GaWC Research Bulletin 44

I use this title of Wallerstein's latest collection of essays² as my starting point. He employs this phrase in two senses: that the modern world-system is approaching its demise, and that the salience of our knowledge of the system is therefore reaching its sell-by date. As we enter a phase of increasing uncertainty everything is up for grabs as it were, not least in world-systems analysis itself. It is in this spirit that the argument below is developed. It seems to me that if indeed the world is in a process of fundamental transition then the familiar cast in the modern world-system story - households, classes, nations, states - will need revisiting. No doubt these four institutions will continue to be salient in their various combinations but they will be subject to challenges from other social arrangements that can arise in the new circumstances. In particular the nation-state is commonly viewed as being under threat in some sense from globalization. Within this context I will formulate a challenge to conventional world-systems thinking by bringing cities and their networks to centre stage. It is a challenge because cities, in any major contemporary form - mega-cities, metropolitan centres, world cities, global cities - are conspicuous by their absence in Wallerstein's essays³ and indeed in other arguments about the future from a world-systems perspective.⁴ 

IMPLICATIONS OF RECASTING

Casting cities into a starring role in a world-systems analysis of the transition produces some new and perhaps surprising outcomes. I provide a brief reprise here to show where the argument leads. I must begin, however, by stressing that I am not in the business of providing atheoretical futuristic speculation of the kind that can be found in the cities literature.⁵ Any 'outlandish' ideas are derived from within the parameters of world-systems thinking and their interest is not to be found simply in their unusual nature. The important point is that they are derivable through world-systems analysis once we change the cast. For instance, the argument concludes with a position that says we have never been capitalist.⁶ The modern world-system is interpreted as a unique hybrid situation where politics and economics have been 'balanced', mutually supporting each other. This has been a political economy, no less but also no more, so that the traditional 'man on horseback', the dominant power for 5,000 years, has retained much of his power throughout the 'modern era'. As may be implied from this, I will argue that rather than politics being the 'missing link' in world-systems analysis,⁷ we have remained too state-centric in our thinking despite the change in object of study from society-state to world-system. This accounts for the relative neglect of cities in most world-systems analysis once the story gets beyond seventeenth century Dutch hegemony. However focusing on cities as networks in the contemporary world-economy suggests that despite increasing material polarisation in the world it is the core category within the world-economy, not the periphery, that is most in danger of disintegration. I will argue that there are current processes within world cities which are leading to a conversion of core into semi-periphery. Hence the prediction that demise of the modern world-system will be manifest in the elimination of the core. To reach these alternative world-systems positions requires a serious consideration of contemporary globalization. 

THE IMPORTANCE OF GLOBALIZATION

Any serious consideration of contemporary globalization within a dissident/critical/radical/emancipatory analysis leads to a risk of being accused of 'celebrating' global capitalism. This is understandable given the pedigree of the concept⁸ and the tendency to equate globalization with global neo-liberalism. There is also the shallowness of much globalization writing where world-systems analysis is treated as a precursor⁹ and Wallerstein has a role as a sort of 'honorary pioneer"¹⁰. Notwithstanding this indictment, contemporary globalization should not be so easily dismissed. There are many different positions on the contemporary salience of globalization which range from 'hyperglobalists' who think we have entered a new global age to the sceptics who see only a continuation of 'internationalisation' with nation-states still very much in control.¹¹ I am going to take a geohistorical position that treats globalization as a fundamental change to the way in which we view the world. It represents an alteration in the geographical scale of how social change is widely understood and as such presents a broad challenge to the embedded statism of the social sciences. This is most certainly not a new global era separate from what we have described as the modern world-system but it is a key break in how that social world is conceptualised. I am going to interpret globalization as a metageographical transition but before I can do this the nature of the spaces in globalization need attending to. 

THE GLOBAL SPACE OF FLOWS

There is one prominent radical theorist who does treat contemporary globalization as a fundamental alteration in the way society operates. Castells' 'network society' is the most influential new conceptualisation of society at a global scale.¹² He argues that there has been a fundamental change in the nature of space with a new space of flows replacing traditional spaces of places. Related to the rise of new enabling technologies in computing and communication, network society has developed since the 1970s in a new informational age. As Arrighi has pointed out,¹³ there were, of course, many spaces of flows operating in the capitalist world-economy before the 1970s so that identifying network society as a wholly new social form is problematic. Nevertheless, Castells' argument does inform the question of globalization and the state: the latter represents a mosaic-based society (space of places) which is being challenged by a new network-based society (space of flows). My interpretation is that globalization represents a new balance between these spatial formations with spaces of flows coming to dominate spaces of places. New global spaces of flows take several forms; the 'most direct illustration' is, according to Castells,¹⁴ the world city network.

SERVICING GLOBAL CAPITAL: THE WORLD CITY NETWORK

In Friedmann's seminal 'world city hypothesis'¹⁵, it is postulated that a global hierarchy of cities developed as 'command and control centres' housing the headquarters of multinational corporations who were in the business of creating of a new international division of labour. Subsequent researches have shown that the key characteristic of these world or global cities is their concentration of advanced producer services.¹⁶ More than just international financial centres, these cities are the locales for the production of knowledge-rich service products such as in inter-jurisdictional legal services, in place-sensitive international advertising campaigns, and in many new financial instruments. World cities are defined in terms of the critical masses of creative and professional labour organised through global service firms. The latter are crucial because they create the connections between world cities through their office networks. Hence the archetypal skylines of office tower blocks are the concrete landscapes of senders and receivers in global spaces of information flows. In aggregate these service firms have created world cities as the nodes of a global network of cities.¹⁷ Figure 1 shows the global connectivity of the leading123 world cities measured in terms of networks of accountancy, advertising, banking/finance, insurance, law, and management consultancy firms.¹⁸ Although it is not surprising to see London and New York with the highest connectivity, it is noteworthy that all world regions are represented on the map, albeit with different densities of world cities across regions. This is a unique spatial structure. To be sure, Braudel¹⁹ identified 'world-cities' in the past but these have been singular 'hegemonic cities'²⁰ not intense system-wide networks. There have, of course, been system-wide networks of ports but these towns and cities have commonly been subordinate to the capital cities of their states. In contrast, contemporary world cities service global capital through the networks of large and powerful service corporations. The world city network constitutes a critical spatial framework of monopoly powers underpinning contemporary globalization.²¹ 

GLOBALIZATION AS METAGEOGRAPHICAL TRANSITION

The combination of a relative decline of the states and the rise of world cities has led me to explore globalization as a metageographical transition.²² Metageographies are the unexamined spatial frameworks through which people order their knowledges of the world.²³ Taken for granted, metageographies appear in all manner of interpretations of our world without revealing their constraints on our thinking. This 'common-sense geography' has taken several forms in the history of the modern world-system. I have identified a topological metageography of trade routes in the mercantile period, a centripetal metageography of a progressive centre against stagnant (but raw material rich) non-progressive zones in the industrial period, and a mosaic metageography of nation-states in the recent consumer period. Each metageography underpinned the knowledge of its times. The topological framework described the (non-European) world in largely non-hierarchical terms as a cornucopia: the myth of 'unknown lands' lay behind exploration, the search for new trades. The centripetal framework was strictly hierarchical: the myth of a racial hierarchy lay behind imperialism, the search for new raw materials. The mosaic framework viewed the world as a collection of separate, nominally equal, societies: the myth of sovereignty lay behind international relations, the search for peace/pacification and new 'national' markets. The mosaic metageography was constructed as part of the nationalist movement that defined 'nations' as the 'natural' units of humanity²⁴ and turned states into nation-states. This theory of 'imagined communities'²⁵ has been physically represented by the world political maps that adorn classroom walls in schools across the world. It is the classic global space of places, a static image, a world patchwork of allegedly homogenous spaces. The core meaning of globalization is that it breaks out of this boundary-obsessed world. Hence I interpret it as a metageographical transition from the current mosaic spatial order to a new as yet unknown replacement. Following Castells I suggest that this will take the form of a future network metageography and that the world city network might well be an incipient expression of this new geographical imagination of a global space of flows. 

DISEMBEDDING SOCIAL KNOWLEDGE FROM IMAGINED COMMUNITY

The implications of this metageographical transition are profound for the social science trinity of sociology, economics and political science. The rise of the mosaic metageography through the first half of the twentieth century was accompanied by the creation of mosaic social knowledge.²⁶ A remarkable spatial congruence was assumed whereby social, economic and political processes all operated in exactly the same space, the sovereign territory of the state.²⁷ This embedded statism was the result of a nationalisation of social knowledge so that society became equated with 'national society', economy with 'national economy', and polity with 'national government'.²⁸ To nationalise was to naturalise: in these three cases the adjective 'national' is redundant within mosaic social science - how else could society be organised? For most of the twentieth century no alternative was contemplated but with globalization other options are commonly proffered. Ideas about global society, global economy and global governance have become commonplace destroying the simple certainties of mosaic social knowledge. The old spatial underpinnings of social knowledge can no longer be assumed; the metageographical transition presupposes a new network-based social science. 

CITIES AS ACTUAL COMMUNITIES

Cities are very different from nations and their states. Cities are actual communities, places where people live and work in an organised manner. Cities embody a continuity that states lack: dozens of states disappeared in the twentieth century; which was the last great city to disappear and when?²⁹ It follows that most cities are much older than their current states - the main exceptions are a few state-created capital cities. Even where states collapse, as in the recent case of Zaire/Congo, cities continue to keep going as Kinshasa has illustrated.³⁰ Where cities suffer a catastrophe as with the sacking of Antwerp in 1585, the citizens rebuild their community elsewhere in another city, in this case Amsterdam.³¹ As Hall has so comprehensively illustrated, cities are the cultural crucibles of civilization.³² From the Renaissance as a movement of the cities (Florence) to modernism (Paris) and even to 'country' music (Nashville), cultural production and innovations are pre-eminently city-based.³³ Cities are what hold civilizations together. This was most starkly illustrated in the targeting of nuclear missiles on cities during the Cold War: 'taking out' the cities was, according to Bunge³⁴ intended to 'decapitate' society. By this he meant that by eliminating the upper levels of the urban hierarchy, all the key services, both public and private, necessary for the organisation of society would be destroyed. More generally, cities are the crossroads of society and as such are cosmopolitan unlike nations that aspire to purity and often use states to further this end. No city can survive alone, they only exist within a network of flows that links them to other cities. Their raison d'être is the organisation of spaces of flows - of people, of ideas, of commodities, of information, of money, of services.

FROM NATIONAL ECONOMY TO CITY ECONOMY

Cities as networked entities have been the great victims of the nationalization of social knowledge behind territorial bounds.³⁵ Boundaries are anathema to the operation of networks and therefore to cities. This is exemplified by 'state-istics' that always emphasise attribute measures over relational measures.³⁶ Thus there is massive effort in all functioning states to measure aspects of the 'national economy' in order to implement policy within their sovereign territory. But this is all based upon the myth of the national economy. Here I follow Jacobs³⁷ and her identification of cities and their regions as the prime entities of economic change. She argues that the idea that economies are 'national' is a false assumption which she traces back to both mercantilists and their political economist critics. Sustained to the present day by macro-economists, they have created a 'fool's paradise' for policy and prediction.³⁸ Cities are where economies grow and develop and influence other areas. So-called 'national economies' are simply artificial amalgams of actual city economies. Of course, there are two city-states that have been pre-eminently successful in the world-economy - Singapore and, until recently, Hong Kong - but otherwise city economies have to take their chance in terms of the nature of the state in which they are located. Where one city dominates the economics in a state its large influence on statistics (and state politics) means that state policies will favour this city economy over other city economies within the state's territory. For instance, from at least the 1925 decision to return to the Gold Standard onwards, the UK government's economic policy has systematically sacrificed the economies of all other city economies in its territory in order to maintain the prosperity of the London economy. Thus, the Governor of the Bank of England in setting high UK interest rates has argued recently that the outcome of increasing unemployment in the north east of England (Newcastle) is worthwhile if this helps keep down inflation in the south east of England (London). Clearly these are two different economic entities that require different economic policies but which is politically inconceivable because of the myth of a national economy. This identification of city regions as the fundamental economic entities in the world-system reinforces the conjecture that cities are the strongest candidates to replace states in the transition from a mosaic knowledge to a network knowledge of social change. 

DISEMBEDDING RADICAL POLITICS FROM SOCIALIST STATISM

But there still remains the matter of politics, a thoroughly nationalised practice through emphasis on gaining access to the powers of the state. Like innovations in cultural and economic activities, political innovations - the development of new political movements - have been pre-eminently produced in cities but this feature has been generally been lost in the rise of the democratic nation-state. The classic case is the story of the nationalization of the socialist movement from its nineteenth century anti-state origins to its great statist 'betrayals' in the twentieth century starting with the debacle of 1914.³⁹ This political turnaround has been truly phenomenal but is often forgotten: socialism has actually come to mean 'state control' through the practices of both its reformist and revolutionary wings in the twentieth century. The key process was an organisational one.⁴⁰ By converting from movement to party, socialists, despite many dire warnings, became incorporated into state political processes as 'national parties' aiming for control of the state apparatus. In the late nineteenth century debates on this strategy there was one political movement that could never be incorporated via statism: the anarchists. Their anti-state position was not just an end, it included all means to that end. However, in addition, they also encompassed an anti-urbanism that was part of a widespread contemporary negative reaction to the industrial urbanisation of the times. Normative blueprints of rural idylls were their stock in trade and where the urban was incorporated it was usually built upon nostalgic visions of the supposed freedoms of medieval town guilds. Bookchin⁴¹ has reinserted cities into anarchist thinking and placed them at the very centre of his analysis by contrasting 'urbanization' as the mere geographical aggregation of people to 'cities' as locations of participatory citizenship. His recasting of anarchism has some affinity to my bringing cities to centre stage, which might be described as a 'world-systems anarchism'. However, while agreeing with the promotion of citizenship in cities, we do not need to dismiss states and the many gains made for ordinary people through the nationalised politics of the parties. It seems to me that a two pronged strategy is needed by treating the states as sites of defence of past gains and the cities as sites of creativity for future politics. The latter requires invention of a popular network politics to replace the simple territorial politics we have become so accustomed to in our old-fashioned mosaic world. 

SERVICING GLOBAL SURVIVAL: A CITY-CENTRED TRANSITION

Let's go back to Figure 1. Here we have a description of a global network of cities created for contemporary global capitalism. This is, of course, a very partial description of city connectivity: cities are much more than capital service centres no matter how creative the service products.⁴² Can this network, created to help global capitalism as it careers towards unsustainability, be useful to those concerned for societal survival? There are parallels here with the state in the nineteenth century that Marx interpreted as 'the organising committee of the bourgeoisie' but which subsequently came to be a useful tool for democratic forces. World cities can be said to constitute an 'organising committee for the global bourgeoisie' in the twenty first century but, as a social infrastructure, use of these nodes and their connectivity does not have to stop with capitalism. Already some of the cities have been the sites of anti-global capitalism protests - Seattle, Prague, Washington, London, Berlin have become a battle roll call. Cities are where things happen, where things get done and therefore they are the sites of protest and petition. They also combine features of the space of flows and the space of places: as well as being global nodes, cities are home to billions of people. It is here that global capitalism comes face to face with democratic forces in a political dialectic to resolve the tensions between flows and places. And the intensity of this tension is expressed in acute economic polarisation that is a key feature of world cities: those that service global capitalism for huge salaries are themselves serviced by people on very low wages.⁴³ The latter are sometimes referred to as the 'third world within' but I think it best to conceptualise this in world-systems terms as the creation of a new semi-periphery at the very heart of the world-economy. This semi-peripheralisation of the core is possible precisely because the mosaic space of states as places is giving way to new spaces of flows in a networked world. It is clearly no accident that the most interesting politicians of our times are city mayors leaving national politicians looking more and more like figures from reruns of old movies.⁴⁴ 

THE END OF THE PAST AS WE KNOW IT: WE HAVE NEVER BEEN CAPITALIST

World-systems analysis treats history as a dialogue between the present and the past with an eye on the future. Hence the past is inherently unstable as both the present situation alters and our views of the future change. In the case of this recasting of world-systems analysis for the demise phase of the modern world-system, this allows us to go forward to the past.⁴⁵ Reinterpreting the past from a city-centred world-system requires us to revisit the equating of modern world-system with capitalist world-economy. Looking at the modern world-system through city lenses, it would seem that states have taken much more out of the system than they have put in. In fact the system does not look all that capitalist given that a major portion of the upper strata have not been primarily concerned with ceaseless capital accumulation. Rather they have been pursuing very traditional political pursuits of living off wealth producers via monopoly of the tools of violence, a practice we can trace back to Sargon and his men swooping down on the cities of Mesopotania over 4,000 years ago.⁴⁶ From this perspective, the world-system looks like a political economy hybrid, a unique system created in the long sixteenth century (c.1450-1650) in which a balance between the political/military and the economic/market was achieved to the medium-term benefit of both sides. This was ultimately made possible by the success of the Dutch cities in the first half of the seventeenth century in both harnessing a global space of flows and creating a territorial shell behind which their wealth could be protected. This was the first time a geographically concentrated set of cities successfully defended their wealth from the men on horseback thus setting the system on a path to capitalism.⁴⁷ But it was only a path. The necessary compromise between politics and economics remained so that with the nationalization of the state, cities were effectively eliminated as major powers within the system.⁴⁸ Until, of course, contemporary globalization and the new global space of flows resurrecting cities as outlined above. As we move towards a period of transition to a new system, the obvious question is whether our unique 'political economy' is becoming uncoupled under conditions of contemporary globalization? Is this the beginning of the process of societal bifurcation that Wallerstein predicts⁴⁹ providing a 'choice' between a new politics of place around centralisation/decentralisation (political) issues and a new politics of flows around market/monopoly (economic) issues?