This Research Bulletin has been published in City, 6 (3), (2002), 351-368
under the title
WORLD CITIES: A FUZZY CONCEPT?
'One shouldn't complicate things for the pleasure of complicating, but one should also never simplify or pretend to be sure of such simplicity where there is none. If things were simple, word would have gotten around' (Derrida, 1988, p. 118).
Despite many attempts to clarify the nature of world cities and refine their analysis, serious doubts remain about their ontological, epistemological, and methodological status. For example, in a recent attack on the proliferation of 'fuzzy concepts' in urban and regional studies, Markusen (1999) singled out 'world cities,' 'networking,' and 'co-operative competition' as exemplary forms of fuzzy thinking. Simplifying to the extreme, she argues that while fuzzy concepts connote, unfuzzy concepts denote.1 On that score, 'world city' is a fuzzy concept because it has not one meaning, but at least four: a leadership role at the international scale (whether through emulation or domination); an external orientation towards the global economy; a high ranking in the world's urban hierarchy; and a major gateway for immigration. Of course, this polysemy can be extended indefinitely, not least by amplifying the differences between the economic, political, social, and cultural criteria for designating the elusive quality of world-cityness. Significantly, most commentators have tended to accord the economic a privileged position in their analyses: initially in terms of the presence of transnational corporation (TNC) headquarters; then in terms of financial markets (especially capital and equity markets); and most recently in terms of a broader range of advanced producer services, such as credit rating, multi-jurisdictional law, and risk management (see, for example, Beaverstock et al., 1997; Friedmann, 1986; Short and Kim, 1998; Taylor, 1997). The multiple connotations of 'world city' thus engender the kind of ill-definition that eludes an empirical grasp: obviously, if sense is unstable then reference is severely hampered.
Hence, Markusen calls for a denotation that would lead to high-definition and a corresponding ability to identify real-world referents. Accordingly, she suggests a simple test to ward off fuzziness: 'How do we know it when we see it?' (Markusen, 1999, p. 870). The difficulty with this argument is threefold. First, since denotation depends upon the play of connotation, undecidability is ever-present. Polysemy and dissemination are the norm, not the deviation. Second, polysemy and contradiction are not at all incompatible with rigorous conceptualization-quite the contrary. For example, world-cityness may be framed dialectically as the outcome of systemic contradictions¾such as the need for centralized reflexive monitoring of economic activity¾or else as the co-relation of multiple economic and non-economic variables. Finally, the empirical cannot be reduced to what is supposedly fully present. Not only do most things have immaterial and invisible components (from ideas to software), but everything ultimately exists in a network of relations that partly¾and in some cases wholly¾constitute it (from the differential relations that lend things specificity to the flows of energy that animate and sustain them). If reality were reducible to what is merely present, then relations and structures would simply cease to exist! And yet for all that, Markusen is right to suggest that the notion of 'world cities' does indeed require further clarification: not by eliminating its polysemy in the vain hope of isolating a kernel of truth, but by taking its seemingly oxymoronic status literally. By conflating scales the term 'world city' appears to be a contradiction in terms and a chaotic conception par excellence. Referring to a plurality of 'world cities' only compounds the difficulty. How can a city be a world? How can London encompass New York and vice versa? How would we identify and come to know such a city even if we saw it?
In this paper we wish to suggest that clarity will only emerge when the worlding of cities is taken literally¾that is to say, materially, not just metaphorically. The 'world' of 'world city' refers neither to an orientation (a city turned towards the world) nor to a status (a city in the world): it refers to an actuality¾the city does indeed exist as a world. But each city exists as a world differently. In other words, our understanding of the city needs to be radically re-scaled, with notions that a world city is a bounded place plugged into a global space of flows rejected in favour of a perspective that sees world-cityness as an achievement of performances and encounters that are distributed, with varying degrees of clustering and dispersion across the world-city network. But it is not just a case of lifting the city up and spreading it out across the globe across. Rather, it is a matter of considering scale itself as something that is produced rather than as something given: as both relational and relativized (Law, 2000).
Elaborating this argument, the paper is organised into three sections. The first reviews successive attempts to conceptualise world-cityness, further underlining Markusen's argument that 'world city' is a fuzzy concept. We depart from Markusen, however, by concluding that world-cityness can never be defined empirically, only theoretically, and that rigorous conceptualisation means taking on board poststructural arguments about the 'undecidability of meaning, the constitutive power of discourse and the political effectivity of theory and research' (Gibson-Graham, 2000, p. 95). The second section elaborates the consequences of this for the policy-oriented literature that considers how cities can best promote themselves and enhance their competitiveness in a global economy. As is the case in much of the world-cities literature, we argue that this body of work remains similarly fixated on a place-based perspective that suggests cities prosper on the back of their indigenous characteristics (i.e. the presence of 'critical infrastructure' or 'institutional thickness'). The final section thus offers a re-conceptualisation of urban competitiveness in a global space of flows, and makes the case for an urban politics that would be multi-scalar, intransitive, and attentive rather than local, reactive and transitive.2
CONCEPTUALISING WORLD-CITYNESS: FROM ATOMISM TO STRUCTURALISM AND BEYOND
The notion of world cities is by no means new, having first been coined by Patrick Geddes in 1915 to refer to those places where a disproportionate amount of the world's business was conducted. Despite the apparently self-evident nature of this definition, the meaning of the phrase 'world city' has become increasingly convoluted as different features have been considered as evidence of world-cityness. Hence, to qualify as a 'world city,' a city ostensibly needs to possess a disproportionate quantity of some essential attribute.3 This may be assessed against some abstract yardstick on a city-by-city basis (e.g. if the producer-service sector is the largest sector, then city X is a world city) or else determined immanently vis-à-vis other cities (if the producer-service sector of city X is greater than the producer-service sector of N other cities, then city X is a world city). Either way, the disproportionate quantity serves as a critical threshold beyond which a city can be said to qualify as a world city. In this sense, a world city is simply defined as a 'big' city by global standards: large, powerful, important, happening...
Extending this logic, many writers have used the method of comparing quantities as a way to establish a world-wide urban hierarchy, within which one can find or inscribe various levels or degrees of world-cityness (e.g. Singapore may be more of a world city than Hong Kong, but less of a world city than Tokyo). Common attributes for determining world-city status include presence of TNC headquarters, numbers of service-sector employees, proportion of foreign residents, and equity market capitalization. Yet there is little agreement on which attributes are most suitable for measurement purposes. Friedmann (1986), for example, suggests seven indicators of world-cityness: financial assets; transport infrastructure; population size; business services; manufacturing output; TNC headquarters; and the presence of international institutions. Latterly, other researchers have added the presence Internet domain names (Townsend, 2001), public-private partnerships (Kresl, 1995), and measures of cultural vitality (Smith and Timberlake, 1995). Additionally, Short et al. (1996, p. 698) have suggested a wide range of supplemental data sources to redress the data deficiency that is 'the dirty little secret of world cities research.' In their work, the accumulation of new data sets is depicted as a precursor to the adequate specification of world-cityness. Such is the drift from the metaphoric use of fuzzy concepts to the obsessive compulsion of encyclopædism.
However, and despite the complexity of many existing analyses, it is important to note that these quantitative, attributive, and comparative approaches remain fundamentally atomistic: each city is a clearly-defined, bounded spatial entity with quantifiable attributes that are in its exclusive possession. In this regard, a world-city balance sheet of assets and liabilities is much like a company balance sheet: both maintain exact boundaries in order to determine what is included/excluded (i.e., London is not New York, and London bankers are not New York bankers). Following the model of private property, exclusivity holds: of both the subject and its possessions. Only on that basis may one calculate the 'disproportionate' thresholds, levels, and degrees of world-cityness. Consequently, it is not surprising the resulting league tables are invariably framed as a zero-sum game, where one city prospers only at the expense of others.
Above all, then, it is this atomistic conception of world cities that we oppose in this paper. In its place we propose a conceptualization of world cities that is relational and relativized. Nevertheless, as we move towards a full-blown worlding of cities, we will also need to take issue with two forms of structuralism: one that is structuralist in essence; and one that is structural in appearance.
Structuralism: Interpolation and Functionalism
The first form of structuralism derives not from a quantitative analysis of the proportions of various world-city attributes held within cities, but from a qualitative analysis of the world economy as a global structure. In this account, one begins with the structure of the capitalist world system¾the rules of the game, so to speak¾within which regional, national, and local space economies take their place (particularly with regard to their form and function).4 For structuralists, the system is not reducible to either its parts or even the sum of its parts. The system exceeds its manifest content. So, the system of world cities exists over and above the cities that ostensibly constitute it. World cities are a special effect of the structure, interpolated into what Guattari (2000) calls Integrated World Capitalism. Such reasoning often has a functionalist and teleological slant: the effect solicits its own cause; the result engenders its own computation. For example, if the system 'needs' to be commanded, controlled or reflexively monitored, then certain places will come to fulfil those needs. The shift from atomism to structural functionalism is decisive. It is the first step in the worlding of cities. Hereinafter, each city is an effect not of its intra-urban attributes, but of its relative location within an inter-urban system. Such is the move from introversion to extroversion, and from exclusivity to connectivity.
However, one does not need to be a structuralist to recognize that there is clear evidence that world cities have taken on a new strategic importance in the contemporary world, acting as crucial hubs in a complex network of international linkages. As we have argued, this requires a shift of emphasis from an atomistic concern with comparative intra-urban studies to an appreciation of inter-urban connectivity. This enhanced role is primarily interpreted as the result of globalization, a suite of processes involving distanciation (the stretching of all kinds of social, cultural, political, and economic relations across space and time) and time-space compression (the apparent annihilation of space by time as a result of a wide variety of media and telecommunications technologies) as 'speed distance' reconfigures many of the taken-for-granted correspondences between social space and physical distance (Bauman, 1998; Harvey, 1989; Virilio, 1999). Recent interpretations suggest this has ushered in a new spatiality of international capital transactions, transnational migrations and epistemic communities which in many respects circumvent the long-assumed centrality of the nation-state.
On this basis, it has been suggested that we are witnessing the emergence of a truly global political economy characterised by a new geometry of flows that is overcoding the patchwork of local, national, and regional space economies. Influential here is Castells (2000a), who argues that global networks are becoming the new organising and managing principles (or social morphology) of 'informational capitalism,'5 with recent advances in information technology providing the material infrastructure for the diffusion of the networking principle throughout society. Insofar as the logic of networks is spreading into all spheres of human activity, Castells (2000b, p. 9) contends that it is also radically modifying 'relationships of production/consumption; relationships of experience; and relationships of power.' As he describes it, contemporary network society is a space of flows, an entangled knot of increasingly dense linkages, connections, and relations stretched out across space and time: or rather, which reconfigure space and time in their own image (cf. Schivelbusch, 1980). Indeed, many writers have noted with dismay just how difficult it is becoming to envisage a form of capitalism that appears to be increasingly diffuse and virtual (e.g. Buck-Morss, 1995; Gibson-Graham, 1996). For some, 'the world is simply too messy, too fractured, too full of vested interests and competing perspectives to be tidily displayed' (Barnes, 2001, p. 560).
Ignoring such obvious difficulties of making the strange familiar through graphic representation, Castells' texts are illustrated with maps and diagrams that allow us to 'see' the structures¾if not the embodied and disembodied practices¾that connect the nodes of a network society (for a critique of this representational strategy, see Pile and Thrift, 2000). As a means of envisioning a world-city network, Castells' diagrams offer a highly original interpretation of world-city processes, showing the strategic position of cities in a network that links global 'core' and 'periphery.' Yet in many ways, Castells' definition of world cities is quite traditional, following, for example, Friedmann's (1986) identification of London, New York, and Tokyo at the top of a global urban hierarchy. However, according to Taylor (1997), the principal contribution of Castells to the world-cities literature is to shift the root metaphor from hierarchy to network, so that it becomes necessary to conceive of a world-city network that is not reducible to the cities that constitute it (i.e. a network that 'has the capacity to work as a unit on a planetary scale in real or chosen time'- Castells, 2000b, p. 10)6. Castells' writing on the command and control functions enacted by this inter-urban structure thus provides a richer and more comprehensive theoretical context for world-city research, making world cities one important layer within a broader and essentially heterogeneous 'global space of flows.' Such is the spatial logic of the informational age.
Hence, while many accounts of the strategic importance of world cities remain based on an analysis of the characteristics of individual cities (e.g. presence of corporate headquarters, number of banks, stock-exchange turnover, etc.), a new literature on world cities focuses on their connectivity within an inter-urban network that functions as the central nervous system of the global economy (see Sassen, 1991; Storper, 1997). However, we remain sceptical about this structuralist framework. First and foremost, it devotes insufficient attention to how the structure is established and maintained. It is taken as a given, rather than as an ongoing achievement; as axiomatic, rather than performative (see, for example, Sassen, 2000, on the notion of locating cities on the specialised economic circuits that are the 'organisational architecture' for cross-border flows or Castells, 2000b, p. 17, on the 'global financial markets and their management networks' that 'constitute an automated network'). We suggest instead that if there is to be a structure, it needs to be considered as an immanent and aleatory effect of contingent encounters. The central nervous system of global capitalism is created, sustained, and attended to¾not least by the 'capitalocentric' business discourses that propose the existence of a coherent global economy populated by 'fast' managerial subjects (see Gibson-Graham, 1996; Thrift, 2000).
Pointillism: Structuralism in Appearance
With an eye towards the contingency and immanence of structure, there is another form of structuralism that we wish to highlight. If we return to the notion of an excess of form over content (such that the world-city network stands in excess of the sum of world cities), then structure can be defined as that holds good between contexts. Structure comprises the rules of the game: they are always in play but always beyond the reach of play. And yet this does not necessarily mean that structure is eternal and ahistorical. The recurrence that is structure can either be created catastrophically in a once-and-for-all eruption¾as with the case of Castells above7¾or else it can be an ongoing and emergent property that remains in perpetual suspense: just like the Law is for Josef K. in Kafka's The Trial. With the latter we are very close to poststructuralism: structure is set in motion; a variable consistency rather than an immutable constant; always in play and never out of play. But before we make this move, consider the structuralism of GaWC (see especially Beaverstock et al., 1997; Taylor, 2000).8 The 'roster' of 123 world cities that they have drawn up is based neither on a comparative analysis of intra-urban attributes, nor on the interpolation of an arbitrary number of cities into a pre-existing world system endowed with structural imperatives, but on the superimposition of networks of accountancy, advertising, banking/finance, insurance, law, and management consultancy firms. Examining the correlation of the office networks of these advanced producer-service firms¾and adjusting them in order to take account of differences in size and function-has enabled a league table of world cities to emerge that considers both their centrality and nodality in the world-city network (Beaverstock et al., 1997).
At first glance, the GaWC roster approximates to a more traditionally conceived urban hierarchy, with London, New York, Paris, and Tokyo exhibiting the highest degree of global connectivity. Although not surprising, it is noteworthy that all world regions are represented in the roster, albeit with different densities of world cities. This means that there are many cities that have not been spoken of as world cities but which nonetheless are well connected within the three major 'globalization arenas' of Northern America, Western Europe, and Pacific Asia. The resulting variation in regional concentration is quite remarkable: a concrete expression of what has been called 'uneven globalization.'9 So, within Western Europe for example, the fact that Paris and London exhibit an extraordinary degree of global connectivity does not preclude Amsterdam, Barcelona, Frankfurt, and Prague from also playing a significant role in global networks. In other words, the differentiation of world cities is not a zero-sum game. All of these cities are successful not only because they are the pre-eminent cities in their own national space economies, but also because they facilitate and mediate global flows.
This take on world cities thus extends Castells' structural account by considering the way that the 'new' economy requires nodes tied into a global network of financial flows. But, in contrast to Castells' (2000b) account, this new economy is not set in advance, deemed to 'frame' the relationships of production/consumption, experience, and power. Rather, it is seen as the ongoing effect of contingent encounters: structure as pure co-relation. However, since GaWC continues to follow a logic of exclusivity (of both world cities and their attributes), this analysis is still essentially atomistic and positivistic, although it is clearly turned towards structuralist, and perhaps even poststructuralist, ends.10 It prescribes privileged actors (such as the offices of advanced producer-service firms), and depends upon the transitivity of offices, cities, nation-states and the world economy (whereby each is seen to exist at a given scale, so that offices are contained in cities, cities in nations, and nations in the world).
Poststructuralism: Performativity, Contingency, and Intransitivity
From a poststructuralist perspective, it is the movement of heterogeneous materials that constitutes the global economy, a spatial assemblage that is a Body-without-Organs, a figure held together by the forces that compose it, rather than by the functional imperatives of the organism as a whole (Doel, 1995). Doing away with distinctions of inside and outside (e.g. the idea that places are contained within a global system that shapes their destiny), this is an account in which the flow becomes all. To that extent, what is interesting about world cities is not their fixed position-both spatially and functionally-within a structured global network that may be imploding (Virilio, 1999), but the way they bring relations into being: the way flows drift in and drift out, speed up and slow down, contract and expand within them, folding and unfolding space (Doel, 1999). The intransitivity of social and spatial relations is the norm, not the deviation. Indeed, it takes a vast amount of effort to keep things 'in their place,' not least because everything is in a perpetual state of becoming, mutability, and dissemination (Law, 2000). In short, global network formation is always subject to a double movement of speed and slowness, of contraction and expansion, of formation and deformation (Doel, 1999).
This relational materialism demands consideration of the way that human and non-human actants are implicated in the articulation of the global network. Economic geographers have recently argued that the global space economy is less coherent and all-encompassing than might be initially supposed, and is in fact a complex and precarious achievement in need of constant attention and maintenance. For example, Amin (2000) claims that the global economy is the product of the institutional reflexivity that is characteristic of contemporary corporate strategies, knowledges, and practices. This suggests that globalisation is not universal or unitary, but heterogeneous and fragmented: a geography created by the discourses and practices of economic life that stress the need to 'think global' and 'act local' - and vice versa. To put it another way, the global space economy is not a 'given' brought into being by technological innovation and communication revolution. It is emergent¾a 'doing' that, like all space, is 'the articulation of relational performances' (Rose, 1999, p. 248). As Amin argues, global space is not just technological, it is also emotional, organisational, affiliational, and discursive, necessitating an enculturated understanding of political economy (cf. Sayer, 1994; Thrift, 2000). In short, world cities must be continuously performed¾or else they are not.
The realisation that the commonplace activities of everyday life are not residual products of globalisation but are themselves formative of the global is important in encouraging geographers to conceive of space and place in a more imaginative way. While positivist, humanist, behavioural, and Marxist theorists have been guilty of either spatial atomism or spatial pointillism (i.e. treating places as discrete and fixed spatial entities), poststructural takes on globalization analyse the flows and interruptions that individuate portions of the world and thereby lend the global its consistency (Gibson-Graham, 1996). This emphasis on connectivity, flow, and the traffic in things is informed by debates about displacement, movement, and speed, as well as work on actor-networks, performativity, and intransitivity (see Thrift, 1996). Unfortunately, given the tendency for geographers to focus on what exists within places, rather than on what individuates them, there is a serious deficit of data concerning the nature of these flows and connections, and their performativity and mutability. Nonetheless, information about the volume and direction of flows of workers (via international migration), capital (via international banking), products (via import/export), ideas (via broadcasting and the media), and even pollution (via dumping policies) begins to indicate the existence of a heterogeneous space of flows that requires constant attention and translation (see Taylor, 1997).
This take on global space suggests the need for a revised ontology of space-one that escapes taken-for-granted notions of 'local' and 'global,' 'concrete' and 'abstract,' 'fixed' and 'mobile,' and 'embedded' and 'ethereal' to explore space in relational terms. Hence, there is currently much interest in Lefebvre's (1990) notion of irreducible spatial practice, which is embodied, lived, and performed by all manner of human and non-human actors, and de Certeau's (1984) insistence on the incalculability of such spatial practice. These stress that space is produced in an ever-shifting context from materials and processes that are themselves heterogeneous and manifold. Consequently, a key requirement is an ability to translate: to render the incommensurable commensurable. Credit rating, risk management, and multi-jurisdictional law are paradigmatic practices of translation (Beaverstock and Doel, 2001). Indeed, if one were looking for one word to convey the essential quality of world cities, then our candidate would be translators, rather than either commanders or controllers.
With poststructuralism, 'a particular form of urban theory which sees the city as the stamp of great and unified forces which it is the task of the theorist to delineate and delimit is left behind. Instead, the city is seen as 'a partially connected multiplicity which we can only ever be know partially and from multiple places' (Thrift, 1997, p. 143). By emphasising partial, ongoing, and polycentric networking-a process that requires constant effort and attention from numerous actors in a multiplicity of places-the idea of a world city as an enduring performance avoids both the pitfalls of atomistic urban studies, which tend to be either additive or comparative, and the fait accompli that is the hallmark of structuralist urban studies. Such an over-emphasis on the functional differentiation and systemic efficacy of the global space economy occludes both the contingency of the world-city network and the plasticity of networked space-time; not only do network formation and maintenance take time and effort (Beaverstock and Doel, 2001; Bingham, 1996), systemic regularity and functional reproduction are irreducibly unintended outcomes of manifold agents and events. In a related paper (Beaverstock et al., forthcoming) we thus argue that world cities are mediators, translators, and transformers of the global space of flows. However, rather than being self-evidently assured, their strategic position in this network needs to be maintained by a variety of attendants.11 At least four of these appear pivotal-firms, sectors, states, and cities12 (each of which is attended to in turn by a host of human and non-human actors). In the remainder of this paper we focus simply on the latter, exploring the way that city networks may attend to the global space of flows so as to influence the position of a particular city within this network. As we will see, a world city is always a work in progress. Its status is always subject to the clause: 'until further notice.'
TAKING WORLD CITIES SERIOUSLY? CITY COMPETITIVENESS AND PLACE PROMOTION
Insisting that the dominant mode of thinking about the global economy is as a patchwork of place-based economies in a constant struggle for competitive advantage over and against one another-spatial atomism¾Amin (2000) suggests that a widespread reading of urban economic success is that it arises from the spatial agglomeration of know-how and capacity in the city. This is certainly evident in the mainstream literature on the competitiveness of cities, which, via business gurus like Michael Porter, suggests that competitive advantage arises from the strategic manipulation of local assets. For example, and somewhat counter-intuitively, Kresl (1995, p. 54) is adamant that external aspects-both national and international-must be excluded from any analysis of the determinants of a city's competitiveness. Likewise, he insists that 'a city's international competitiveness is quite different from the concept of an international city.' While Kresl argues the latter concerns connectivity (at the international scale), the former concerns only the city in question: in his view, a city can be extremely competitive without being connected into a network of other cities, just as a city can be fully plugged into a network without being at all competitive. This is why Kresl (1995, p. 52) can claim that 'a city may dramatically increase its competitiveness, even its international competitiveness, without being or increasing the degree to which it is an international city.'
For Kresl and others (e.g. Duffy, 1995; Oatley, 1998), city competitiveness and success unquestionably derive from the internal characteristics of a city. Similarly, Deas and Giordano (2001, p. 1413) claim that the source of urban competitiveness is the 'initial stock of assets present in a geographical unit,' and that its outcome is 'the result of firms to exploit these assets.' This form of spatial atomism exemplifies the oxymoronic notion of 'hierarchies without networks' which is implicit in the vast majority of work on city competitiveness, even when the context is explicitly global. Likewise, in the influential work of Krugman, which dismisses the notion of global competition and holds to the idea of strategic economic complementarity, urban and regional success derive from agglomeration economies and the benefits of untraded interdependence at the local scale (Boddy, 1999). Hence, the idea that building up economic infrastructure is essential for economic success is endemic among those groups that act and speak for the city-those 'growth coalitions' or 'urban regimes' that characteristically unite both private and public sectors in a pro-growth politics. In the account offered by Deas and Giordano (2001, p. 1413), these are the actors who are best able to intervene between sources of competitiveness and their outcomes: those 'who are able to create, exploit, supplement and replenish city asset bases, and transform liabilities into assets.'
Although some growth coalitions are more pro-active than others in this regard, most are allocating increasingly large budgets for promoting the city as a favourable environment for business and commerce (see Hall and Hubbard, 1996). There are often remarkable similarities in the key elements of such promotional policies, leading many to talk of a generic 'entrepreneurial' approach to place promotion (Gold and Ward, 1994). Indeed, almost every 'wannabe' world city now has its requisite series of promotional pamphlets, posters, and other cultural products communicating selective images of the city as an attractive, hospitable, and vibrant international city in which to live and work. What is also increasingly evident is that this marketing of place seldom restricts itself to extolling the existing virtues of a given city, but seeks to re-invent the city as a knowledge-rich and culturally-diverse technopole by manipulating its indigenous assets.
This conscious manipulation and promotion of the city, evocatively termed 'imagineering' by Rutheiser (1996), has been subject to close examination from academics who have pointed out the ways in which city identities are sanitised, commodified, and distorted in accordance with the perceived demands of the global marketplace (see especially Gold and Ward, 1994). Indeed, given the pervasive nature of place marketing and other simulacra, it might be argued that it no longer makes much sense to distinguish between the myths and realities of the city, as the images of the city incorporated in the promotional brochures, adverts, guidebooks, and videos come to define the essence of the city as much as the city itself. Nonetheless, the large-scale physical redevelopment of significant portions of the city has taken centre stage in this process of enhancing the city's image. No aspiring world city has been seen as complete without its requisite signature-architect designed spaces. Short (1999, p. 45) describes the 'desperate scramble for big-name architects, art galleries and cultural events' as 'a fascinating part of the place wars amongst cities aiming for the top of the urban hierarchy.' This has been particularly evident in the construction of developments (variously described as 'flagships,' 'urban megaprojects' or 'prestige projects') that have aimed to emulate the perceived success of the rejuvenation of Baltimore's inner harbour in the 1960s (Olds, 1995). The names of these monumental spaces and buildings have quickly become synecdoches for the cities in which they are located, and the quintessence of their world-cityness: London's Canary Wharf, Barcelona's Olympic Marina, Paris' La Défense, Vancouver's Pacific Place, New York's Battery Park, Atlanta's Peachtree Center, Sydney's Darling Harbour and so on. Baudrillard (1990a) argues these are the defining buildings of the late twentieth century-the 'monsters' of the city. Almost without exception, these projects are inherently excessive, spectacular, and wasteful. Individually and collectively, they promote the city through a strategy of seduction and symbolic exchange (Baudrillard, 1990b, 1990c, 1998). They do not so much re-present the city as issue a challenge to other cities. The logic of prestige is based on provocation, outbidding, the escalation of stakes, and¾at the limit¾sacrifice (which is why prestige products are ultimately inimical to both economic calculation and the logic of production and accumulation. They are fundamentally expendable).13
Together with mega-events such as World Expos, City of Culture celebrations, and, perhaps most significantly, the Olympic Games (where even the bidding process has become a major marshalling point for urban boosterism and civic peacockery), the transformation of the fabric of cities via Prestige Projects has been interpreted as a fundamental means by which city governors have attempted to provide previously industrial cities with a new post-industrial identity geared to the needs of a globalised economy (Short, 1999). To date, academic attention has been devoted to documenting the largely deleterious social, cultural, economic and environmental impacts of these projects at the local scale. This critique has, for instance, been articulated in Harvey's (1989) structural reading of entrepreneurial governance when he talks of the zero-sum game resulting from the drive for place differentiation (via the promotion of local character) versus the standardisation implicit in the notion of 'global' development. In his estimation, the serialisation of Prestige Projects bequeaths no significant competitive advantage to any particular city, merely fuelling further speculative development which is of negligible import in terms of assuaging economic problems at the local level. In Harvey's thesis, it is the fear of losing (apparently locally-embedded) capital that triggers the urge to develop Prestige Projects, with each city feeling obliged to create a succession of projects just to 'keep up.' Accordingly, many of these projects have been shown to be financially unviable in their own terms, and have been propped up by public money on the (unproven) understanding that such projects nonetheless are a crucial lure for global investors (Loftman and Nevin, 1996). Like A Field of Dreams, Prestige Projects rest on the ultimate act of faith: 'build it and they will come.'
RE-CONCEPTUALISING URBAN COMPETITIVENESS IN A GLOBAL SPACE OF FLOWS
Considering practices of place promotion in the light of the emerging world-city literature challenges the simple assertion that manipulating the assets of an individual city can provide the basis for economic success and world-cityness. To recapitulate our relational perspective on world-city formation; a world city is not defined by its agglomeration of particular businesses, knowledgeable individuals or economic infrastructure, nor secured by the development of Prestige Projects, nor brought into being by the demands of systemic and functional imperatives. Viewed from a poststructuralist perspective, a world city is less a site whose strategic importance and success in the global economy can be determined by its connectivity, than a configuration of heterogeneous materials and practices that attend to the intersection, bifurcation, and cultivation of innumerable flows: especially practices that allow the incommensurable to be rendered commensurable and the incompossible to be made compossible (everything from data processing and econometric analysis to risk assessment and multi-jurisdictional law). Simply put, a world city should be thought of as a division of labour (which embraces both human and non-human actors) and a constellation of practices (which may or may not have coherence, consistency, and efficacy) rather than as a portion of the Earth per se. It is the sheer volume of flows (of money, ideas, knowledge, and people) that are attended to in world cities that bequeaths them their pre-eminent role in a global space of flows. This contrasts with the assumption underpinning entrepreneurial efforts at place promotion that 'hypermobile' capital can be embedded in the locality through policies designed to enhance the attractiveness of the city.
Acknowledging that world cities are not bounded phenomena forming 'islands of economic competitiveness' (Amin, 2000, p. 18), but networked sites of translation thus undermines the spatial logic of fixity and embedment that underpins entrepreneurial place promotion. The alternative is to see each individual city as necessarily open to a constitutive outside and perpetually in process. When one stops thinking of cities as fixed places with tightly-circumscribed administrative boundaries, and starts thinking of spaces of flow, the notion of city competition changes. Here, the urban economy is no longer reducible to its locally-embedded resources, but concerns the flows that it works on. To talk of these flows benefiting one city to the detriment of all others is clearly non-sensical. From both structuralist and poststructuralist perspectives, these various 'centres of calculation' do not primarily compete with each other since there needs to be collaboration and a division of labour for the global space economy to 'work.' Sassen (1991) illustrates this with reference to the financial deregulation of the 1980s. This occurred at a time when the relation among the three major centres-London, New York, and Tokyo-was seen as one of straight competition (so that the 1986 Financial Deregulation Act, for example, was supposed to enhance the competitiveness of the City of London). Sassen suggests that this deregulation favoured no one city over the others, but enabled all three to enhance their position in the global space of flows. Likewise, New York's domination of investment banking and fund management may very well have been considerably more important than economic growth within the UK space economy for strengthening London's position in the world-city network. Further, she suggests that just because London became a more important hub in a global space of flows in the 1980s does not mean that cities like Bristol or Birmingham (in the UK) became less important. To the contrary, the enhanced importance of London in the network meant that many other cities were more fully integrated in the European globalization arena.
The fact that there seems to be important synergies and linkages between cities in the global space of flows does not mean that the notion of competition disappears altogether (as Krugman implies). Indeed, it is relevant to talk of competition in the sense that the position of cities in the world-city network can change for better or worse. For example, it is apparent that Frankfurt and Miami become more significant as translators and mediators of flow at the same time that Milan and Detroit have become less strategically important (Taylor, 1997). This, as we suggested above, is a conjoint but unintended outcome of the actions of four key attendants of the global space economy-firms, sectors, states, and cities. Within geography, most attention has been devoted to the former, with the locational decision-making of TNCs often depicted as the key factor shaping global flows through the world-city network (Dicken, 1992). But this should not overshadow the work performed by state organisations (whether national or international), sectoral organisations (organized labour and business associations) and indeed cities themselves in attending to the world-city network.
What we are suggesting here-that the city can attend to its own position in a global space of flows-may initially seem confusing. This may be clarified if we distinguish between the world city as a site in a world-wide network of heterogeneous flows and the world city itself as a heterogeneous assemblage of practices, materials, and actants drawn from within and without the city limits. By taking world cities literally, we arrive at the conclusion that there is a potentially world-wide network that attends to the relative importance of a particular city in a global space of flows. This 'city' network comprises a multiplicity of groups and individuals who feel they have a stake in the economic prosperity of a specific city. In a seminal analysis of how networks of actors make and carry out governing decisions for the city, Stone (1989) suggested that these 'urban regimes' characteristically include property interests, rentiers, utility groups, trade unions, universities, and 'local' businesses. Consequently, when one refers to London's city network, one would routinely include local institutions and agencies such as the City Corporation, the London Assembly, the London Boroughs, the Cross River Partnership, the London Pride Partnership, and so on. But, in addition, one must follow Cox and Mair (1988) by including groups and individuals who are not necessarily based within that city, and who may well be scattered across the globe: network translators may act at a distance. Thus, one must remember that world-city networks may involve the epistemic communities, businesses, and knowledge-rich individuals that are often described as being disembedded from the cities in which they work, rest, and play. Far from being hypermobile, these 'fast subjects' necessarily exist in city networks which embody circulation and provide the basis for the creation of new business practices (Thrift, 2000). Yet there is no guarantee that a given city network will provide a successful basis for commanding and controlling: the architecture of this city network is therefore often flimsy, requiring constant effort to maintain it. Nonetheless, the agency of the city (i.e. its evidential ability to attend to the world-city network) results from the work performed by this much broader constellation of actors.
So, the suggestion that world cities are irreducibly networked phenomena does not mean that Prestige Projects and place promotion cannot play a role in enhancing the position of a city in the world-city network. To the contrary, we suggest that city networks can be pro-active in manipulating the global space of flows to the benefit of particular cities via such projects. Yet to do so it is crucial that these city networks replace their place-based way of thinking with a focus on connectivity, performance, and flow. Understandings of both the local and the global need to change, and city networks need to extend through time and space as they seek to perform world-cityness. Hence, for a city to become plugged into the world-city network the local authority needs to use such projects to establish connections and manipulate flows. As an analogy consider the construction of an international airport as a means to enhance a city's place in the world-city network. Building a fully-functioning airport is not enough. For it to work one must also attend to and manipulate existing flows of air traffic by establishing flight routes, making connections with international carriers, liasing with air traffic controllers, securing international freight, passengers, and transfers, etc. This means the work involved in creating connectivity-and urban success-may be widely dispersed. On the other hand, it might not (cf. Thrift, 2000).
Our argument, then, is not that work beyond the confines of the city needs to be prioritised over work within. Nor are we arguing for a 'geographical imagination which can look both within and beyond the city and hold the two things in tension' (Massey, 1998, p. 191). Rather, we emphasise a need for policy-makers and academics alike to revise understandings of inside and outside, local and global, near and far. Instead of a place-based politics of competition, we hold to a distributed politics of flow whose degree of concentration, dispersion, consistency, and efficacy will be both contingent and context dependent. Prestige Projects may play a role in this, but what seems to be badly lacking at the moment is any monitoring of the way projects succeed in establishing flows and the work they do in enrolling people, ideas, products or capital into city networks.
Crucial here is the idea that these projects are aggressively marketed to an international audience through their 'overexposure' in the international media (Virilio, 1999). This relies on them being projected and mediated by global advertising, public relations, and promotions executives who ensure that these projects and events are internationally recognised and acknowledged (Short, 1999). Yet assessments of multiply-mediated global spectacles like London's Millennium Dome, World Expos, and Olympic Games focus overwhelmingly on local reception, national mediation, and job-creation rather than on international projection and success in attending to global flows (see Loftman and Nevin, 1996). The bottom line with any Prestige Project is not how it connects with the local economy and the local community, but how it enhances global connectivity and collaboration. If we accept this argument, we can begin to see that there is no necessity for a Prestige Project to be within a particular city for it to enhance that city's position in the world-city network. A more successful policy may well be to establish a series of projects elsewhere (e.g. a series of Millenium Domes in Mumbai, Islamabad, Chicago, Beijing, and so on). This would immediately involve movements of capital, investment, people, and commodities, creating new global flows and bringing new actants into the city network.
On the surface, encouraging city governors to invest in developments that are not even sited in their locale seems to fly in the face of (capitalist) reason. To make this suggestion is to offer a wider critique of an urban politics obsessed not only with how the city is performing against its 'competitors' but also with the 'petty fear' of losing its prized assets and belongings: jobs, capital, people, knowledge, prestige, etc. Needless to say, both of these fuel uneven development insofar as they devalue some places in relation to others. If city governors begin seeing other cities as potential collaborators rather than competitors, then the possibility of a more radical form of urban politics may emerge. Instead of seeing their role as attracting a share of footloose 'global' capital to benefit a 'local' population (or an élite segment of that population), urban policy-makers would be responsible for managing flows and connections in collaboration with other city governors. Like the flows that make up the global economy, urban politics need to become both truly transnational and fully intransitive.
Signs of this new politics of flow, and recognition that cities can be marketed 'elsewhere,' are emerging. For example, during a week-long promotion of Wales in the USA entitled 'Festival of Wales 2001,' Rhodri Morgan, the Welsh Assembly First Minister, addressed a meeting of the Welsh American Society of Northern Californian in San Jose, where he called upon Welsh ex-patriates for their assistance: 'We want to recruit you as unpaid ambassadors for Wales.'14 It was here, in the heart of Silicon Valley, that a new, multi-lingual promotional website for Wales was launched: 'Wales. World Nation' (URL <http://www.walesworldnation.com>). The main image on the site's homepage is centred on a composite Welsh and American flag set against a map of North America. 'Wales in Europe' takes precedence over 'Wales in a devolved UK.' The website is available in English, French, German, Japanese, and Spanish¾ but not Welsh.
As a symptomatic effort at attending to global flows we cannot comment as yet on the success of either the Festival of Wales or Wales: World Nation. Nor can we, in the absence of critical geographic scrutiny, ascertain the success of Hannover 2000 in establishing mini-expo sites in other cities; London Fashion Week in relocating to New York; the New York Yankees opening stores in Manchester, and so on. Yet questions as to how best to manipulate and to attend to global flows-and how to perform world-cityness-are currently high on the political agenda in the wake of the attack on New York's World Trade Center. The obliteration of this building form the skyline of Manhattan, and the repercussions this has had for both the image of New York and world-wide financial markets, seems to point overwhelmingly to the importance of such place-based developments in providing an infrastructure geared to the global economy. The destruction of this building-and the loss of the knowledge-rich individuals it housed, many of whom were actively attending to other world cities, such as London, as much as they were attending to New York15-led to immediate questions about the future viability of New York as the pre-eminent world city, with many proposing the need for a new global financial network in which the city would play a diminished role. Responding to this, and emphasising a 'business as usual' rhetoric, policy-makers, including Mayor Giuliani, have stated their intention to ensure this does not happen. Redeveloping Ground Zero may be important for repairing the flows and connections that were so suddenly ruptured, but as we have already seen from New Jersey to Kabul, New York's world-cityness also needs to be worked at elsewhere.
We no longer live in a space of places, but a global space of flows. Extending this logic, and elaborating an ontology of space that holds to the idea that world cities are networked phenomena, this paper has sought to clarify the consequences of this for urban competitiveness. This has made clear that the city is not a localised place of economic competitiveness (i.e. a hub in this space of flows), but is a constellation that lends consistency to these flows, working on their heterogeneity. Following this, it is suggested that cities can only enhance their competitiveness by recognising that world-cityness is not determined by a city's location in a pre-existing structure, but needs to be performed and worked at. It is this work that creates the global space of flows, a complex and contingent achievement that requires constant attention. This becomes most evident in moments of urban crisis, when the role of a city in translating flows is brought into question (for example, the financial crises that have swept through a host of emerging markets in Central Europe, East Asia, and Latin America, as well as Japan's ongoing banking catastrophe). But the network is always becoming, and there will always be winners and losers (i.e. cities that are becoming more or less important). However, our conclusion is that the principal losers will be those who imagine themselves to be in competition with the winners, and focus on manipulating 'indigenous' assets rather than attending to the global space of flows. Collaboration, not competition, is the key to urban success.
We would like to thank the Economic and Social Research Council for supporting this project (Award number: R000222693). We are also grateful to the Deutscher Geographentag, who invited us to present a previous version of the paper at the Stadt und Region conference, Leipzig.
Amin, A. (2000) Spatialities of globalisation Global Conference on Economic Geography, National University of Singapore, 5-8 December.
Barnes, T. (2001) 'Retheorising economic geography from the quantitative revolution to the cultural turn', Annals of the Association of American Geographers 91, pp. 546-565.
Baudrillard, J. (1990a) Cool memories. London: Verso.
Baudrillard, J. (1990b) Fatal strategies. London: Pluto.
Baudrillard, J. (1990c) Seduction. London: Macmillan.
Baudrillard, J. (1998) The consumer society: myths and structures. London: Sage.
Bauman, Z. (1998) Globalization: the human consequences. Oxford: Polity.
Beaverstock, J. and Doel, M. (2001) 'Unfolding the spatial architecture of the East Asian financial crisis the response of global investment banks', Geoforum 32, pp. 15-32.
Beaverstock, J., Doel, M., Hubbard, P. and Taylor, P. (2002) 'Attending to the world competition, co-operation and co-efficiency in the world city network', Global Networks forthcoming.
Beaverstock, J., Taylor, P. and Smith, R. (1999) 'A roster of World Cities', Cities 16, pp. 445-458.
Bingham, N. (1996) 'Object-ions from technological determinism towards geographies of relations', Environment and Planning D Society and Space 14, pp. 635-658.
Boddy, M. (1999) 'Geographical economies and urban competitiveness: a critique', Urban Studies 36, pp. 811-843.
Bromley, S. (1999) 'The space of flows and timeless time: Manuel Castells's The Information Age' Radical Philosophy 97, pp. 6-17
Buck-Morss, S. (1995) 'Envisioning capital political economy on display', Critical Inquiry 21, pp. 434-467.
Castells, M. (2000a) The rise of the network society, second edition. Oxford: Blackwell.
Castells, M. (2000b) 'Materials for an exploratory theory of the network society', In British Journal of Sociology 51, pp. 5-24.
Cox, K. and Mair, A. (1989) 'Levels of abstraction in locality studies' Antipode 21, pp. 121-132.
Deas, I. and Giordano, B. (2001) 'Conceptualising and measuring urban competitiveness in major English cities', Environment and Planning A 33, pp. 1411-1429.
Derrida, J. (1988) Limited Inc. Evanston, IL: Northwestern University Press.
Dicken, Peter (1992) Global shift. London: Arnold.
de Certeau, M. (1984) The practice of everyday life. Berkeley: University of California.
Deleuze, G. and Guattari, F. (1984) Anti-Oedipus: capitalism and schizophrenia. London: Athlone.
Deleuze, G. and Guattari, F. (1988) A thousand plateaus: capitalism and schizophrenia, London, Athlone.
Doel, M. (1995) 'Bodies without Organs: deconstruction and schizoanalysis', in Pile, S. and Thrift, N. (eds) Mapping the subject: geographies of cultural transformation. London: Routledge.
Doel, M. (1999) Poststructuralist geographies: the diabolical art of spatial science, Edinburgh University Press: Edinburgh.
Duffy, H. (1995) Competitive cities succeeding in the global economy. London: Spon.
Fainstein, S. (1994) The city-builders property and planning in New York and London. Oxford: Blackwell.
Friedmann, J. (1986) 'The World City hypothesis', Development and Change 17, pp. 69-83.
Gibson-Graham, J.K. (1996) The end of capitalism (as we knew it) a feminist critique of political economy. Blackwell: Oxford.
Gibson-Graham, J.K. (2000) 'Poststructural interventions', in Barnes, T. and Sheppard, E. (eds) A Companion to Economic Geography. Oxford: Blackwell.
Gold, J. and Ward, S. (eds) (1994) Place promotion the use of publicity and marketing to sell towns and regions. Chichester: John Wiley.
Guattari, F. (2000) The three ecologies, London: Athlone.
Hall, T. and Hubbard, P. (1996) 'The entrepreneurial city new urban politics, new urban geographies?', Progress in Human Geography 20, pp. 153-74.
Harvey, D. (1989) 'From managerialism to entrepreneurialism the transformation of governance in late capitalism' Geografiska Annaler 71B, pp. 3-17.
Kafka, F. (1992) Kafka the complete novels. Minerva: London.
Kresl, P. (1995) 'The determinants of urban competitiveness a survey', in Kresl, P. and Gappert, G. (eds) North American cities and the global economy, Beverly Hills: Sage.
Law, J. (2000) 'Transitivities', Environment and Planning D Society and Space 18, pp. 133-148.
Lefebvre, H. (1991) The production of space. Oxford: Blackwell.
Loftman, P. and Nevin, B. (1996) 'Going for growth prestige projects in three British cities', Urban Studies 33, pp. 991-1019.
Lyotard, J.F. (1993) Libidinal economy. London: Athlone.
Markusen, A. (1999) 'Fuzzy concepts, scanty evidence and policy distance the case for rigour and policy relevance in critical regional studies', Regional Studies 33, pp. 869-884.
Massey, D. (1999) 'On space and the city', in Massey, D., Allen, J. and Pile, S. (eds) City Worlds, London: Routledge.
Murdoch, J. (1997) 'Inhuman/nonhuman/human actor-network theory and the prospects for a nondualistic and symmetrical perspective on nature and society', Environment and Planning D Society and Space 15, pp. 731-756.
Newman, P. and Thornley, A. (1997) 'Fragmentation and centralisation in the governance of London', Urban Studies 34, pp. 967-988.
Oatley, N. (1998) Cities, economic competition and urban policy. London: PCP.
Olds, K. (1995) 'Globalization and the production of new urban spaces Pacific Rim megaprojects in the late twentieth century', Environment and Planning A 27, pp. 1713-1743.
Pile, S. and Thrift, N. (2000) 'Technical note', in Pile, S. and Thrift, N. (eds) City A-Z, London: Routledge.
Rose, G. (1999) 'Performing space', in Massey, D., Allen, J. and Sarre, P. (eds) Human geography today, Cambridge: Polity.
Rutheiser, C. (1996) Imagineering Atlanta. London: Verso.
Sassen, S. (1991) The global city. Princeton: Princeton University Press.
Sassen, S. (1994) Cities in a world economy. Thousand Oaks: Pine Forge.
Sassen, S. (2000) 'Locating cities on global circuits' GaWC Research Bulletin 46 URL <http//www.lboro.ac.uk/gawc/rb/rb46.html>
Sayer, A. (1994) 'Cultural studies and "the economy stupid."', Environment and Planning D Society and Space12, pp. 635-637.
Schivelbusch, W. (1980) The railway journey: the industrialisation of time and space in the 19th century. Berkeley: University of California.
Short, J., Kim, Y.H., Kuus, M., Wells, H. (1996) 'The dirty little secret of World City research-data problems in comparative analysis', International Journal of Urban and Regional Research 20, pp. 697-719.
Short, J. and Kim, Y.H. (1998) 'Urban crises/urban representations selling the city in difficult times', in Hall, T. and Hubbard, P. (eds) The entrepreneurial city, Chichester: John Wiley.
Short, J. (1999) 'Urban imagineers', in Jonas, A. and Wilson, D. (eds) The urban growth machine critical perspectives two decades later, Albany: State University New York.
Smith, D. A., and Timberlake, M. (1995) 'Conceptualising and mapping the structure of the world systems city system', Urban Studies 32, pp. 287-302.
Stone, C. (1989) Regime politics governing Atlanta. Lawrence: University of Kansas.
Storper, M. (1997) The Regional World: Territorial Development in a Global Economy. New York: Guilford Press.
Taylor, P. (1997) 'Hierarchical tendencies amongst world cities a global research proposal', Cities 14, pp. 323-332.
Taylor, P. (2000) 'World cities and territorial states under conditions of contemporary globalization', in Political Geography 19, pp. 5-32.
Thrift, N. (1996) 'New urban eras and old technological fears reconfiguring the goodwill of electronic things', Urban Studies 33, pp. 1463-1493.
Thrift, N. (1997) 'Cities without modernity, cities with magic', Scottish Geographical Magazine 113, pp. 138-149.
Thrift, N. (2000) 'Performing cultures in the new economy', Annals of the Association of American Geographers, 90, pp. 674-692.
Townsend, A. (2001) 'The Internet and the rise of the new network cities', Environment and Planning B Planning and Design 28, pp. 39-58.
Virilio, Paul (1986) 'The overexposed city', in Fehrer, M. (ed) Zone 1, New York: Zone.
1.Although there is some resonance, this definition departs from the more common treatment of 'fuzzy concepts' as chaotic, rather than merely polysemic, conceptions or abstractions. The phrase 'unfuzzy concepts' is our own. Although she refers to 'concepts,' 'new concepts,' 'valid concepts,' and a process that would 'pare fuzzy concepts down to a sharp and clear profile,' Markusen only attempts to pin down 'fuzzy concepts' and even 'fuzzier concepts' (Markusen, 1999: 873). All remain polysemic. So much for 'conceptual clarity' and 'plainly stated theories' (Markusen, 1999: 870 and 881).
2. Transitivity should be understood as referring to the logic of a relation, such that if a relation holds between the first and second terms, and also between the second and third, then it necessarily holds between the first and the third. However, since scale is a social construct-something that is produced rather than something that is given-intransitivity is the norm, not the deviation.
3. Even if one goes along with this definition, virtually every word is up for grabs.
4. Much of the world-cities literature is strongly associated with Dependency Theory, World Systems Theory, and other forms of global structuralism.
5. A more appropriate term might be 'infantile capitalism,' not least because it re-inscribes capital within a libidinal economy of desiring production. 'Informational capitalism,' on the other hand, tends to dissimulate the subjectivity of capital (Lyotard, 1993). Deleuze and Guattari have done much to understand capitalism as a semiotic regime, whose principle output is not material products but subjectivities (Deleuze and Guattari, 1984, 1988; Guattari, 2000).
6. Ironically, Bromley (1999, p.13) points out that 'Castells says remarkably little about the nature of networks'; in over 1,200 pages of text, he makes only a few passing references.
7. It is no coincidence that Castells opens The rise of the network society with a reference to catastrophism over and against gradualism. 'My starting-point, and I am not alone in this assumption, is that, at the end of the twentieth century, we lived through one of these rare intervals in history. An interval characterized by the transformation of our "material culture" by the works of a new technological paradigm organized around information technology' (Castells, 2000a, p. 28). This quotation demonstrates that Castells' structuralism is rooted in technological determinism. Unlike his earlier work, Castells now regards the Subject of History as a unilinear technological imperative rather than as an antagonistic social relation. In this regard, the Information Age follows on from the Stone Age, the Iron Age, and the Machine Age.
9. Hence Castells' (2000b, p.10) argument: 'Globalization is highly selective. It proceeds by linking up all that, according to dominant interests, has value anywhere in the planet, and discarding anything (people, firms, territories, resources) which has no value or becomes devalued, in a variable geometry of creative destruction and destructive creation of value.'
10. 'Pointillism' is perhaps a better term that 'atomism' in this regard, since while atomism suggests that the whole is entirely reducible to its parts¾and in that sense the whole is an unnecessary illusion¾, pointillism does not. With pointillism, the juxtaposition of elements gives rise to a structural effect or structuralist impression. Whence the ambivalence of the GaWC approach: it is structuralist in appearance rather than in essence.
11. Sassen (1991) prefers to talk of 'articulators,' amongst whom she identifies advanced-producer service firms as the principal articulators of the world-city network.
12. Sectors embrace both business associations and organized labour, while states also include numerous international institutions.
13. Obviously, the logic of prestige expressed through symbolic exchange is not restricted to massive urban redevelopment projects. It is also expressed in public art, and even in graffiti, vandalism, and burn-out cars. All of these practices refuse the rational calculations of instrumental reason and structural functionalism.
14. 'Morgan tells US "promote homeland"' BBC News Online (04/09/01). URL <http://news.bbc.co.uk/hi/english/uk/wales/newsid_1523000/1523378.stm>
15. Of course, it needs to be remembered that many, perhaps most, of those affected were part of a servicing rather than service class, individuals who played a significant and crucial role in the performance of New York as a world city (cf. Allen and Pryke, 1994; Sassen, 1994).
Edited and posted on the web on 26th April 2002
Note: This Research Bulletin has been published in City, 6 (3), (2002), 351-368