute to building and rebuilding the ‘infrastructural' fabric of cities (buildings, networks, etc.).

5. Since the 1990s, this new 'metageography' has faced a now well-informed empirical challenge, i.e. the so-called dearth of empirical data (Short et al., 1996) which has driven numerous studies on the flows exchanged between world or global cities (for a discussion on the differing theoretical canvas between these two notions, see Derudder, 2006). Researchers have made numerous attempts to grasp the quantity and quality of inter-city flows associated with telecommunications networks (Rutherford et al., 2004), airline flights (Derudder et al., 2007), foreign student exchanges (Cattan, 2007), the location of multi-office advanced services firms (Hall and Pain, 2006), the network of multinational headquarters and their subsidiaries (Alderson and Beckfield, 2004; Rozenblat and Pumain, 2006; Carroll, 2007) and the geography of firms' telephone calls (Halbert, 2007), in order to capture the production and processing of globalisation flows.

6. Some (but not all) of the recent work on neoliberalism and its ‘locally' contingent expressions can be said to fall into this trap.

7. The other implication of ‘assemblage' is that there is a coherent end product or achievable objective to be had from assembling, combining or cobbling together the different processes and practices that constitute cities, when in fact cities are always in the process of being made (and remade) and never constitute even temporarily stable entities.

8. Even the very term ‘relational' implies here some form of passive dissociation of flows and places by suggesting a ‘relation' between the two rather than their unitary nature.

9. Commutation in the telecommunications sector is a term which actually describes three sets of operations: the connection of a transmitter and a receiver; the processing of a call through the establishment and the interruption/breaking of the link; the control functions (fixing of a price, maintenance, system operation) (Musso, 1997, page 229; see also Curien and Dupuy, 1996).

10. Castells developed the notion of ‘switchers' to describe the actors who hold the most power in the space of flows by acting as the intermediaries between differing networks (Castells, 2000). If we prefer the term ‘commutation', it is primarily because this seems to better capture the unitary, simultaneous flow- and place-based logics of urban development.

11. Cities such as Mumbai, Bangalore, Shanghai, Sao Paulo and Mexico can be said to be ‘emerging' in at least two ways: i) as nodes of production/control in the global economy; ii) as increasingly important poles attracting international investments to support the development of their material infrastructure (transport, buildings, basic services like water, sanitation and electricity).